The Language of the Sangleys: A Chinese Vernacular in Missionary Sources of the Seventeenth Century 9004184937, 9789004184930

Citation preview

The Language of the Sangleys

Sinica Leidensia Edited by

Barend J. ter Haar Maghiel van Crevel In co-operation with

P.K. Bol, D.R. Knechtges, E.S. Rawski, W.L. Idema, H.T. Zurndorfer

VOLUME 98

The Language of the Sangleys A Chinese Vernacular in Missionary Sources of the Seventeenth Century

By

Henning Klöter

LEIDEN • BOSTON 2011

Cover illustration: ‘Sangleyes, o Chinos’ (part of the Carta Hydrographica, y Chorographica de las Yslas Filipinas, shelf mark: Maps K. Top. CXVI. 37), courtesy of the British Library. This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data Klöter, Henning, 1969The language of the Sangleys : a Chinese vernacular in missionary sources of the seventeenth century / by Henning Kloter. p. cm. — (Sinica leidensia; v. 98) Includes bibliographical references and index. ISBN 978-90-04-18493-0 (alk. paper) 1. Chinese language—Middle Chinese, 1200-1919. 2. Chinese language—Dialects— Hokkien 3. Southern Min dialects. 4. Missions—Linguistic work. I. Title. II. Series. PL1081.K56 2010 495.1’724—dc22 2010039398

ISSN 0169-9563 ISBN 978 90 04 18493 0 Copyright 2011 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing, IDC Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

für Chin-hui

CONTENTS Figures and Tables ....................................................................... Acknowledgements ..................................................................... Technical Conventions ................................................................ A. Abbreviations .................................................................... B. Symbols ............................................................................. C. Typographical Conventions ............................................... D. Rendering of Handwritten Diacritics ................................. E. Transcriptions and Translations .........................................

xi xv xix xix xix xx xxi xxii

PART ONE EARLY MANILA HOKKIEN: MANUSCRIPTS, LANGUAGE AND METALANGUAGE Chapter 1: Introduction ................................................................ A. The Topic of the Study ...................................................... B. Hokkien Dialects ............................................................... C. The Arte de la Lengua Chio Chiu ...................................... 1. Extant Manuscripts and Authorship .............................. 2. Year of Compilation ...................................................... 3. Place of Compilation ..................................................... 4. Previous Documentation and Research ......................... D. Missionary Linguistics ...................................................... 1. Missionaries as Fieldworkers ........................................ 2. Research Directions and Topics .................................... 3. Language, Metalanguage, and Influence ....................... 4. Perspective and Purpose ................................................ 5. Reported Titles and Extant Sources .............................. E. Editorial Principles ............................................................. F. Research Questions and Chapter Division .........................

3 3 3 6 6 8 8 9 11 11 11 12 14 15 16 17

Chapter 2: Hokkien Dialects in European Sources....................... A. Introduction ....................................................................... B. Regional Vernaculars in Missionary Linguistics: Previous Research .............................................................

20 20 21

viii

CONTENTS

C. Analytical and Ideological Dimensions of Missionary Linguistics ................................................. D. The Sixteenth Century: ‘…All That They Doe Write is by Figures’ ...................... E. Dominicans in the Philippines, or: ‘Hay Particular Lengua’ .................................................... 1. EMH Documentation: Sources and Contexts ............... 2. The Analytic Approach of the EMH Sources ............... F. Jesuits and Mandarin, or: ‘Even the Women and Children Understand It’ ................ 1. The Linguistic Tradition of the Jesuits ......................... 2. Tradition, Strategy, or Environment? ........................... G. The neglect of Chinese Dialects, or: ‘They Speak Clownishly’ ................................................. H. Protestant Contributions, or: ‘An Independent Language’ ............................................. I. Summary and Concluding Remarks ................................... Chapter 3: Lexicography ............................................................. A. Introduction ....................................................................... B. EMH Sources: Reported Titles and Extant Sources .......... C. The Dictionarium Sino Hispanicum .................................. 1. Historical Context ......................................................... 2. Lexicographic Make-up ................................................ D. A Lost Dictionary: The Diccionario de la Lengua Chincheo .......................... E. The Bocabulario de la Lengua Sangleya ........................... F. Two Anonymous and Undated Dictionaries in the UST Archives .......................................................... G. Summary and Concluding Remarks .................................. Chapter 4: Language and Metalanguage ..................................... A. Introduction ....................................................................... B. Paradigm and Purpose ....................................................... C. Explanatory Devices: From Non-Terminology to New Terminology ................. 1. Non-Terminological Devices ........................................ 1.1. Macrostructure ...................................................... 1.2. Written Representation of Data ............................

23 26 31 31 33 34 34 38 42 46 48 51 51 52 56 56 59 66 68 73 81 83 83 84 87 90 90 92

CONTENTS

ix

2. Avoiding Terminology .................................................. 3. Questioning of Existing Terminology ........................... 4. Terminological Innovation ............................................ 4.1. Modes of Pronouncing .......................................... 4.2. Classifiers as ‘Proper Numerals’ ........................... D. Constructed Language? ..................................................... E. Summary and Concluding Remarks ..................................

97 98 100 100 104 106 110

Chapter 5: Phonology and Orthography ...................................... A. Introduction ....................................................................... B. Tone Diacritics and Tones ................................................. 1. Tone Diacritics .............................................................. 2. Tone Categories ............................................................ 3. Phonetic Properties of EMH Tones ............................... 4. Notation of Tones in Other EMH Sources .................... 5. EMH Tones ................................................................... C. Consonants ......................................................................... 1. EMH consonant spellings and Hokkien consonants ..... 2. Initial and Final Consonants .......................................... 2.1. Initial and Final .............................................. 2.2. Initial and Final .............................................. 2.3. Initial ............................................................ 2.4. Initial and Final .............................................. 2.5. Initial .............................................................. 2.6. Initial .............................................................. 2.7. Initial .............................................................. 2.8. Initial and Final ............................................. 2.9. Initial and final ............................................... 2.10. Initial and Final .......................................... 2.11. Initial and Final

............................................ 2.12. Initial .......................................................... 2.13. Initial ............................................................ 2.14. Initial ........................................................... 2.15. Initial and Final ............................................. 2.16. Initial ............................................................ 2.17. Initial ............................................................ 3. EMH Consonant Phonemes .......................................... D. Vowels ............................................................................... 1. EMH Vowel Spellings and Hokkien Vowels ................

112 112 113 113 115 121 126 129 130 131 132 132 133 135 136 137 137 138 139 139 139 140 140 141 141 142 143 143 144 145 145

x

CONTENTS

2. Interpretation of .............................................. 3. Interpretation of ........................................... 4. Interpretation of and ..................................... 5. Nasalization ................................................................... 6. Other Diacritics ............................................................. 7. EMH Monophthong Phonemes ..................................... E. Summary and Concluding Remarks ..................................

145 146 148 149 150 152 153

Chapter 6: Early Manila Hokkien: A Mixed Dialect .................. A. Introduction ....................................................................... B. New-Dialect Formation ..................................................... C. The Sangleys ..................................................................... D. Linguistic Evidence: Methodological Considerations ...... E. Lexical Evidence ............................................................... F. Phonological Evidence ...................................................... 1. Tones ............................................................................. 2. Finals ............................................................................. 2.1. Zhāngzhōu Affinity ............................................... 2.2. Zhāngzhōu and Cháoshàn Affinity ....................... 2.3. Zhāngzhōu and Quánzhōu Affinity ....................... 2.4. Open Issues ........................................................... G. Summary and Concluding Remarks ..................................

155 155 155 159 160 162 166 166 167 167 168 169 169 172

PART TWO THE ARTE DE LA LENGUA CHIO CHIU (BMS): TRANSCRIPT AND ANNOTATED TRANSLATION Transcript and Translation ..........................................................

176

Appendix: Tone Marking in the Arte and in Douglas (1873) ............................................................ Bibliography ................................................................................ Index ............................................................................................ Plates

371 381 403

FIGURES AND TABLES A. FIGURES Chapter Three Figure 1: Dictionarium Sino Hispanicum, f. 1 ....................... Figure 2: Example sentences of the Dictionarium Sino Hispanicum .........................................

60 64

Chapter Four Figure 1: Arte, f. 7r ................................................................. Figure 2: Arte, f. 5r .................................................................

94 95

B. TABLES Chapter Three Table 1: Missionary sources of Mǐn dialects .......................... Table 2: Dictionarium, f. 10 ................................................... Table 3: Dictionarium, f. 40 ................................................... Table 4: Terms referring to pronunciation in the Bocabulario ............................................................. Table 5: Entries with postposed tau ....................................... Table 6: The representation of Hokkien lexicon in the Manila manuscripts ................................................. Table 7: The headword and its sub-headwords (Dictionario) ................................. Table 8: The headword and its sub-headwords (Dictionario) ................................. Table 9: The headword and its sub-headwords (Dictionario) ................................. Table 10: Entries in Nebrija (1495?), Molina (1571), and the Dictionario ...........................................................

53 61 62 70 72 74 75 76 77 79

xii

FIGURES AND TABLES

Chapter Four Table 1: ‘Modes of pronouncing’ in the BMS and LMS ........................................................

101

Chapter Five Table 1: Sequence of tone marks ........................................... Table 2: EMH tone diacritics and tones ................................. Table 3: EMH tone indication according to TLPA ................ Table 4: Mandarin tones and tone diacritics in Varo (1703) ................................................................... Table 5: Tone descriptions in the Arte ................................... Table 6: Tone description in Varo (1703) .............................. Table 7: Tone descriptions in the Arte and Varo (1703) .............................................. Table 8: Tone distinctions in the Bocabulario ....................... Table 9: Diacritics in the Dictionarium ................................. Table 10: Comparison of tone diacritics ................................ Table 11: EMH tones ............................................................. Table 12: Examples of EMH tones ........................................ Table 13: Hokkien consonants ............................................... Table 14: Initial .............................................................. Table 15: Initial ............................................................... Table 16: Spiritus asper for aspiration ................................... Table 17: Indication of aspiration in EMH sources ............... Table 18: Initial ............................................................. Table 19: Initial .............................................................. Table 20: Initial .............................................................. Table 21: Initial .............................................................. Table 22: Initial ............................................................... Table 23: Initial ............................................................. Table 24: Initial .............................................................. Table 25: Initial

.............................................................. Table 26: Initial ............................................................... Table 27: Initial .............................................................. Table 28: Initial ............................................................... Table 29: Initial .............................................................. Table 30: Examples of EMH consonants ............................... Table 31: Hokkien monophthongs .........................................

114 116 120 121 122 123 125 126 127 128 129 130 131 133 133 134 135 135 136 137 138 138 139 139 140 141 142 142 143 144 145

FIGURES AND TABLES

Table 32: EMH compared with four Hokkien dialects ................................................ Table 33: EMH vs. and Hokkien -o vs. -oo ...................................................... Table 34: Superscript for open nasals .............................. Table 35: Indication of open nasals in different EMH sources .................................................. Table 36: Use of superscript in the LMS ......................... Table 37: Spelling of and in EMH sources ................................................................. Table 38: Examples of EMH vowels .....................................

xiii 147 148 150 150 151 152 152

Chapter Six Table 1: Hokkien personal pronouns ...................................... Table 2: EMH tone descriptions compared with Medhurst (1832) ........................................................ Table 3: Nasalized -uinn and unnasalized -ui ......................... Table 4: Final -ue ................................................................... Table 5: Finals -an and -ang ................................................... Table 6: Finals -u and -i ......................................................... Table 7: Medials -i- and -u- .................................................... Table 8: Finals -e, -ee, -ei .......................................................

165 166 167 168 169 170 171 172

ACKNOWLEDGEMENTS This book is the revised version of my habilitation thesis submitted to the Faculty of East Asian Studies of Ruhr University Bochum in June 2009. For valuable comments on draft chapters, answers to many questions, and other support, I would like to thank Willem Adelaar, Astrid Alexander-Bakkerus, Ramón Arzápalo Marín, Wolfgang Behr, José Eugenio Borao Mateo, Georg Bossong, Pieter van den Broek, Chien Hung-yi 簡宏逸, Maghiel van Crevel, Fang Chen-chen 方真真, Raoul David Findeisen, José Antonio Flores Farfán, Lloyd Haft, Esther Hernández, Huang Tzu-fang 黃資芳, Koos Kuiper, Sim Lee, Lien Chinfa 連金發, Lin Chin-hui 林欽惠, Heinz Lohmann, Tomasz Majtczak, Sachiko Matsumoto 松本幸子, Javier Martinez, Cristina Monzón, Sven Osterkamp, Luisa Paternicò, Alain Peyraube, Horst Pietschmann, Emanuele Raini, Emilio Ridruejo, Thomas Cedric Smith-Stark (1948–2009), Reinhard Wendt, Jeroen Wiedenhof, Otto Zwartjes, and the anonymous reviewers of the book manuscript. This book is the result of several research projects funded by different institutions. Grateful acknowledgement is made to:  Most importantly, the Netherlands Organisation for Scientific Research (NWO), for a research grant in 2005–2007 (VENI project Early Descriptions of Southern Min; 275-70-015), without which the realization of this book would not have been possible;  The Faculty of Humanities of Leiden University, for financial and organizational support in 2005–2007;  The Office of Research and Development of the National Taiwan Normal University, for sponsoring the project A Description of 17th Century Southern Min Phonology and Lexicon (project number: 96091017) in 2008;  The National Science Council of Taiwan, for sponsoring the project The Earliest Southern Min Dictionary: Documentation

xvi

ACKNOWLEDGEMENTS

and Analysis of the Dictionarium Sino Hispanicum (project number: 97-2410-H-003-035) in 2008–2009. I would like to thank the following institutions for providing images, and for the permission to reproduce them in this book:  The University of Barcelona Library, for providing digital images of the Arte de la lengua chio chiu (Ms. 1027), reproduced in the Plates section in the back of this book;  The British Library, for providing a digital image of the vignette ‘Sangleyes, o Chinos’ (part of the Carta Hydrographica, y Chorographica de las Yslas Filipinas, shelf mark: Maps K. Top. CXVI. 37) and digital images of folios 313v/314r of the Arte de la lengua chio (Add. Ms. 25317), reproduced respectively on the cover of this book and in the Plates section in the back of this book; and  The Italian Ministry for Arts and Culture, for providing pictures of folios 1, 52, and 53 of the Dictionarium Sino Hispanicum by P. Petrus Chirino (1604), held by the Biblioteca Angelica in Rome (Ms. 60), reproduced in Chapter 3 of this book. Any type of unauthorized reproduction of these pictures is forbidden. I would also like to thank the John Benjamins Publishing Company (Amsterdam and Philadelphia; www.benjamins.com), for their kind permission to reproduce (parts of) revised versions of two of my previous publications. Chapter 3 of the present study is a revised version of: Klöter, Henning (2009) “The Earliest Hokkien Dictionaries”, in Missionary Linguistics IV: Lexicography. Selected Papers from the Fifth International Conference on Missionary Linguistics, MéridaYucatán, 14–17 March 2007, edited by Otto Zwartjes, Ramón Arzápalo Marín, and Thomas Smith-Stark, 303–330. Amsterdam, Philadelphia: John Benjamins.

Some sections in Chapters 2 and 4 are based on: Klöter, Henning (2007) “‘ay sinco lenguas algo diferentes’: China’s Local Vernaculars in Early Missionary Sources”, in Missionary Linguistics III: Morphology and Syntax. Selected papers from the Third

ACKNOWLEDGEMENTS

xvii

and Fourth International Conferences on Missionary Linguistics, Hong Kong/ Macau, 12–15 March 2005, Valladolid, 8–11 March 2006, edited by Otto Zwartjes, Emilio Ridruejo, and Gregory James, 191–210. Amsterdam, Philadelphia: John Benjamins.

Finally, grateful acknowledgement is made to the Chiang Ching-kuo Foundation for International Scholarly Exchange, for providing a generous publication subsidy in 2010 (project number: SP004-U-09).

Henning Klöter Leiden and Bochum, July 2010

TECHNICAL CONVENTIONS A. ABBREVIATIONS 1 2 3 A BMS BR Ch cl. EMH f. fn. LMS Mnd. p. c. pl PMS Q r RMS sg TLPA v Zh

first person second person third person Amoy 廈門 dialect Barcelona manuscript of the Arte Blair and Robertson (1903–1909), numbers following BR refer to the volume number Cháoshàn 潮汕 dialect classifier Early Manila Hokkien folio footnote London manuscript of the Arte Mandarin personal communication plural Paris manuscript of the Dictionarium Sino Hispanicum (Chirino 1604a). Quánzhōu 泉州 dialect recto side of a folio The Rome manuscript of the Dictionarium Sino Hispanicum (Chirino 1604a) singular Taiwan Language Phonetic Alphabet verso side of a folio Zhāngzhōu 漳州 dialect B. SYMBOLS

 --

A check mark indicates that an expression occurs in the source. A double hyphen in a table indicates that an expression does not occur in the source.

xx ~ 

| = ≈

/ ° ?

TECHNICAL CONVENTIONS

A tilde separates: (a) variant spellings of the same word; (b) different dictionary translations of the same word A right-pointing triangle precedes my English translations of example sentences which are based on the Hokkien sentence. These are added when a Spanish rendering does not clearly distinguish gloss and translation (explained in Chapter 4, used in Part II). A vertical stroke separates different linguistic representations in quoted dictionary entries. An equal sign is written between a variant character and a standard character (used in Part II). An approximately equal sign is written between an EMH form and a form recorded in another Hokkien source if the two are not identical (explained in Chapter 1, used in Part II). A forward slash indicates a line break in a manuscript. A superscripted circle precedes a form posited on the basis of lexicographic conventions. A question mark indicates an uncertain reading or interpretation of language data or an uncertain date. C. TYPOGRAPHICAL CONVENTIONS

«text»

text ‘text’ [text]

{text} text […]

Guillemets enclose (parts of) entries quoted from dictionaries and the Arte that include the use of two or more scripts or languages Angled brackets enclose original spellings. An italicized expression indicates a linguistic form. Single quotation marks indicate a meaning. Square brackets indicate: (a) interpolations within quotations (b) phonetic transcriptions in IPA (c) modern data when compared with historical data (d) conjectural readings (used in Part II only) Curly brackets indicate that text in a manuscript was added or inserted, presumably by another hand. Strikethrough indicates that one or more words in a manuscript are crossed out. Three periods in square brackets stand for one or more

xxi

TECHNICAL CONVENTIONS

[x] [xxx] {xxx}

words in a manuscript that are missing and not conjecturable. An x in square brackets stands for an illegible letter or character in a manuscript. Three x’s in square brackets stand for one or more illegible words in a manuscript. Three x’s in curly brackets stand for one or more added or inserted illegible words. D. RENDERING OF HANDWRITTEN DIACRITICS

See explanations in Chapter 1 (Section E). Handwriting

Description horizontal stroke

Rendering cūn

grave accent

nò

háček

bǎn

acute accent

cuá

circumflex

sû

vertical stroke

làg'

spiritus

cùnc

two dots

kē:

superscript n

chéngn

xxii Handwriting

TECHNICAL CONVENTIONS

Description superscript nn

Rendering kêngnn

Diacritics 1–5 may occur in isolation, diacritics 6–10 always occur in combination with one or more other diacritics. In my conversion, diacritics 6–10 are always attached to the transcribed syllable. I have decided to render the handwritten spiritus in row 7 with a superscript c, as the shape of the latter more closely resembles the original handwriting. The two dots (row 8) are represented by a colon written after the vowel to which the diacritic applies. Due to technical reasons, there are some situations where what appears in the manuscript cannot be as well rendered in electronic media. For example, when a macron or longer line appears above a syllable containing the letter , or when a diacritic is handwritten above a syllabic . In such cases, I have rendered the diacritic independently, before the grapheme to which the diacritic applies. E. TRANSCRIPTIONS AND TRANSLATIONS Romanization of Mandarin expressions follows the Hànyǔ Pīnyīn 漢 語拼音 system (DeFrancis, 1996:1341–1349). Southern Mǐn/Hokkien expressions are transcribed according to the Taiwan Language Phonetic Alphabet (Tung, 2001). To achieve uniformity, and enable comparisons across different sources, all transcriptions quoted from Southern Mǐn sources, such as Douglas (1873), have been given according to the TLPA system. Taiwanese place names are spelled according to international convention. The transcription of Taiwanese personal names is according to private usage if known to the author. All translations are my own, unless otherwise indicated.

PART ONE EARLY MANILA HOKKIEN: MANUSCRIPTS, LANGUAGE, AND METALANGUAGE

CHAPTER ONE

INTRODUCTION A. THE TOPIC OF THE STUDY In the first decades of the seventeenth century, a Spanish missionary based in the Philippines wrote a grammar entitled Arte de la lengua chio chiu (hereafter: Arte). The two extant manuscripts of this grammar are considered the oldest documented grammatical analysis of a Chinese language. Part I of this study is a historical and linguistic analysis of the Arte, followed by a transcript of the major parts of the original text and an annotated English translation in Part II.1 B. HOKKIEN DIALECTS The place name in the title of the Arte very likely refers to Zhāngzhōu 漳州 district in the province of Fújiàn 福建, which is located on the south-eastern coast of China. One piece of evidence is the entry «chión chiú: provinçia | 漳州», found in a dictionary that was likewise compiled by a Western missionary in the Philippines during the seventeenth century (Dictionario, f. 111r; see Chapter 3). Under the entry , the name is added. The dictionary entry thus contains the same two names and that occur in the first paragraph of the Arte (see Part II, pp. 176–177). In some Spanish sources, the different spellings distinguish the district Zhāngzhōu from its district capital. Very often, however, the use of the two place names both in Spanish and also in Portuguese sources is rather inconsistent (for details, see Boxer, 1953:313–326). In general, however, any link to Fújiàn province supported by textual evidence is consistent with the recorded language data, as the variety documented in the Arte can unambiguously be identified as belonging to the Mǐn 閩 group, the group of Chinese dialects spoken in Fújiàn. The subdivision of Mǐn dialects remains a matter of controver-

–––––––– 1

Folios 17r–22v of the Arte are a list of Hokkien numbers; these are only partly documented in Part II.

4

CHAPTER ONE

sy. Traditionally, the northern group (Mǐnběi 閩北) is distinguished from the southern group (Mǐnnán 閩南). Some linguists, however, argue in favor of a major division between a western and an eastern group (Norman, 1988:233) or an inland and a coastal group (Branner, 2000:109–116). Speakers of Southern Mǐn dialects traditionally live in an area in Fújiàn province that stretches some 200 miles from the modern city of Quánzhōu, via Zhāngzhōu down to the Cháoshàn 潮汕 area in modern Guǎngdōng 廣東 province. The Cháoshàn dialect region comprises Cháozhōu 潮州 and Shàntóu 汕頭 prefectures. The data in the Arte have a strong affinity with those analyzed as Southern Mǐn, including the varieties of Xiàmén 廈門 city, referred to as Amoy, and the districts of Zhāngzhōu 漳州 and Quánzhōu 泉州. According to Ethnologue figures (Lewis, 2009), there are currently more than 25 million Southern Mǐn speakers in China, more than 15 million in Taiwan and some four million speakers in Southeast Asia. In reference to the Southern Mǐn dialects of Southeast Asia, the name Hokkien is commonly used. It derives from the Hokkien pronunciation hok4-kien3 of Fújiàn. As the Arte was written among overseas Chinese of the Philippines, the name Hokkien is used in this study as well. Historically, the Mǐn group is known as the first Chinese variety to have split off from Old Chinese. The cultural and linguistic sinicization of Fújiàn province is associated with different waves of migration between 100 BC and 900 BC (Norman, 1979, 1991; LaPolla, 2001; Lǐ, 2005; Dīng, 2006, 2007; Kwok, 2006). In modern dialects, many phonological and lexical archaisms, as well as non-Sinitic vocabulary, are preserved. Typologically, Hokkien dialects, like other Sinitic languages, are tonal and have a certain degree of analytic features, together with little derivational morphology (Chappell, 2001:4). Hokkien dialects and those of other Chinese varieties, including Mandarin,2 Yuè 粵 (Cantonese), Wú 吳, Kèjiā 客家 (Hakka), Gàn 贛, and Xiāng 湘, are not mutually intelligible. Within the Mǐn group, linguistic diversity prevails, and there is a low degree of intelligibility be-

––––––––

2 The term Mandarin is ambiguous, as it can refer to the Mandarin dialects collectively referred to as běifānghuà 北方話 (cf. Norman, 1988:181ff.) and the modern standard language based on the pronunciation of the capital Běijīng 北京. In reference to the modern standard language, Mandarin corresponds to the Chinese terms pǔtōnghuà 普通話 (lit. ‘common language’) used in China and guóyǔ 國語 (lit. ‘national language’) used in Taiwan. In premodern contexts, Mandarin equals the Chinese term guānhuà 官話 (lit. ‘official’s language’).

INTRODUCTION

5

tween speakers of northern and southern dialects. As I argue elsewhere (Klöter, 2009), the fact that local varieties are labeled as dialects (fāngyán 方言) in China, despite their mutual unintelligibility, has in the first instance ideological connotations. Southern Mǐn dialects have a written tradition distinct from written Mandarin and classical Chinese. The oldest sources are printed editions of stage plays in different Southern Mǐn dialects. The oldest extant edition is the Lìjìng jì 荔鏡記 ‘Story of the Lychee and the Mirror’ of 1566. The written representation of linguistic regionalisms in these editions is based on an adaptation of the Chinese character script. These popular writing conventions never enjoyed cultural prestige and were also never standardized (for details, see Klöter, 2005:58–87). Some of the plays have been edited by Wú Shǒulǐ 吳守禮 (Wú, 2001a–d, 2002a, b); linguistic analyses of the data have been published by Lien (2000, 2001b, 2004, 2006a–c, 2008, 2009, 2010) and Zēng Xiàntōng (1991). The lack of official recognition of local vernaculars like Southern Mǐn is also reflected in the history of native Chinese language studies. It was not before the early nineteenth century that phonological analysis of Southern Mǐn data led to the compilation of local rhyme dictionaries (for details, see Klöter, 2005:64–71). A Chinese tradition of grammar writing was only initiated in the late nineteenth century.3 It is thus safe to claim that the documents left behind by the missionaries are by far the oldest systematic analyses of Southern Mǐn dialects. State-of-the-art modern linguistic research on Mǐn dialects reflects a somewhat paradoxical situation. Zhāng’s (2004) bibliography of Mǐn dialect research comprises more than 8,000 Chinese, Japanese, and Western titles. On the other hand, as I have pointed out previously (Klöter, 2005:252), a comprehensive grammatical overview is still missing.

–––––––– 3

The first Chinese scholar to write a Chinese grammar was Mǎ Jiànzhōng 馬建忠 (1844–1900). His grammar entitled Mǎshì wéntōng 馬氏文通 ‘Basic Principles for Writing Clearly and Coherently by Mister Ma’ (1898) was strongly influenced by Western grammar (Peyraube, 2001).

6

CHAPTER ONE

C. THE ARTE DE LA LENGUA CHIO CHIU 1. Extant Manuscripts and Authorship One of the two extant manuscripts of the Arte is kept in the British Library in London (hereafter: LMS), the other in the library of the University of Barcelona (hereafter: BMS). Measuring 14.5 by 10.3 centimeters, the LMS comprises 23 double-sided folios; the BMS has 32 folios (64 pages) measuring 20.5 by 14.5 centimeters. As the BMS is more comprehensive and also more systematic in the documentation of language data, both parts of this study are based on this manuscript. Thus, any reference to the Arte is by definition to the BMS, unless indicated otherwise. Facsimiles of the first two folios of the LMS and of the BMS sections transcribed and translated in Part II are shown in the Plates section in the end of this book. The titles of the two documents are almost identical, viz. Arte de la lengua chiochiu (LMS) and Arte de la lengua chio chiu (BMS). As pointed out above, the toponym chiochiu or chio chiu very likely refers to the Zhāngzhōu district in China’s Fújiàn province. The title page of the BMS bears the title ‘Gramatica Chino’ (‘Chinese Grammar’) and the quotation ‘hombre nacido de mujer’ (‘Man, who is born of a woman’, Job 14:1). Although the names of two persons can be found in the text, neither of the manuscripts provides safe evidence of authorship. The very first line of f. 1r says: ‘For the use of Fr. Raymundo Feyjoó of the Order of the Preachers.’4 Folch points out that the Arte once belonged to the Dominican monastery of Santa Caterina (1995:158). Raymundo Feyjoó was a member of the monastery, which is evidenced by a list of obituaries of members of the monastery. The year of his death is not noted. As Feyjoó is identified as a monk of the lowest rank, Folch assumes that he died young while preparing for a stay in Asia, presumably using the Arte to study the language (ibid.). The second name, fr. Melchior de Mançano, appears as a signature at the very end of the BMS (f. 44r). Chappell and Peyraube assume that de Mançano (also spelled Melchor Manzano, ?1579–?1630) is the author of the Arte (2006:976). This assumption is compatible with historical data pointing out that Melchior de Mançano was a Dominican missionary based in Manila during the early seventeenth century.

–––––––– 4

In order to avoid redundancy, Part I of this study quotes my English translations of the Arte only. The original quotations can be found in the transcript in Part II.

INTRODUCTION

7

His name can be found on a list of missionaries who, in 1605, embarked for the Dominican mission in Manila known as the ‘Provincia del Sancto Rosario de la Orden de Predicatores’ (‘Province of the Holy Rosary of the Order of Preachers’). The entry in the list reads as follows (Aduarte, 1605; trans. BR 14:87): Father Fray Melchor de Mançano, a native of Villaseusa de Aro, a son of the convent of Santo Domingo at Ocaña, aged twenty-six years, nine years in the order, studies completed.

Mançano arrived in Manila in 1606. In 1617, he was elected Head and Superior of the province (Aduarte, 1640; trans. BR 32:67). Aduarte writes (ibid.): He was a very prudent and devout character, a professed son of the convent of the order in Ocaña; and had been made, on account of his great ability and his successful studies, a theologue at the college of Sancto Thomas at Alcala. In this province he had governed many of the best convents with great approbation; and his term as provincial was very useful to the province, augmenting it greatly, as will be narrated.

Blair and Robertson add that, in 1621, Mançano ‘was appointed procurator of the province at Madrid; and he died in Italy, about 1630, as Bishop-Elect of Nueva Segovia’ (BR 32:67, fn. 14). In 1627, he was appointed Censor of the Inquisition (Folch, 1995:158). After his return from Manila, Mançano successfully petitioned to King Felipe IV to relax some of the strict decrees that the Spanish colonial administration had enforced against the Chinese population. For example, a decree issued by the King in 1627 explicitly mentions Mançano’s intervention. The passages of the decree read as follows (Contreras, 1627; trans. BR 22:166–167): Fray Melchor Manzano, of the Order of St. Dominic, has reported to me, in behalf of the Sangley Chinese living in the Parián outside the walls of the city of Manila, that they experience much extortion and injury, on account of not only what pertains to the Christianity that they profess, but their liberty, possessions and honor, by making them cut their hair when they become Christians—a thing regarded as ignominious by their nation, and which is an obstacle to their conversion […] Fray Melchor has petitioned me that, attentive to the above, I be pleased to order that those converted to our holy Catholic faith be not obliged to cut their hair […]

Hence, historical evidence not only confirms that Mançano was based in the Philippines during the early seventeenth century; historical doc-

8

CHAPTER ONE

uments also point to his close connections with the Chinese community of Manila. All these pieces of evidence, however, do not prove his authorship of the Arte. One argument speaking against the authorship of Mançano is the place of the signature in the manuscript. It occurs at the very end, preceded by ten blank folios. If he had signed as an author, we would expect less distance between the place of the signature and the main text. In light of the administrative authority he gained in 1617, he may have signed the Arte as a sign of formal approval for usage, or as the owner of the manuscript. Folch suggests that Mançano may have signed in his capacity as Censor of the Inquisition (1995: 158). Thus, in short, I have found no convincing evidence in favor of Mançano’s authorship, and both manuscripts are therefore listed under ‘anonymous’ in my bibliography. 2. Year of Compilation No date of compilation is indicated in the manuscripts. On f. 31r of the BMS, however, one important hint can be found. A section on counting and dates contains the example «Bǎng lèg' sỳ chàp' pê' nī | 萬 曆四十八年» ‘48th year of the Wànlì emperor.’ According to historical records, the 48th year was the last year of the Wànlì reign, corresponding to the year 1620 of the Gregorian calendar. The Wànlì emperor died on 18 August 1620 (Hucker, 1976:324); the Wànlì reign formally ended on 27 August 1620. This does not, however, imply that the Arte was written between the beginning of the 48th year of the reign (4 February 1620, according to the Gregorian calendar) and its official end. It must have taken some time before news of the emperor’s death reached the Chinese community in Manila. Even if they had known quite early of the events in the Chinese capital, they could still have stuck to the old reign for the indication of dates. Thus, if we accept that the example «Bǎng lèg' sỳ chàp' pê' nī | 萬曆四十八年» reflects the year in which the Arte was written, we can conclude that it was written in 1620 or, less likely, in the first days of 1621. 3. Place of Compilation I have previously mentioned that the Arte was written in the Philippines. This claim is sufficiently supported by internal evidence. For example, Manila is the only place name occurring in example sen-

INTRODUCTION

9

tences (on ff. 9v, 13r, 14r). As importantly, on ff. 2v and 13r, the Arte mentions the ‘Sangleys’ (also spelled Sangleyes in other sources). The word commonly referred to the Chinese in the Philippines. The origin of the word ‘Sangley’ remains unknown, proposed etymologies include siang5 lai5 ‘constantly coming (from China to Manila)’ and sing1-li2 ‘trade’ (cf. Boxer, 1953:260, fn. 2). Boxer assumes, however, that the word is of non-Chinese origin. He furthermore points out (ibid.): [T]he term [Sangley] was originally applied by the Spaniards to Chinese in general but more particularly to those who came from Fukien [Fújiàn] ports. In the seventeenth century, the word gradually came to mean Chinese who were resident in the Philippines (as distinct from the visiting traders to whom it had first been applied) or those who had Filipino blood in their veins.

The Chinese migration to the Philippines and the social interaction of the Sangleys with Spanish colonizers and missionaries on Philippine soil has been analyzed in various historical studies (e.g., Horsley, 1950; Weightman, 1960; Felix, 1966, 1969; Zēng, 1998; Chang, 2002; Chia, 2006; Fang, 2006; Yáng, 2007). The broader historical context of the Arte is thus firmly established in previous research. However, many specific questions associated with the Arte as a historical document in its specific social and intellectual contexts have thus far remained unaddressed, as I will point out presently. 4. Previous Documentation and Research Thus far, two transcripts and two translations of parts of the Arte have been published. Without exception, the previous documentation is exclusively based on an analysis of the Spanish metalanguage, which inevitably results in transcription and translation errors. The first translation of the Arte (LMS) is Theophilus Siegfried Bayer’s (1694–1738) partial Latin translation ‘Grammatica Linguae Sinicae popularis in Provincial Chin Cheu,’ included in his famous Museum Sinicum (1730:137–160). The Museum Sinicum is ‘a collection of theoretical essays, long and short, on the Chinese language, literature, grammar, origins of the script, lexicography, dialects, and materials leading towards a full-fledged dictionary, based largely on the works of earlier Jesuits’ (Honey, 2001:24). According to Lundbæk, Bayer included the grammar, as he ‘felt it was important as an example of a language or dialect in which the sounds and the tonal

10

CHAPTER ONE

system were very different from those of the standard Mandarin’ (1986:129). The Taiwan-based Yuen Ze University (Yuánzhì Dàxué 元智大學) provides a scan of a black and white photocopy of the BMS on its homepage (YZU, 2007). As the photocopy is of insufficient quality, various passages of the manuscript are illegible. The scan is complemented by transcripts and translations of the legible sections. As these are few, the result is at best fragmentary. A major weakness of the transcript lies in the fact that no attempt was made to render the original romanized transcriptions. A better transcript of the text can be found in Ishizaki (2006). His edition includes the original romanized transcriptions, including the tone marks. However, apart from some erroneous renderings, most of which result from indifference towards the Hokkien data, the quality of the edition suffers from a lack of fidelity to the original arrangement. As will be shown subsequently, the arrangement of example sentences in the Arte is closely linked with the analysis of sentence structure. Thus, the disruption of the original arrangement in Ishizaki’s transcript conceals an important analytical dimension of the original document. Moreover, his critical apparatus is exclusively devoted to comparisons of the original Spanish spellings with modern Spanish orthography and thus of little relevance to sinology and Chinese linguistics. In contrast to the earlier documentation, my transcript and translation presented in Part 2 of this study follows an integrative approach. This approach is based on the premise that the Spanish metalanguage must be read in close conjunction with its object of analysis, i.e., the language of the Sangleys. Furthermore, the language data must be reanalyzed in the context of the descriptive and analytical framework applied in the Arte. In linguistic research, the Hokkien data documented in missionary sources have likewise received surprisingly little attention. Exceptions are Yue-Hashimoto’s (1991) article on stratification in Hokkien, as well as the publications on diachronic Sinitic grammar by Chappell (2000, 2006) and Chappell and Peyraube (2006). Previous research owes much to the documentary and analytic foundation laid by van der Loon’s seminal article The Manila Incunabula and Early Hokkien Studies (van der Loon, 1966, 1967). Although van der Loon’s analysis includes references to the LMS and Hokkien dictionaries, his focus lies on language data selected from Hokkien translations of Christian

INTRODUCTION

11

doctrinal manuals and prayer texts. On the basis of this relatively short collection, he reaches remarkably precise conclusions in the field of early Hokkien phonology and its orthographic representation (1967: 144–186). As the focus of his research lies on non-linguistic sources, however, the analysis of linguistic metalanguage does not receive much attention. This is mainly due to the fact that van der Loon was apparently unaware of the existence of some important linguistic documents analyzed in this study, notably the BMS and two comprehensive dictionaries analyzed in Chapter 3. These sources testify to the fact that early Hokkien grammars and dictionaries represent the richest linguistic documentation of any non-Mandarin Chinese vernacular prior to the nineteenth century. D. MISSIONARY LINGUISTICS 1. Missionaries as Fieldworkers The Norwegian linguist Even Hovdhaugen once referred to missionaries as ‘the first professional fieldworkers’ (1996:14). In various respects, this is an apt description of a largely neglected aspect of missionary work. After the sixteenth century, missionaries established churches in parts of the world which had previously been unknown in Europe. As it is at the core of missionary work to communicate with other people, they were obliged to learn the languages spoken in their immediate surroundings. In this respect, missionary work can indeed be compared to linguistic fieldwork: Missionaries had to listen carefully to speakers of unknown languages, they had to analyze unknown sounds, words, and sentences; and they had to convert spoken languages into some kind of written form. This aspect of missionary work ultimately resulted in a huge body of linguistic documentation of the world’s languages, such as dictionaries, grammars, vocabularies, teaching manuals, etc. In many cases, these missionary documents remain the only sources of previous stages of a particular language available today. 2. Research Directions and Topics Internationally, in the past years, the contributions of missionaries to the documentation and analysis of the world’s languages have attracted increasing scholarly attention. International conferences on

12

CHAPTER ONE

missionary linguistics have been held regularly since 2003. The conference series has resulted in four collections of essays (Zwartjes and Hovdhaugen, 2004; Zwartjes and Altman, 2005; Zwartjes, James, and Ridruejo, 2007; Zwartjes, Arzápalo Marín, and Smith-Stark, 2009). Two editions of missionary linguistic documents, Alexander-Bakkerus (2007) and Zwartjes (2009), have been published in the new ‘Lingüística misionera’ series. In Chinese linguistics, the English translation of the first printed Mandarin grammar (Varo, 1703) by Coblin and Levi (2000) is a significant contribution to the field.5 On the basis of the English translation, a Chinese translation has been published (Yáo and Mǎ, 2003). Publications by Chappell (2000), Chappell and Peyraube (2006) and Masini (2000) on Southern Mǐn; Chappell and Lammare (2005) on Hakka; and Breitenbach (2004, 2005), Coblin (1996, 1997, 1998, 2003, 2006) and Masini (1996, 2003, 2005) on Mandarin also testify to the growing academic interest in diachronic data of Chinese languages as documented by Western missionaries. The abovementioned contributions focus on the interpretation of historical language data in missionary documents. This approach can be characterized as diachronic language research through a re-reading of historical metalanguage. Another disciplinary direction within missionary linguistics is of a more sociolinguistic nature, focusing on the multiple roles of missionaries in the history of European-Asian encounters. Their primary role as proselytizers is not only inextricably linked to their language research, but also to their roles as translators and mediators of theological and philosophical concepts and their roles as educators and, by extension, as representatives of European colonial powers. This approach is pursued in recent studies by, inter alia, Errington (2001, 2008), Gilmour (2006), and Vollet and Castañeda (2004). 3. Language, Metalanguage, and Influence My analysis of the Arte takes a comparative approach that is not restricted in regional terms. Linguistically, as pointed out, the data recorded in the sources unquestionably belongs to the Southern Mǐn

––––––––

5 It is widely held that Varo’s Mandarin grammar, posthumously published in 1703, was the first printed Mandarin grammar in history. However, Paternicò (2010 forthcoming) convincingly argues that Martino Martini’s (Wèi Kuāngguó 衛匡國, 1614–1661) grammar was published in 1696 already, as an appendix to the second edition of Melchisedec Thévenot’s Relations des divers voyages curieux.

INTRODUCTION

13

group of Sinitic languages. However, the analytical framework and the terminology applied for the analysis of data in the Arte is by no means bound to a particular region. As missionary grammar and dictionary writing did not start from the scratch, any extant source devoted to the analysis of a language was to some extent influenced by previous scholarship. Thus, when examining the use of explanatory devices in a particular work like the Arte, it must be asked whether a device was influenced by some other work or whether it influenced linguistic analysis in a later period. A common distinction in this context is that between language and metalanguage. Koerner defines metalanguage as ‘the use of language for the description of linguistic concepts, ideas or theories of earlier periods’ (1989:31; see also Koerner, 1987). He distinguishes three criteria upon which a particular source of influence can be identified. The first criterion refers to connections between an ‘author’s background, family tradition, schooling, studies and particular interests and pursuits during his formative years […] that may lead to evidence of […] borrowing, integration and assimilation of particular ideas, concepts, or theories.’ The second criterion relates to ‘textual parallels between a particular theory or concept and supposed sources.’ Thirdly, ‘the most important evidence in favor of a claim of influence may result from direct references by an author to the work of others’ (1989:40–41). Many studies deal with the influence of the Spanish grammarian and lexicographer Antonio de Nebrija (also known as Elio Antonio de Lebrija, Antonius Nebrissensis, etc.; 1441–1522) on missionary linguistics.6 Whereas the significance of Nebrija should not be denied, this study shows that a narrow comparison of Nebrijan terminology with the metalanguage of a particular work hides other sources and directions of influence. In particular, it must be emphasized that the compilation of the Arte involved communication between its compiler and one or more Hokkien native speakers. This in turn leads to the question whether this communication entailed the transfer of Chinese linguistic concepts. Here it must be kept in mind that the language of the Sangleys differed in many respects from Latin and Castilian, the

––––––––

6 A few examples include Braselmann (1991), Breitenbach (2000, 2004, 2005), Breva-Claramonte (2000), Calvo Pérez (2000, 2005), Esparza Torres (2000, 2002, 2007), Esparza Torres and Calvo Fernández (1994), Hernández (2009), Manrique (2000), Monzón (2000), Percival (1994/2004, 1997/2004, 1999/2004), Smith-Stark (2009), Suárez Roca (2000), Sueiro Justel (2007) and Zwartjes (2000, 2002).

14

CHAPTER ONE

languages upon which Nebrija’s analytical framework was based. This leads to the question as to how the compiler analyzed these unfamiliar language features and whether Chinese linguistic traditions were of any significance in this analysis. 4. Perspective and Purpose Studies on language analyses by missionaries often contain quality statements which are based on comparisons of past scholarship with modern research. Emphasizing the ‘mistakes’ in a particular source, such statements tend to carry a patronizing undertone, arguing that the missionaries in the old days did quite well, but were as yet not able to recognize such-and-such phenomena. If restricted to such quality statements, this analytic approach certainly misses important dimensions of missionary language documentation. Thus, in order to achieve a more nuanced reading, a particular mode of explanation must be seen in close connection with the analyzed language as its object of analysis, and vice-versa. The analysis of the interplay of explanandum and explanans should moreover by complemented by an examination of the social contexts in which a particular source was used. To be sure, missionary sources, like modern linguistic analyses, are not free of doubtful explanations, inconsistencies, and mistakes; and it would certainly be wrong to leave them unmentioned. What I refute, however, is a perspective that is overly corrective. In other words, ‘wrong’ and ‘erroneous’ are not very productive, yet at times unavoidable categories in analyses of missionary linguistic sources. In many cases, however, they reflect a prescriptive bias deriving from the allegedly objective knowledge of modern language structure and reliability of modern linguistic methodology. The focus of my analysis lies on historical language data. Thus, in a very broad sense, the following chapters aim at examining the language spoken by the Sangleys in the early seventeenth century. Methodologically, the examination of Sangley language data is inextricably linked to an analysis of the way the data is documented and analyzed in the Arte and other sources. As a proper understanding of missionary metalanguage requires comparisons with other sources, my analysis, as pointed out, includes references to various other missionary grammars, including grammars of non-Sinitic languages. It must be emphasized, however, that my aim is not to compare as many instances of the usage of a certain term as possible. Instead, my comparison will be

INTRODUCTION

15

limited to those examples which are relevant for the interpretation of the language data. 5. Reported Titles and Extant Sources Written sources are at the core of missionary linguistics. The search for and consultation of sources is therefore an indispensable part of any research project. In order to find documents, however, especially rare and unedited manuscripts, one often has to embark on long and tedious journeys; and even optimal preparation of such journeys does not guarantee that a particular manuscript can actually be consulted. This leads to an inevitable conflict. On the one hand, any serious analysis of unedited manuscripts requires the consultation of the original sources. On the other hand, however, due to the enormous time and budget expenditure required for the search of manuscripts, it is beyond the capacity of any research project to strive for completeness. It should therefore be a maxim of document-based research to work within strict time constraints and to strictly distinguish between reported titles and extant sources. This, in turn, has important consequences for the presentation of results. Any qualitative statement, such as ‘oldest,’ ‘first,’ ‘most comprehensive,’ etc., implies the constraint ‘within the range of extant sources.’ By extant, I refer to those sources which can be unambiguously identified with regard to their present whereabouts. A reference in a bibliography, as I argue in Chapter 3 of this study, is an insufficient criterion for claiming that a certain source actually existed. In other words, missionary linguistics is, by definition, an exercise in stocktaking. Thus, the validity of any hypothesis posed on the basis of one particular source will be subject to verification and continuous reevaluation when more sources have been discovered. We are therefore reminded of the bon mot of the linguist William Labov who wrote that ‘[h]istorical linguistics can […] be thought of as the art of making the best use of bad data’ (1994:11). By analogy, we can conceive of missionary linguistics as the art of offering the best hypotheses on the basis of incomplete documentation.

16

CHAPTER ONE

E. EDITORIAL PRINCIPLES My reluctance to subscribe to a corrective perspective has important implications for the presentation of data. The chapters in the first part of this study quote many phrases from handwritten sources. The second part contains a transcript of most parts of the Arte, likewise a manuscript. My documentation of handwritten data is based on the principles of diplomatic editing which, according to Williams and Abbot (1999:71; see also Boyle, 1976): […] aims to reproduce a manuscript […] as a historical artifact. It presents a text as it was available at a particular time in a particular document. Such editing is noncritical in that it does not emend the text, even a text that may not accurately reproduce an author’s words.

Following this principle, the transcript in Part II leaves the original arrangement of examples intact. As argued in Chapter 4, an editorial rearrangement would inevitably conceal important analytic dimensions associated with the original document. Inevitably, however, any attempt to represent handwritten data in typed text faces various limitations. One example is the conversion of the handwritten romanized transcriptions of Hokkien words. In contrast to modern romanization systems, diacritics in the handwritten romanization do not occupy a fixed spot or size in relation to the letters of the word. For example, although most diacritics occur above a word or letter, a diacritic may also appear next to a word or letter. The same line may sometimes be as long as a written word, and sometimes as short as one letter. As analyzed in Chapter 4, some diacritics also occur in combination, but there is no fixed convention for the arrangement of combined diacritics. On the one hand, modern font editing software could certainly cope with these challenges of document editing. On the other hand, however, any device specifically designed for the purpose of editing the paleographical features of the Arte would have limited compatibility outside its specific software environment. My edition is therefore based on the principle that legibility of converted diacritics should not depend on special software requirements. Following this principle, my digital conversion of handwritten diacritics only uses signs which are available in common software (see Conventions, pp. xxi-xxii). This decision is motivated by the fact that the data recorded in the Arte represents only a small fraction of all extant early Hokkien data re-

INTRODUCTION

17

corded in other sources. It is hoped that future projects will result in the release of digital databases based on other sources. These databases should in turn convey a clear impression of how the data was represented in the original manuscripts. Using converted transcriptions of handwritten words, I present the historical data as closely as possible as they are recorded in the sources. For example, if a certain tone can be posited on the basis of my interpretation of the transcription system, the tone is indicated as such. If a comparison with other reference works suggests that the tone indication in the Arte may have been incorrect, my notes juxtapose tone recording in the Arte and tone indication in other sources by using an approximately equal sign ≈ (see Conventions). Thus, a formula like in the Arte stands for phah8 ‘beat, strike’, which is recorded as phah4 ‘strike’ in other Hokkien reference works. Chinese characters are also recorded as written in the Arte, even if a particular character is ‘incorrect’ from a philological perspective. A systematic analysis of character use would be beyond the scope of the present study. Therefore, in the notes of Part II, only those characters are listed which are not attested in historical dictionaries. Next to these, I indicate—without any implication with regard to etymological correctness—which attested characters resemble the ‘peculiar’ Arte characters most closely. F. RESEARCH QUESTIONS AND CHAPTER DIVISION Although the focus of the present study will be on the Arte, I do not analyze it as an isolated phenomenon. Instead, the chapters which follow place the Arte into broader contexts. As pointed out above, at the beginning of the seventeenth century, Western missionaries based in the Philippines devoted themselves to the study of the spoken language of the Sangleys. As I will discuss in Chapter 2, within the broader context of Chinese language studies by Europeans prior to the eighteenth century, this particular choice of an object of analysis is quite unusual. Specifically, if we compare the Arte and its related documents with language studies by the Jesuits based in China and missionary linguistics in the Americas, it becomes obvious that the Arte represents a distinct chapter in the history of Chinese language studies by Europeans. The analysis in Chapter 2

18

CHAPTER ONE

therefore compares different traditions in Chinese missionary linguistics by addressing the following research questions: (a) What is the object of analysis in linguistic documents? (b) Is the object of analysis in linguistic documents associated with a particular order? (c) What is the reason behind the choice of this particular object of analysis? (d) Does the choice represent a new direction in Chinese missionary linguistics, or is it a continuation of an earlier tradition? The documentation of Hokkien dialects by Western missionaries in the seventeenth century is by no means restricted to the Arte. Chapter 3 explores the quantitative and analytical dimensions during the earliest stage of Hokkien language documentation. As all other extant documents are dictionaries, the analysis will focus on lexicography. One purpose of this chapter lies in taking stock of extant sources. As the inclusion of two hitherto unanalyzed dictionaries shows, Hokkien dictionary compilation reached greater dimensions than observed in previous research. The lexicographic approach manifest in the extant titles is rather heterogeneous. In this chapter, I highlight the analytic dimensions of lexicographic arrangements by discussing the following questions: (a) Which influences are visible in the lexicographic arrangement of the dictionaries? (b) What is the relationship between lexicographic arrangement and language analysis in the dictionaries? Chapter 3 concludes that linguistic analysis goes beyond written words and therefore does not require explicit terminology. This conclusion is picked up in Chapters 4 and 5 which examine linguistic metalanguage and other explanatory devices employed in the Arte. Chapter 4 addresses the common critique that the use of linguistic metalanguage by missionaries is an inappropriate enforcement of traditional European grammatical paradigms. It is argued that this critique disregards the multiple roles of the missionary linguist; it also ignores various explanatory devices which are not bound to the appli-

INTRODUCTION

19

cation of an existing terminology. The following questions are examined: (a) Which explanatory devices are employed in the Arte? (b) Which sources of influence are manifest in these devices? (c) Does the analysis yield results with regard to the analyzed language that are consistent with the data in Hokkien sources of the seventeenth century which were not compiled by missionaries? Chapter 5 pursues the interplay of explanandum and explanans further by analyzing phonological aspects of the language of the Sangleys through the lens of missionary metalanguage. These questions will be addressed: (a) What do the letters and diacritics of the transcription system used in the Arte reveal about the phonology of the recorded language? (b) Are conclusions drawn on the basis of the previous question consistent with earlier research on diachronic Hokkien phonology? (c) How consistent is the transcription of Hokkien sounds in the extant sources? On the basis of the findings of Chapter 5, Chapter 6 asks: (a) Which Hokkien dialect(s) did the Sangleys speak? It is claimed that, in light of the internal diversity within the Southern Mǐn group, a general identification of the data as ‘Hokkien’ or ‘Southern Mǐn’ is insufficient. On the basis of lexical and phonological evidence, the chapter challenges a previous hypothesis according to which the Sangleys spoke a Hokkien variety of one particular place in the district of Zhāngzhōu. Employing the sociolinguistic notion of new dialect formation, I argue that the data represent a distinct Hokkien variety formed through the contact of migrant speakers of different social and geographical origins. The resulting contact language should be treated as an independent Hokkien variety, to be labeled Early Manila Hokkien.

CHAPTER TWO

HOKKIEN DIALECTS IN EUROPEAN SOURCES A. INTRODUCTION This chapter places the Arte in its historical context. This contextualization will show that peculiarities of the Arte will only become visible in comparison with sources compiled by missionary linguists in China during the same period, notably those of the Jesuits. The characteristics of Jesuit scholarship, in turn, can only be properly understood when compared with missionary linguistics outside China, especially in the Americas. The very fact that the object of analysis of the Arte is a regional vernacular has important implications for the present analysis. After all, although China is known for its high degree of linguistic diversity, historiographers of Chinese linguistics tend to take for granted that missionaries focused on the three linguistic codes associated with China’s political elite, i.e., Mandarin, the classical Chinese written language wényán 文言, and the Chinese character script. As exemplified by the Arte and other EMH sources introduced in Chapter 3, this is, however, certainly not the case. Hence a more nuanced approach is needed; an approach which systematically examines missionaries’ responses to multilinguistic environments and which thus questions the assumption—often taken for granted—that ‘Chinese’ refers to one of the three above-mentioned codes. The approach in this chapter thus goes beyond the chronological listing of linguistic sources on Chinese vernaculars. Instead, it also aims at presenting explanations for the choice of language in missionary linguistics and the influence of missionary linguistics on European scholarship after the seventeenth century. Integrating analytical and ideological aspects of missionary linguistics, the following questions will be addressed: What exactly is the object of analysis in a particular source? Which method is applied for its analysis? What is the motivation behind the choice of this particular object of analysis? Does the choice represent a new direction in Chinese missionary linguistics, or is it a continuation of an earlier tradition?

HOKKIEN DIALECTS IN EUROPEAN SOURCES

21

B. REGIONAL VERNACULARS IN MISSIONARY LINGUISTICS: PREVIOUS RESEARCH Previous studies on Chinese missionary linguistics are rather unspecific with regard to the varieties of Chinese recorded by Western missionaries. Frequently, it is simply stated that after the initiation of permanent work in China in the sixteenth century, missionaries started to learn and record ‘the Chinese language.’ This collective term implies that regional languages are closely related varieties of a single Chinese language. Such a monolinguistic view on China is rather common and not restricted to linguistic historiography. In China, this conceptualization is in turn shaped, to a large extent, by non-linguistic considerations. In the words of Ramsey (1987:16): The Chinese believe that they speak dialects of a single language not because they are unaware of the objective linguistic facts, but because of certain cultural considerations. Unlike the people who speak Romance, the Chinese are not divided into a number of national units corresponding roughly to the several groups of closely allied dialects. Rather, the Chinese language is spoken by a single group of people with a common cultural heritage. China is not only the most populous country on earth, it is also the oldest social institution, and the Chinese people belong to and follow cultural and national traditions that have continued since the days of the Han Empire and before.

The ‘myth of monolingualism,’ as Mair (2007) calls it, is also manifest in Chinese linguistics, both inside and outside China. Yuen Ren Chao, without a doubt one of the most eminent scholars in the field, argued in his landmark Grammar of Spoken Chinese that ‘the greatest degree of uniformity is found among all the dialects of the Chinese language’ (1968:13) and thus claimed that ‘one can say that there is practically one universal Chinese grammar’ (ibid.). Recent scholarship has refuted this claim, arguing that, as Matthews and Yip put it, Chao’s concept is ‘the result of several factors, including an inadequate database, motivated wishful thinking, and an anachronistic view of what grammar entails’ (2001:266). In a similar vein, many scholars now emphasize the diversity among China’s linguistic varieties and tend to speak of Chinese languages in the plural. The long-standing persistence of the monolinguistic view on China has important implications for the historiography of Chinese missionary linguistics. When we look at previous research on missionary contributions to the study of China’s languages, it seems that local ver-

22

CHAPTER TWO

naculars of the common people were a late discovery, as the historiography of Chinese missionary linguistics has thus far focused on missionary accounts of the languages of China’s elite. The initiation of missionary research of Chinese vernaculars, on the other hand, is often dated to the early nineteenth century. For example, the oldest source mentioned in Yóu’s (2002:122–191) inventory of Western missionary accounts of Chinese dialects is the first printed Hokkien-English dictionary by the English Congregationalist Walter Henry Medhurst (1796–1857), compiled at the beginning of the nineteenth century and published in 1832 (Medhurst, 1832). The Catholic discovery of the Chinese vernacular is taken to have happened even later. According to Yang’s (1960) inventory, the study of Chinese dialects by Catholic missionaries did not take off before the late nineteenth century, with the publication of the two-volume dictionary of the Shànghǎi 上海 dialect by the French Jesuit Paul Rabouin (1828–1896) in 1894 and 1896. The question of which language the missionaries recorded in their linguistic treatises has been interpreted in the context of different social philosophies of missionary work. Heylen, for instance, contrasts the elite-oriented Catholic tradition with the non-elitist approach of the Protestants. She argues that (2001:141): The Roman Catholics had based their proselytising strategy on adapting themselves to the cultural outlook of the educated Chinese elite. Attempts to reach out directly to the illiterate masses were not prevalent. By winning the respect of the scholar class, the Roman Catholic liturgy was written in literary Chinese or ‘court Mandarin’ […]. This literary style was […] unintelligible to the common people.

Assumed differences in missionary philosophies in terms of elitist/ Catholic vs. non-elitist/Protestant and previous research on Chinese missionary linguistics seem to be compatible. However, various studies have pointed out that the Catholic missionary linguistic tradition outside China by no means excluded local languages (e.g., Foertsch, 1998a, 1998b; Zwartjes, 2002; Ridruejo, 2005; Sueiro Justel, 2005). According to Foertsch (1998b:76), before their prohibition by the Pope in 1773, the Jesuits alone compiled 164 dictionaries, 165 grammars, 167 catechisms, and over 430 other texts in 134 languages and six dialects. Thus, within the global context of missionary linguistics, it seems hardly conceivable that missionary work in China could neglect local languages for more than two hundred years. Indeed, a

HOKKIEN DIALECTS IN EUROPEAN SOURCES

23

few studies (van der Loon, 1966, 1967; Masini, 2000) have shown that linguistic documentation by Catholic missionaries did include Chinese vernaculars. As will be explained presently, the oldest Western sources documenting Chinese regional vernaculars are almost as old as the oldest documentation of Mandarin by the Jesuits. Due to their spread along the south-eastern coast of China and, more importantly, in Southeast Asian Chinese communities, Hokkien dialects received the most attention in the early history of Chinese missionary linguistics. This chapter divides the history of Hokkien dialect research into two periods, both of which are associated with language studies by missionaries. The first period comprises the first decades of the seventeenth century, when missionaries, mostly of the Dominican order, documented the Hokkien dialects spoken by the Chinese population of the Manila area. The second period started in the first half of the nineteenth century, when Protestant missionaries of different orders recorded Hokkien dialects spoken in Southeast Asia and, after the enforced opening of Chinese cities in the aftermath of the First Opium War in 1842, in China itself. C. ANALYTICAL AND IDEOLOGICAL DIMENSIONS OF MISSIONARY LINGUISTICS This chapter distinguishes the analytical and the ideological dimensions of missionary linguistics. Missionary linguistics in its analytical dimensions looks at the interplay of the explanation and the thing explained in the sources, i.e., methods employed for the analysis of unknown phonologies and morphosyntax, the adaptation and redefinition of linguistic metalanguage, etc. As I have pointed out previously (Klöter, 2005:89–130, 2006), phonological analysis and orthography design were intimately connected aspects of missionary work. Apart from its analytic side, the use of the letters of the Roman alphabet for writing non-European languages also has an important ideological dimension. As alphabetic writing was associated with a widespread attitude of cultural superiority over non-European languages and their speakers, alphabetic writing became a visible symbol of cultural hegemony. In this respect it must be emphasized at the outset that missionaries were not only linguists who analyzed language structure, but also language planners who promoted linguistic standards. In Chapter

24

CHAPTER TWO

4, I will analyze links between methods of linguistic analysis and the multiple role of the missionary linguist. Both of these dimensions of missionary linguistics—the analytical and the ideological—are closely associated with the Spanish scholar Antonio de Nebrija (1441–1522). Nebrija’s Latin grammar Introductiones latinae (1481) and his Castilian grammar Gramática castellana (1492/1981)1 both influenced missionary linguistics in the following centuries. As will be discussed in Chapter 3, Nebrija’s linguistic influence is not restricted to grammar writing, but also manifest in missionary lexicography. After the Spanish discovery of the Americas, Nebrija’s grammars and dictionaries became important reference works for missionaries writing grammars of indigenous American languages. The second dimension of Nebrija’s influence, according to Mignolo, ‘is political and ideological and relates to the programs Nebrija attached to his grammar of Latin, on the one hand, and to his grammar and orthography of Castilian, on the other hand’ (1995:48). Ideological values associated with alphabetic writing were an idea ‘fermenting in Nebrija’ (ibid.:41) and formulated in more detail in the writings of Bernardo de Aldrete (1565–1645), especially in his famous treatise on the origins of the Castilian language (1606). Following the footsteps of Nebrija, Aldrete’s ‘framework was bounded […] by Latin as the language of learning and civility, and by Castilian as the language of the nation’ (ibid.:37). Closely related to this, he believed in ‘the connection between alphabetic writing and civility’ (ibid.:35). This connection, in turn, was the ideological basis for disrespect towards indigenous languages and script traditions reflected in missionary accounts of regional languages. Commenting on missionary linguistics in the Pacific area during the nineteenth century, Mühlhäusler observes a ‘perceived imperfection of languages and their speakers’ among missionary linguists (1996:141). He writes that indigenous languages were seen as ‘primitive’ and ‘repositories of wickedness and moral degradation’ (ibid.:140). It seems safe to assume that similar attitudes also prevailed when the missionary linguistic enterprise started to unfold after the discovery of the New World. It

–––––––– 1

Previous studies have pointed out that Nebrija’s Latin grammar was far more influential than the Castilian grammar. According to Mignolo, for example, the Latin grammar was held in most monastic libraries. It ‘went through no less than fifty editions during the author’s lifetime,’ while the Castilian grammar ‘was never reprinted until the second half of the eighteenth century’ (1995:49; see also Braselmann, 1991: 63–68).

HOKKIEN DIALECTS IN EUROPEAN SOURCES

25

has to be emphasized, however, that Nebrija-inspired ideology did not necessarily translate into disrespectful attitudes. In contrast to Mühlhäusler and Mignolo, Zwartjes denies that a feeling of cultural superiority led to negative attitudes towards languages not associated with European civilization. He points out that many missionaries ‘accept the cultural superiority of the Christian nations, but when analyzing the languages of the indigenous tribes, they admiringly talk about “copiousness” and “elegance” of these languages, often ranking them in high rank of refinement’ (forthcoming:13). One core element of Nebrijan language ideology never fully gained ground in the Americas, namely, the replacement of indigenous languages with Latin and Castilian. 2 As Mignolo points out, Nebrija ‘knew that the power of a unified language, via its grammar, lay in teaching it to the barbarians’ (1995:39). Missionary linguistics in the Americas was, however, ‘a curious inversion of Nebrija’s programs […]: while Nebrija proposed learning Latin as means of unification and consolidation of the Christian republic, the friars in the New World had the option of Amerindian languages as a means to fulfill the same goal’ (ibid.:54). In other words, when Christian texts were first translated by Jesuits and Dominicans into classical Chinese and Hokkien respectively, the path towards the use of non-European languages in missionary practice had already been opened in the Americas. What remained was ‘Nebrija Lite,’ so to speak, i.e., a tradition of alphabetic orthographies for, and analysis of, indigenous languages within a paradigm inspired by European tradition. Chapter 4 of this study will deal in more detail with the question of whether and how the analytical approach of the Arte was inspired by Nebrija, and to what extent other sources of influence can be determined. The acceptance of indigenous languages for religious practice yielded a growing scholarly interest in the world’s linguistic diversity. As Bossong argues, the discovery of the New World in its linguistic dimensions marks an important break in the history of linguistics. He writes (2007:124):

––––––––

2 To be sure, the sociolinguistic dominance of Spanish in present-day Central and South America goes back to the Spanish expansion after the sixteenth century. However, many local languages continue to be used alongside Spanish. Interestingly, by contrast, three hundred years after the beginning of Spanish rule in the Philippines, less than ten percent of the local population spoke Spanish (cf. Wendt, 1992:189).

26

CHAPTER TWO

The scholastic philosophers [of the Middle Ages] were not interested in specific languages, but in language in general. Their attitude can be summarised by an oft quoted dictum of Roger Bacon (ca.1220–1292): grammatica una et eadem est secundum substantiam in omnibus linguis, licet accidentaliter varietur (‘grammar is substantially one and the same in all languages, although it varies accidentally’) (Bacon, 1902 [ca. 1250]:27). […] From the very dawn of humanism in Italy, another approach was gaining ground. With Dante Alighieri (1265–1321), a new outlook on language began, leading to the Renaissance with its insistence on linguistic diversity.

As a consequence, ‘Latin ceased to be the exclusive yardstick of logic in language as its relativity became obvious in contact with unfamiliar linguistic worlds’ (ibid.:125; a similar point has been made by Oesterreicher and Schmidt-Riese, 1999:69). As pointed out above, the acceptance of indigenous languages as a medium of linguistic practice and, in consequence, of scholarly analysis at first followed practical considerations; it therefore does not necessarily represent a cultural revaluation of these languages in the eyes of European missionaries. Moreover, as Errington claims, the ‘missionaries sought to describe the languages in order to create literate colonial subjects, which meant that they transposed practices of literacy from European nations to “native” communities’ (2008:16). In this respect, however, the situation in the Americas and Asia was fundamentally different. As Bossong correctly claims, in contrast to the Americas, Asia ‘was the cradle of some of the oldest and most developed cultures; the Europeans had to face civilizations on the same level as, or a superior level to, their own’ (2007:126). As I show below, Bossong’s claim is amply supported by the example of the Jesuit presence in China. In the field of language learning, the Jesuit fascination for China’s culture translated into a focus on the languages of China’s cultural and political elite. This fascination for the ‘high’ languages and Chinese writing was so intense that it effectively suppressed any interest in linguistic diversity. This focus was, however, not initiated by the Jesuits, as I show in the following paragraph. D. THE SIXTEENTH CENTURY: ‘…ALL THAT THEY DOE WRITE IS BY FIGURES’ From the very beginning, European scholarly interest in the Chinese language was shaped by a fascination for Chinese character writing.

HOKKIEN DIALECTS IN EUROPEAN SOURCES

27

The Flemish missionary and explorer Willem van Ruysbroeck (William of Rubruck, ca.1215–1270) notes in the thirteenth century already that the ‘inhabitants of Cathay3 […] make several letters in a single character to make up one expression’ (quoted by Harbsmeier, 1998:9). 4 Harbsmeier claims that ‘[m]ore detailed knowledge about the Chinese language is almost entirely due to reports from the Jesuits’ (ibid.). It is certainly true that the Jesuits played a crucial role in the formation of early European conceptions of the Chinese language. This does not, however, imply that European scholarship, in the centuries between van Ruysbroeck’s remarks and the beginnings of the Jesuit mission in China in 1579, lacked any interest in the Chinese language. It also does not imply that the Jesuits were the only missionaries who documented and analyzed the Chinese language. Instead, parallel to the China-based Jesuits, missionaries of different religious orders had initiated linguistic research among overseas Chinese in Southeast Asia, most notably the Philippines. Their approach to Chinese language documentation differed fundamentally from that of the Jesuits, both in terms of method and with regard to the language chosen for analysis. Interestingly, sources on Chinese compiled outside China received much less attention among European scholars than those written by the Jesuits. Prior to the establishment of permanent missions in China and Southeast Asia, European perceptions of China were largely shaped by Juan González de Mendoza’s (ca.1540–1617) Historia de las cosas más notables, ritos y costumbres del gran Reyno de las China (hereafter: Historia). This book owed much of its popularity to the fact that it was, soon after its first publication in 1585, translated into seven different European languages ‘and had become one of the bestsellers of its day’ (Lach, 1965:743f.). The first English translation by Robert Parke appeared in 1588. As its author never visited China, Mendoza’s book cannot be considered a firsthand account. In 1580, he was a member of a group of Augustinian emissaries bound for China, but, due to political circumstances, the mission was suspended before it

––––––––

3 The word Cathay (Latin Cathaya) derives from Qìdān 契丹 (Khitan), the founders of the Liáo 遼 dynasty (916–1125 CE). Cathaya was used during the European Middle Ages (Wilkinson, 2000:753). 4 According to Nachod (1923:237) the first mention of a Chinese script was made by the Franciscan envoy Plano di Carpini, whose report of his visit to Asia was written a few years before Ruysbroeck’s remarks. Recent studies on early European encounters with the Chinese script include Rusk (2007) and Kim (2009).

28

CHAPTER TWO

could reach Asia (ibid.:747). So, when Mendoza was commanded by Pope Gregory XIII to write a book on China after the return from his failed mission, he had to rely on accounts previously published by others. This is why Mendoza, despite the great success and wide circulation of his Historia, was also ‘accused of telling tales, fabricating his data, and wholesale plagiarizing’ (ibid.:745). As shown in Lach’s detailed analysis (ibid.:742–794), the sources consulted by Mendoza likewise heavily relied on accounts written by others. Due to this circulation of facts and figures, it is often difficult to spot the original source behind a particular bit of information in the Historia. In two cases, however, it is possible to identify Mendoza’s use of first-hand accounts of Europeans who had actually been to China and who had written something on the linguistic situation. The first such account is the Tractado em se cõtam muito por estẽso as cousas da China (hereafter: Tractado) published by the Dominican friar Gaspar da Cruz (?–1570) in 1569. Cruz had been stationed in a number of Asian countries before his sojourn led him to the southern Chinese city of Canton (Guǎngzhōu 廣州), where he spent a few weeks in the winter of 1556. The Tractado, written after his return to Europe and shortly before his death, was the first European book exclusively devoted to China (ibid.:742; Boxer, 1953:ixii). However, as Boxer points out, the Tractado never reached a wide circulation, one reason being the limited attention to Portuguese sources outside Portugal (1953:lxiv). The second firsthand account consulted by Mendoza is the Relación of the Augustinian friar Martín de Rada (1533–1578) who, together with Jerónimo Marín (?–1606), led the first Spanish mission to China in 1575. For English translations of Cruz’ Tractado and de Rada’s Relación, I refer to Boxer (1953). Mendoza also consulted Bernardino de Escalante’s Discurso de la navegacion que los Portugueses hazen à los Reinos y Provincias del Oriente, y de la noticia q̃ se tiene de las grandezas del Reino de la China of 1577, which was the second European book on China. However, unlike Cruz and de Rada, Escalante never visited China (Lach, 1965:742). The chapters in Mendoza’s Historia cover a wide range of topics, such as China’s climate, rituals, temples, administration, etc. Information on language is rather scarce. Still, the few bits of linguistic information provided by Mendoza and the aforementioned sources seem to have set the tune for how European thinkers perceived of the Chinese

HOKKIEN DIALECTS IN EUROPEAN SOURCES

29

language in the seventeenth century. Most importantly, Mendoza mentions the Chinese script, which he characterizes as universally understood across linguistic boundaries. He claims that linguistic intelligibility only exists in writing, whereas the different spoken languages are mutually unintelligible. The important passages appear in Chapter 13 of the Historia, entitled ‘Of the characters and letters that the Chinos do vse, and of the colleges and scholes that are in this kingdome, and of other curious things’ (trans. Parke, 1588:92–95). The remarks on the Chinese script read as follows (p. 92): You shall finde verie fewe in this kingdome but can both write and reade, yet haue they not the alphabet of letters as we haue, but all that they doe write is by figures, and they are long in learning of it, and with great difficultie, for that almost every word hath his character. […] They do vse more than sixe thousand characters different the one from the other, and they doo write them verie swiftly […]; it is a kinde of language that is better vnderstood in writing than in speaking (as the Hebrue toonge), by reason of the certaine distinction of points that is in euery character differing one from the other, which in speaking cannot be distinguished so easilie.

It is almost trivial to notice that, from a modern linguistic perspective, the claim that formal and semantic distinctions only exist in writing and not in speech is simply wrong. This would imply that spoken Chinese lacks the properties of a natural language. It is, however, exactly this conception of Chinese as a huge inventory of universally understood written symbols which would fuel European discourses on a universal language in the seventeenth century. Mendoza’s idea that Chinese, due to the nature of the written language, can be understood across linguistic borders is reflected in the following lines (p. 93): […] one figure or character vnto them all doth signifie one thing, although in the pronouncing there is difference in the vowels. The character that doth signifie a citie is this , and in their languyage some doo call it Leombi, and others Fu,5 yet both the one and the other doo vnderstande it to bee citie; the like is in all other names. And in this order doo communicate with them the Japones, Lechios, those of Samatra, and those of the kingdome of Quachinchina and the borders

–––––––– 5

Mendoza quotes this example from Escalante (1579:30). In Mendoza’s quotation, is clearly identifiable as 城 chéng ‘city’. In Escalante’s book, on the the character other hand, similarity with any existing Chinese character is hardly recognizable. Note that none of the readings fu or leombi is attested for Mandarin. Escalante (or Mendoza) may have mixed up the characters 府 fǔ ‘district, prefecture’ and 城 chéng ‘city’. The possible non-Mandarin source of the reading leombi is unclear.

30

CHAPTER TWO

vnto them: whereas in their speech or language, there is no more vnderstanding then is betwixt Greekes and Tuskanes.

As pointed out, Mendoza’s reflections are not original, but rather reflect thoughts formulated in earlier sources. Prior to Mendoza, Cruz had described the linguistic diversity in China, the mutual unintelligibility among the various languages, and the important function of the written language as a universally-understood tool of communication. Cruz wrote that (1569; trans. Boxer, 1953:73): […] in China there are many differences of language, for the which reason many of them do not understand each other’s speech, yet they understand each other’s writing, as do likewise the inhabitants of the islands of Japan, who understand the Chinese through their writing although they have a different language.

In a similar vein, de Rada mentions that ‘the same document can be read in all the tongues of China’ (1575; trans. Boxer, 1953:297). He also notices that he ‘saw documents written in the court script which was different from that of Hocquien [Hokkien]’ (ibid.). This early reference to the Hokkien language is rather ambiguous. One possible reading is that the Hokkien language differs from the court language Mandarin. It is also possible that de Rada points to local Hokkien characters, which are based on Chinese writing traditions but distinct from standard Chinese characters (see Chapter 1; Klöter, 2005: Chapter 2). Despite its ambiguity, this remark is very likely the earliest explicit reference to the Hokkien language in a European source. The previous quotations on China’s regional diversity fit well into the bigger picture of the country’s geographical vastness, as sketched in all sources. Lach points out that ‘[l]ike Escalante and Rada, Mendoza divides “this mightie kingdome […] into fifteen provinces, that every one of them is bigger then the greatest kingdome that we doo vnderstand to be in all Europe”’ (Lach, 1965:755). Cruz devotes a whole chapter to ‘the provinces into which China is divided’ (trans. Boxer, 1953:88–91). This chapter also contains a few lines on a ‘province called Fuquem [Fújiàn],’ which is described as ‘one of the largest and noblest provinces’ (Boxer, 1853:88). However, no mention of the language of Fújiàn province is made. In short, European sources of the sixteenth century do not establish any links between linguistic diversity and China’s provincial division.

HOKKIEN DIALECTS IN EUROPEAN SOURCES

31

E. DOMINICANS IN THE PHILIPPINES, OR: ‘HAY PARTICULAR LENGUA’ 1. EMH Documentation: Sources and Contexts Systematic documentation of Hokkien dialects by Europeans started in the early seventeenth century. The extant documents owe their existence to the presence of European missionaries and Chinese Sangleys (see Chapter 1) in the Philippines. The Spanish-Chinese encounter in the Philippines contributed an important, albeit largely neglected chapter to the history of Chinese linguistics. For the first time, the friars based in the Philippines recorded the Hokkien language spoken by the Sangleys in various documents. As I will argue in Chapter 6, the language recorded in these sources should be treated as a separate Hokkien dialect, Early Manila Hokkien (EMH). The extant documents are similar to those from other countries where European missionaries recorded local languages. One group includes translations of Christian catechisms in the local vernacular. Translations of the Christian doctrine are known as Doctrina Christiana, the earliest of which were compiled in Central and South America in the mid-sixteenth century (Amado Aymoré, 2004). The second group of texts includes those which, in a very broad sense, analyze language. These can be simple word lists, comprehensive dictionaries or grammars. I refer to these texts as linguistic documents. The four extant EMH dictionaries are introduced and analyzed in Chapter 3. As pointed out in Chapter 1, the Arte must be considered the earliest extant linguistic treatise on any variety of Chinese. As the present study is primarily concerned with questions related to linguistic analysis, references to EMH translations of the Doctrina christiana will be kept to a minimum. For comprehensive studies, I refer to van der Loon (1966, 1967) and Hashimoto (1999). In contrast to the EMH Doctrina christiana, which was printed in the early seventeenth century (Keng, 1605?; facsimile in Gayo Aragón and Domíguez, 1951), all linguistic documents are manuscripts. Only one of these, the Dictionarium Sino Hispanicum, provides information about authorship and date of compilation. Its unnumbered title folio was signed by the Jesuit Pedro [Petrus] Chirino (1557–1635) on 31 March 1604 (for details, see Chapter 3). None of the other extant documents provide comparable information. It has been pointed out, however, that most linguistic documents are of Dominican provenance

32

CHAPTER TWO

(van der Loon, 1966, 1967; Breitenbach, 2000:30; see also Chapters 1 and 3). This does not imply that missionaries of other orders were not interested in language analysis. Most importantly, it is known that Franciscan friars also preached to the Sangleys in Hokkien (van der Loon, 1967:105f.). So far, however, no manuscript documenting Franciscan approaches to the study of EMH has been discovered.6 The fact that the Spanish-Chinese linguistic encounter took place on Philippine soil is no historical coincidence. It is mostly assumed that first Chinese contact with the Philippines took place during the Sòng 宋 dynasty (960–1279 CE). However, recent research claims that the earliest Chinese presence even dates back to at least 661 CE (Yáng, 2007). It was, in any case, only with the colonization of the Philippines by Spain in 1565 that a firm basis for Chinese trade and Chinese settlement was established (Weightman, 1960:47–64; Wickberg, 1965: 3). Wickberg reports that, in the early 1570s, the Spaniards ‘found a small settlement of about 150 Chinese’ in the Manila area. Within 30 years, this number soared to 20,000 (ibid.:4–6). In October 1603, the Spanish colonial government suppressed an uprising, resulting in the massacre of almost the entire Chinese population (ibid.:10; Chang, 2002). Although the number of Sangleys quickly recovered to the original level, relations between the colonial administrators and Sangleys remained hostile (ibid.:4ff.). This was largely due to a number of restrictive measures imposed on the Sangleys. For instance, out of fear of Chinese invasion or uprisings, the Spaniards unsuccessfully tried to limit the number of Sangleys to 6,000 (ibid.:11). Another regulation ordered the Sangleys to stay in their own residential areas.7 The dimensions of the tensions between the Spanish colonizers and the Chinese population cannot be underestimated. According to Horsley, the Spaniards, less than forty years after their initial settlement, ‘had formed a pattern of hatred against

–––––––– 6

The grammars of the Franciscan missionary Melchor Oyanguren de Santa Inés (1688–1747), who was also based in the Philippines, include various sections on Mandarin but no explicit reference to Hokkien documents (Klöter and Zwartjes, 2008; Zwartjes, 2009). 7 Historical research has focused on the residential and commercial area known as Parián, which was established in 1581 and relocated several times in the following decades (Chia, 2006:516). Leaving the Parián was subject to strict regulations (Weightman, 1960:68ff.; Wickberg, 1965:11; Chia, 2006:516). Other Chinese settlements in the Manila area, but outside the city walls, were located in Binondo, Tondo, Santa Cruz and Quiapo (Chia, 2006:520).

HOKKIEN DIALECTS IN EUROPEAN SOURCES

33

the Chinese that persisted for more than three centuries’ (1950:1). However, despite these hostilities, the presence of Spaniards and Sangleys in Manila provided a win-win situation for both sides. Both Spaniards and Sangleys made economic profits; the Spaniards moreover utilized the presence of Chinese to effectively prepare for the expansion of their mission to the Chinese mainland (analyzed in Wills, 1994; Menegon, 2009). 2. The Analytic Approach of the EMH Sources The extant sources represent EMH as a natural, living language, with its own phonological and morphosyntactic properties. As will be pointed out in Chapter 4, the linguistic analysis goes beyond the uncritical application of Greco-Latin grammatical paradigms. Instead, typological features which are not part of the Greco-Latin grammar, like tones and classifiers, are duly recognized. In this respect it must be kept in mind that the missionary compilers of the sources did not intend to feed any theory. Instead, the dictionaries and grammars were supposed to assist missionaries in learning a foreign language. To any missionary in the world, the importance of foreign language learning went beyond the linguistic translation of religious rituals. Communicative skills in foreign languages were arguably the most crucial foundation of missionary work. Without linguistic skills, the social and organizational challenges of any missionary enterprise would have been doomed to failure. In order to understand how the Manila-based missionaries conceived of the language of the Sangleys, it is not sufficient to look at what the sources say explicitly; one must also consider what appears implicitly. Most importantly, although all example sentences in the Arte are rendered in Chinese characters, not a single explanation is provided on the Chinese character script. Apparently, reflections on Chinese writing and its nature were remote from the communicative needs of everyday missionary life. In other words, EMH sources are completely irresponsive to the rising interest in Chinese writing among European scholars of the sixteenth century. The choice of a Chinese vernacular as an object of analysis and documentation is likewise noteworthy. To be sure, for the missionaries based among the Sangleys, Hokkien was the language spoken in their immediate surroundings. Hence, what could have been more obvious than learning Hokkien? From the first paragraph of the Arte it

34

CHAPTER TWO

becomes evident that the compiler was fully aware of the linguistic diversity in China. He also knew that Mandarin and not Hokkien was the dominant language of China’s cultural and political elite. The paragraph reads as follows: The common language of the kingdom of China is Mandarin, which is current in the whole kingdom; and also in the province of Chincheo, which has its own language. All those who can read understand Mandarin. The province of Chio Chiu has its own language, which is the one that is discussed here. But it should be observed that in this province there are five languages which are somewhat different from each other, as are Portuguese, Valencian, Aragonese, Castilian and others. The most common is that of Chio Chiu, the one that is most widely spoken here, for which reason the grammar and dictionary will be in this language, which is better than to mix them all up—just as one should learn the language of Toledo rather than mixing up Portuguese, Spanish, and so on. (trans. van der Loon, 1967:100; see Part II, pp. 176–177)

F. JESUITS AND MANDARIN, OR: ‘EVEN THE WOMEN AND CHILDREN UNDERSTAND IT’ 1. The Linguistic Tradition of the Jesuits The significance of the choice of Hokkien becomes more evident when we compare the linguistic approach of the Dominicans in the Philippines with that of the China-based Jesuits. The establishment of the Jesuit mission in China dates back to 1579 and, thus, almost exactly coincides with the first arrival of Dominicans in Manila in 1587 (Wills, 1994:115). In contrast to the Philippine-based friars, however, linguistic analyses of the Jesuits focused exclusively on the quasi-official court language Mandarin and the classical written language. This approach followed an explicit decision made by the Visitor Alessandro Valignano (Fàn Lǐān 范禮安, 1539–1606) even before the Jesuits had built their first church in China.8 Valignano not only insisted ‘on a regime of language studies that would precede proselytizing activities’ (Brockey, 2007:246), but, understanding that Mandarin was widely used among officials throughout the country, he also issued a clear order that the language to be learned was to be Manda-

––––––––

8 The office of Visitor entailed ‘an extraordinarily wide range of powers to oversee and govern the mission as he saw fit’ (Üçerler, 2003:340). Üçerler’s article is a concise biography of Valignano.

HOKKIEN DIALECTS IN EUROPEAN SOURCES

35

rin (Witek, 2001:154, see also Gāo, 2008). Within a few decades, Valignano’s order turned into a full-fledged language program. The first missionary who entered China and studied Mandarin following Valignano’s order was Michele Ruggieri (Luó Míngjiān 羅 明 堅 , 1543–1607). In the absence of bilingual textbooks, Ruggieri approached local teachers who reverted to native Chinese materials and methods of language teaching. Brockey points out that ‘[e]vidence preserved in the Roman Archives of the Society of Jesus shows that some of the first missionaries employed the same texts that Chinese children learned by rote to increase their vocabulary and practice writing characters’ (Brockey, 2007:247–249; see also Zhāng Xīpíng, 2009: 41–68). This ‘appropriation of indigenous study methods’ (ibid.) remained a characteristic of the Jesuit approach to Chinese language learning when a formal four-year course was established in the early 1620s. This new program included lessons in spoken Mandarin, the reading of the Confucian canon, and writing and composition (Brockey, 2007:257). The choice of Mandarin was later confirmed by Matteo Ricci (Lì Mǎdòu 利瑪竇, 1552–1610) who wrote (1615/1953:28–29): Besides the various dialects of the different provinces, the province vernacular so to speak, there is also a spoken language common to the whole Empire, known as the Quonhoa, an official language for civil and forensic use. This national tongue probably resulted from the fact that all the magistrates, as we shall explain later, are strangers in the provinces which they govern, and to avoid the necessity of obliging them to learn the dialects of the provinces, a common speech was introduced for transacting official government business. The Quonhua dialect is now in vogue among the cultured classes, and is used between strangers and the inhabitants of the province they may visit. With the knowledge of this common language, there really is no necessity for the members of our Society to learn the dialects of the provinces in which they work. A province dialect would not be used in polite society, although the more cultured classes might use it in their home province as a sign of neighborliness, or perhaps outside of the province from a sense of patriotism. This national, official tongue is so commonly used that even the women and children understand it.

This quotation contains some accurate sociolinguistic information. However, the remark about the widespread use of Mandarin outside literati circles should be taken with a pinch of salt. I will come back to this issue presently.

36

CHAPTER TWO

The Jesuit devotion to the study of Mandarin and classical Chinese yielded remarkable results in missionary linguistics. In the field of lexicography, the accommodationist approach of the Jesuits resulted in the landmark dictionary Xīrú ěrmù zī 西儒耳目資 ‘An Aid to the Ear and the Eye of Western Scholars,’ written by Nicolas (Nicola, Niklaas) Trigault (Jīn Nígé 金尼閣, 1577–1628) with the assistance of the Chinese scholars Hán Yún 韓雲 and Wáng Zhēng 王徴 in 1626 (Masini, 1996:236; a detailed study of the Xīrú ěrmù zī is in Chen, 1987; see also Lo, 1930b; Zēng Xiǎoyú, 1991, Sòng, 2004; Tán, 2006). This dictionary is not only entirely written in Chinese, it also incorporates various Chinese lexicographical traditions and phonological distinctions. The two main volumes of this three-volume dictionary are essentially listings of character readings in the Latin alphabet. The readings reflect, according to Coblin (1997:262; see also Lǔ, 2003, 2007), the standard Mandarin pronunciation of the late Míng 明 dynasty (1368–1644 CE). In the mid-seventeenth century, Jesuit missionaries also started to embark on the writing of Chinese grammars. The most important contributions in this field are the Notitia Linguae Sinicae of Joseph Henri-Marie de Prémare (Mǎ Ruòsè 馬若瑟, 1666– 1736) and the Grammatica Sinica of Martino Martini. Both works circulated as manuscripts during their authors’ lifetimes. It was not before the nineteenth century, however, that Prémare’s Notitia Linguae Sinicae was translated into English and printed by Bridgman (1847/ 2007; for details, see Lundbæk, 1991:64–102). Martini’s (1689) grammar was edited by Bertuccioli (1998). A comparison of the analytical approaches in these grammars and the Arte would be beyond the scope of the present study. It is, however, worth noting that these two works differ fundamentally from the Arte in their ample use of Chinese philological terminology. The employment of indigenous Chinese methods of language learning and language analysis comes as no surprise when seen in the broader context of Jesuit cultural accommodation in China. As a matter of fact, accommodation was the core characteristic of the Jesuit strategy in China. The Jesuits, according to Standaert, ‘adapted themselves to the lifestyle and etiquette of the Confucian elite of literati and officials’ (1999:352). This strategy of accommodation was inseparable from the fascination for Chinese culture shared by many Jesuits. In Mungello’s words, ‘China was in many ways, apart from its lack of Christianity, the equal or superior to Europe’ (2005:81). It is interesting to see how

HOKKIEN DIALECTS IN EUROPEAN SOURCES

37

the accommodationist approach permeated various aspects of the Jesuit encounter with things Chinese, including the choice of the language to study. The Jesuit approach to Chinese language analysis differs markedly from missionary linguistics outside China, most notably in their almost total neglect of the Nebrija-inspired Arte tradition of grammar writing. Although China-based Jesuits did write Chinese grammars, none of these works is entitled Arte (cf. Brockey, 2007:254). In the light of the Jesuit’s accommodationist approach, we may assume that the application of grammatical paradigms stipulated by the Arte tradition would have been incompatible with the adaptation and appreciation of traditional Chinese philology. Moreover, paradigms dictated by an Arte framework would have been of little use in the analysis of Chinese characters. All in all, therefore, the Jesuit contributions to Chinese language studies can be regarded as a prime counter-example to Mignolo’s generalizing claim that ‘the Jesuits would take Nebrija’s Latin grammar as one of their basic textbooks’ (2003:58). The disregard of the Greco-Latin model grammar is not the only way in which the Chinese-learning Jesuits distinguished themselves from missionary linguists of other orders. Another important difference lies in the high esteem for the Chinese language expressed by the Jesuits, an attitude which deviates from the previously mentioned ‘perceived imperfection of languages’ expressed in various missionary sources written in other parts of the world. However, some clarification of the term ‘Chinese language’ is required here. As pointed out previously, the Jesuit focused exclusively on learning Mandarin and studying the Chinese classics. This focus resulted in a lack of any serious interest in local vernaculars. To the best of my knowledge, during the almost 150 years of the Jesuit mission in China, not a single treatise on a Chinese vernacular was written by a Jesuit hand. Brockey points out that Jesuit sources ‘are ambiguous about the issue, offering few references to local dialects’ (2007:258). He writes that (ibid.): One exception is the testimony of Francisco Simões regarding a mission circuit in North Zhili, which suggests that some priests had considerable trouble with local dialects, at least in their first years in China. Most, if not all, of the area’s villages seemed to have their ‘own particular ways of speaking, tantamount to different languages,’ Simões lamented. Such diversity could be found even in hamlets separated by ‘one league or less,’ he complained, making it appear that ‘the Devil mixed the lan-

38

CHAPTER TWO

guage up in such a way as to make preaching the Holy Gospel more difficult.’

In the light of China’s linguistic diversity and the fact that the southern Chinese vernaculars like Hokkien, Yuè, and Wú are mutually unintelligible with Mandarin, the lack of interest in local languages calls for explanations and modifications of previous hypotheses. Most importantly, one cannot claim without qualification that the Jesuits ‘learned the local language’ (Collani, 2001:310). As pointed out in previous studies (Coblin, 1996, 1997), the regional base of the Mandarin variety studied by the Jesuits in the late Míng dynasty was the city of Nánjīng 南京. Outside the Nánjīng area, Mandarin had little communicative utility among the common people. The same is certainly also true of wényán, which, as a purely written language, was out of reach for the uneducated masses.9 The Jesuit focus on Mandarin and wényán and its implicit indifference towards local languages even went beyond the literary practices of Chinese scholars. It must be pointed out that the recitation of classical texts by literati officials was not bound to Mandarin pronunciation, but followed conventions that were closely linked to local vernaculars (cf. Hirata, 2000:541; Kaske, 2008:41, 54). 2. Tradition, Strategy, or Environment? Why did the Jesuits embark on a different path of language learning and documentation from that of the Dominicans? As pointed out, the Jesuits were chiefly engaged in interaction with the Mandarinspeaking literati officials. Thus, the choice of Mandarin is perfectly compatible with the claim that the Jesuits in China pursued a ‘topdown strategy,’ following the idea that the conversion of the masses would follow the conversion of the emperor and the literati officials (cf., inter alia, Standaert, 1999; Collani, 2001:310; Mungello, 2005: 23). The Dominicans, on the other hand, were less devoted to close intellectual interaction with the Chinese elites, even after their expansion of missionary work to China in the 1630s. Menegon points out that the Dominican friars sent to Fújiàn ‘had, with a few exceptions, received a rather narrow and abbreviated education […]. Thus, the Dominicans could offer little to Chinese literati in terms of

–––––––– 9

According to Rawski (1979:23), during the eighteenth and nineteenth centuries, about seventy percent of the population was illiterate.

HOKKIEN DIALECTS IN EUROPEAN SOURCES

39

“learned conversations”’ (2009:57). Moreover, in the well-known controversy concerning the participation of Catholic missionaries in traditional Chinese rites, the Dominicans, in contrast to the Jesuits, pursued a strict anti-accommodationist strategy, explicitly criticizing the tolerant stance of the Jesuits on this issue (Mungello, 2005:81–84). The experience of the Dominican friars prior to their arrival in China also plays an important role. According to Menegon, China-based Dominicans had first been stationed in Mexico or the Philippines, ‘where anti-idolatry campaigns against native religions were the dominant thrust of the missionary enterprise’ (2009:58). It would be beyond the scope of the present study to link these anti-elitist and anti-accommodationist traits in the Dominican strategy with the intellectual tradition of the order. From the above comparison, however, it is evident that the two orders pursued different strategies, and that the Dominican strategy entailed more distance from the values, culture and ultimately the language of the literati officials. The different intellectual traditions and proselytizing strategies of Jesuits and Dominicans, thus, certainly played an important role when it came to the choice of language of education, research, and documentation in missionary contexts. The relevance of different traditions notwithstanding, I argue that the immediate social environment of missionary work in China and the Philippines is of greater importance. From the very beginning of the Jesuit mission, the missionaries’ right to remain on Chinese soil was challenged by China’s political elites. Minor incidents often turned into large-scale anti-Christian movements, which in turn severely jeopardized the Jesuit enterprise in China as a whole. In other words, the Jesuits had no choice but to closely interact with the elites. To be sure, many Jesuit missionaries certainly shared a genuine respect and admiration for Chinese culture. On the other hand, however, a proselytizing strategy not focusing on the language and culture of the elites would have been doomed to failure. In contrast, the Dominicans and missionaries of other orders based in the Philippines had the backing of the Spanish colonial government. As representatives of the colonial order, they were not forced to interact with local elites in order to safeguard their residence status. More importantly, even if the Dominicans had wanted to establish connections with Chinese literati, this would have been rather difficult in Manila and elsewhere, simply because the Sangleys came from the lower echelons of society. The Dominican Juan Cobo (ca.1546–1592)

40

CHAPTER TWO

notes in 1589 that ‘I wish to say that the people who come over are the poor, seagoing people, fishermen and laborers who come to earn a living’ (quoted by Felix, 1966:138). Cobo’s contacts with the Chinese were not restricted to the poor. Wills notes that he also interacted with ‘men of substantial Confucian learning,’ resulting in the publication of ‘a remarkable work of Neo-Confucian-Christian apologetics’ in 1593 (1994:116). This publication, according to Wills, is ‘remarkable evidence for the presence of at least one reasonably learned Chinese scholar in Manila at this time and for the very rapid progress of Dominican dialogue with the Chinese’ (ibid.:117). The example quoted by Wills is certainly an exception and, in the broader picture, Mandarin was of little or no communicative use in the interaction between Spaniards and Sangleys. As the spread of Mandarin outside the capital was largely restricted to literati officials, it is unlikely that the Sangleys could speak it. Although this claim has yet to be substantiated, I assume that the Mandarin proficiency of most Sangleys during the first half of the seventeenth century was comparable to that of overseas Chinese in Malacca two hundred years later. Describing his experience in Malacca during the early nineteenth century, Medhurst writes that ‘no man in five hundred knows any thing of the Mandarin tongue, or can carry on a conversation of more than ten words in it’ (1832:v). Similarly, referring to the linguistic competence of overseas Chinese at the College of Goa, Valignano noted that they knew nothing of the Mandarin language (1542–1564/ 1944:211). It has been reported that the Dominican Francisco Varo (Wàn Jìguó 萬濟國, 1627–1687), author of one of the first printed Mandarin grammars (Varo, 1703/2000), studied Mandarin in Manila for one year before embarking for China in 1649 (Coblin and Levi, 2000:x). However, as stated above, it is hard to conceive that Varo’s first steps in Mandarin learning benefited from the linguistic environment of Manila. We must therefore assume that he received some kind of formal schooling organized by the Dominican order. Thus far, however, little research has been done on the installation of Mandarin education by Dominicans in Manila after the 1630s. As this comparison shows, it is by no means unusual that the linguistic approaches of missionaries of the same order varied considerably in different social environments. Thus, from a broader perspective, it becomes even more evident that missionary linguistic approaches were in the first instance shaped in the specific environment

HOKKIEN DIALECTS IN EUROPEAN SOURCES

41

of missionary work. Significantly, when the Dominicans succeeded in expanding their missionary activities to southern China in the 1630s, they also shifted their focus of language learning from Hokkien to Mandarin.10 This shift was largely motivated by social considerations as the Dominicans, like their Jesuit confrères, were now required to communicate in Mandarin, the prestigious language of the literati officials (Menegon, 2009:56–57; see also Zhāng Xiānqīng, 2009).11 Another example of the influence of social contexts on missionary linguistic approaches is the oldest extant EMH source, the Dictionarium Sino Hispanicum (Chirino, 1604a), which was compiled by the Jesuit missionary Pedro Chirino (see Chapter 3). In contrast to the China-based Jesuits, Chirino had few reservations about devoting himself to the study of the spoken languages of the Philippines, including EMH. His book Relacion de las islas Filipinas (‘Report on the Islands of the Philippines,’ Chirino, 1604b; trans. BR 12:169–320) contains a chapter dealing with native Philippine languages and providing translations of the Ave Maria in Tagal (BR 12:237), Harayan (BR 12:238), and Bissayan (BR 12:239). Chirino was by no means the only Jesuit based in the Philippines who learned and recorded a local vernacular; and it is safe to claim that language documentation by Jesuits in the Philippines represents more diversity and authenticity in their response to the linguistic environment than Jesuit language studies in China (studies on missionary linguistics in the Philippines include Wendt, 1992; Javellana, 1999; Ridruejo, 2004, 2005, 2007). In Japan, moreover, the approach of the Jesuit João Rodrigues (ca. 1561–1633) to language research was significantly different from the path taken by the Jesuits in China. He not only wrote two Japanese grammars of the Arte type (Maruyama, 2004), but also explicitly criticized the approach of the China-based missionaries to learning Chinese through ‘constant dealings with the natives, being always in

––––––––

10 This issue requires further research. Whereas the new focus on Mandarin cannot be denied, some reported sources also hint at a growing interest in Northern Mǐn dialects. Some of these sources are listed in Chapter 3 (Table 1). 11 A few years earlier, in 1626, Dominican missionaries, under the auspices of the Spanish crown, had established a Catholic mission in Taiwan. Apparently, the Dominican friars Jacinto Esquivel (1595–ca. 1633) and Theodore Quirós de la Madre de Dios (dates unknown) composed a dictionary and an Arte of a native Formosan language. These sources are no longer extant (Borao 2001, Klöter 2008). The Catholic enterprise in Taiwan started before large-scale influx of Chinese immigrants had taken place. Thus, from a linguistic perspective, comparisons with Dominican language studies in China would not be meaningful.

42

CHAPTER TWO

their midst’ (quoted and translated by Brockey, 2007:254). These examples all show that conclusions about links between an order’s tradition and the linguistic approaches taken by its missionaries tend to be stated as inaccurate generalizations if the specific social contexts of missionary work are left out of consideration. It remains a task for future comparative research to examine in more detail the interplay of different traditions and social contexts in the formation of distinct missionary linguistic approaches. G. THE NEGLECT OF CHINESE DIALECTS, OR: ‘THEY SPEAK CLOWNISHLY’ The second half of the seventeenth century was, in the words of Cornelius, ‘high-water years for speculations about the Chinese language’ (1965:68). These speculations were to a large extent influenced by accounts from the Jesuits. The EMH manuscripts written by the Dominicans, on the other hand, did not attract any noteworthy attention. Reports on the Chinese language reached European thinkers either through procurators dispatched from the Chinese missions to Europe (cf. Dehergne, 1973:314ff.; Kraft, 1976:94), or, more importantly, through direct correspondence with the missionaries. A wellknown example is the correspondence between the German scholar Gottfried W. Leibniz (1646–1716) and China-based Jesuits (Widmaier, 1990, 2006). A look at some examples reveals that the Jesuits did not modify the direction of analysis taken by Mendoza, de Rada and other writers of the sixteenth century. In essence, the Jesuit reports reinforced the focus on Chinese character writing, which turned into almost obsessive fascination among European scholars (Alleton, 1994; Porter, 2001). Such a fascination for Chinese writing can, for instance, be found in the works of scholars like Francis Bacon (1561–1626), Athanasius Kircher (1602–1680), John Wilkins (1614–1672), and Leibniz. Their fascination must be seen in the context of the European search for a universal language in the seventeenth century. According to Mungello (1985:34): [This] search for a universal language had Biblical roots in the widespread assumption that the Primitive Language—a language of utter simplicity, clarity and uniformity—given by God directly to the first man, Adam, had been lost with the dispersion of tongues which occurred at the Tower of Babel.

HOKKIEN DIALECTS IN EUROPEAN SOURCES

43

In order to overcome this loss, scholars like Bacon expressed the plan to create ‘Real Characters.’ These ‘would be capable of communicating their meaning to all nationalities in a clear, logically self-evident manner rather than in the usual arbitrarily agreed-upon manner of languages’ (Mungello, 1985:35). The concept of Real Characters ‘involved writing that represents not merely letters or words, but things and ideas’ (Mungello, 2005:86; see also Singer, 1989). It was widely believed that Chinese characters could fulfill this role. In sharp contrast to the fascination for Chinese characters, scholarly interest in China’s linguistic diversity was almost non-existent in the seventeenth and eighteenth centuries. In his letters, Leibniz only occasionally requested information on local languages. For example, in a letter to Joachim Bouvet (1656–1730) written on 15 February 1701, Leibniz asks for samples of the Lord’s Prayer and frequently-used words in different Asian languages. He adds that ‘s’il ya dans la Chine même des langues differentes de celle qui est commune dans l’Empire, j’en souhaiterois la meme chose’ (‘if there are in China languages different from the common language of the Empire, I would hope for the same thing’; quoted by Widmaier, 2006:318). Three years later, addressing Jean de Fontaney (1643–1710), Leibniz notes: ‘J’ay lû, qu’il y a des differentes langues particulieres dans diverses provinces de la Chine même’ (‘I have read that in various provinces of China itself there are different languages’). His request for local language data, reiterated in the letter to de Fontaney and again to Bouvet in 1704, remained unanswered. Instead, de Fontaney responded to Leibniz on 10 September 1705: A la Chine on ne parle qu’une mesme langue par tout; on a cependant quelques dialectes en plusieurs provinces, qui n’empeschent pas qu’on ne les entende fort bien, excepté dans la Province de Fokien proche d’Emoüy. Leur jargon en ce quartier là est un chinois si corrompu, qu’on ne l’entend point. Mais tous entendent la langue Mandarine, qui n’est pas une langue diverse et relevée, mais un chinois pur, sans dialecte particulier. Le Tunkinois et le japonois (je parle des langues) sont differents du chinois. Au Tonkin neanmoins et au Japon on connoit les characteres chinois, mais chacun lit en sa langue. (Widmaier, 2006: 508, 510) [In China one and the same language is spoken everywhere. There are, however, some dialects spoken in several provinces. These can all be understood, except in the province of Fújiàn close to Xiàmén. The gibberish there is so corrupt that one does not understand anything of it. But everyone understands Mandarin, which is not a different or exalted

44

CHAPTER TWO

language, but a pure Chinese devoid of dialect features. The Tonkinese and the Japanese (I speak of the languages) are different from Chinese. In Tonkin and Japan they know Chinese characters, but everyone reads in his own language.]

De Fontaney’s remarks contain some inaccuracies which lead to important questions. Obviously, his sketch of mutually intelligible dialects is at odds with Francisco Simões previously quoted claim that each village seemed to have its own language. It is hard to imagine that de Fontaney was not aware of China’s linguistic diversity. Why, then, does he not mention it? As a missionary who had spent many years in China he must have been aware of the fact that, in southeastern China, Mandarin was not spoken by the lower ranks of society. Why, then, does he claim that everyone spoke Mandarin? His reference to the dialect of Fújiàn (Hokkien) obviously has a derogatory undertone. Does this merely reflect de Fontaney’s frustration arising from his inability to understand it? Or does he convey an evaluation of local languages which persists to the present day, i.e., the notion that local languages have an ‘inferior status and deficient nature’ (Mair, 2007) in comparison to Mandarin? I will not attempt to provide any answers to these questions here. Suffice to note that de Fontaney’s sketch fits nicely into the broader picture of a missionary linguistic school that simply had no interest in linguistic variation. De Fontaney’s remark on the ‘gibberish’ of Fújiàn is by no means the first such reference to Hokkien of this kind. Some 35 years before de Fontaney corresponded with Leibniz, the British architect John Webb (1611–1672) had published his Historical Essay, Endeavouring the Probability that the Language of China is the Primitive Language (Webb, 1669). Webb’s Essay is the first lengthy monograph on the Chinese language to be published in the West (Harbsmeier, 1995; Ramsey, 2001). Webb had never visited China and he did not speak any Chinese. The information in his book is thus entirely culled from Western literature on China available during his days. From these sources Webb constructed ‘a case that Chinese was the original language of mankind before the building of the tower of Babel’ (Harbsmeier, 1995: 327). His Essay contains a few lines on the Hokkien language. He writes (p. 183): […] likewise in Fokien where they speak clownishly they usually change N into L, as Lankin for Nankin, and the like. For thus Martinus [Martino Martini, H. K.] also, in his description of Nankin. The

HOKKIEN DIALECTS IN EUROPEAN SOURCES

45

Portugals, saith he, vulgarly call it Lankin, receiving the errour from the Fokiens, with whom they chiefly trade, for these being very rude in speaking by a most common vice of their Countrey are wont to change every N into L.

Interestingly, the very same example occurs in Matteo Ricci’s diaries (Ricci, 1615/1953:268).12 In contrast to Webb, however, Ricci did not attach any negative connotations to this linguistic observation. We are thus left wondering on what basis Webb judged Hokkien pronunciation as erroneous and rude. It is not known when EMH documents first made their way to Europe. The earliest traces of the Arte in European scholarship go back to the first half of the eighteenth century when Theophilus S. Bayer (1694–1738), one of Europe’s pioneer sinologists, copied a manuscript in Berlin and translated some parts of it into Latin. As pointed out in Chapter 1, he included the translation in his famous Museum Sinicum of 1730, where it forms the third and last part of the chapter entitled Grammatica Linguae Sinicae. One hundred years later, the German orientalist Julius Klaproth (1783–1835) included EMH data in his famous work Asia polyglotta (1823; see Lundbæk, 1995:31). Bayer and Klaproth both consulted the LMS of the Arte, the Bocabulario and some religious texts (van der Loon, 1967:106–108). All in all, European scholarship prior to the nineteenth century contains only some occasional references to Hokkien and other Chinese regional languages. This lack of interest is by no means due to a lack of available data, as Hokkien dialects are richly documented in the dictionaries and grammars left behind by the Dominican friars in and around Manila. Why, then, did European scholars show such little interest? One reason is certainly that reflections on the Chinese language were to a great extent shaped by the dialogue between the China-based Jesuits and European thinkers. As argued above, the fascination for Chinese writing and its link with the endeavor to discover a biblical universal language were simply incompatible with the notion of a multilingual China. Another reason is that all EMH documents only circulated as manuscripts. It is not known how many copies of the sources existed. In any case, however, it goes without saying that a few hand-copied dictionaries and grammars could not exercise the

––––––––

12 It is problematic to attribute authorship of the published diaries exclusively to Ricci, as they owe much of their content to Trigault (Schreyer, 1992:13–14; Standaert, 2003:369, fn. 2).

46

CHAPTER TWO

same intellectual influence as printed books. This changed fundamentally in the nineteenth century, with the beginning of the Protestant tradition and its numerous printed Hokkien dictionaries. H. PROTESTANT CONTRIBUTIONS, OR: ‘AN INDEPENDENT LANGUAGE’ It was not until the nineteenth century that detailed data of various Chinese regional languages became available. Like the first period of missionary research on Hokkien dialects, this second phase has an analytical and an ideological dimension. The analytical side of Protestant dialectology in southern China has been discussed in detail in a number of previous studies, e.g., Branner (1997) and Klöter (2005: 89–130, 2006). Like the first period, this second period of Hokkien dialectology was dominated by lexicography, phonology, and orthography design. A Hokkien grammar was not written. The analysis of Hokkien phonology profited to a large extent from the incorporation of traditional Chinese phonology. Another dimension of missionary linguistics in southern China was the birth of comparative dialectology. As Branner points out, it ‘does appear that Western missionaries were the first to do systematic Chinese dialect comparisons. This work began late in history as it did in part, of course, because the Chinese lacked the techniques of phonetic description and comparative method’ (1997:244). Various sources show that, on the ideological side, missionary linguists were divided in their opinion about the role of alphabetic writing vis-à-vis the Chinese script. It is important to note here that in nineteenth-century Europe, enthusiasm for Chinese culture in general and Chinese writing in particular gave way to a widespread rejection of things Chinese (cf. Mungello, 2005:107–130). It is thus by no means uncommon to see various disparaging remarks about the Chinese language in European sources of the nineteenth century. For example, in a newspaper article published in the Edinburgh Review in 1805, Francis Jeffrey writes that there is ‘no instance, we believe, on the face of the earth, of a language so extremely imperfect and inartificial; and it is difficult to conceive how any race of people could be so stupid, or so destitute of invention, as to leave it in such a state of poverty’ (cited by Kennedy, 1951:161). Thus, when Protestant missionaries compiled their Hokkien dictionaries in the nineteenth centu-

HOKKIEN DIALECTS IN EUROPEAN SOURCES

47

ry, they very likely did not receive any written inquiries about the mysteries of the Chinese script. In the following sections, the diverging opinions with regard to the role of alphabetic writing will be introduced by quoting some representative viewpoints. The first printed Hokkien dictionary was published in 1832 by Medhurst. Medhurst’s first station in Asia was Malacca, where he worked among the Hokkien-speaking Chinese community (Klöter, 2005:94, 108–111). Medhurst is known as the creator of the earliest version of the most-widely used Hokkien romanization system, known as ‘Church Romanization’ because of its widespread use in mission schools. He never considered alphabetic writing as a substitute for the Chinese character script. To be sure, when Medhurst wrote his Hokkien dictionary, there was, according to DeFrancis, ‘almost unanimous agreement among experts and amateurs alike that the Chinese language should be characterized as exceedingly difficult, uniquely monosyllabic, and, most important of all, necessarily ideographic in its written form’ (1950:18). Due to the enforced opening of China following the First Opium War and the Treaty of Nanking in 1842, Western missionaries for the first time in history gained widespread access to Fújiàn province, the homeland of Hokkien dialects. In the following sixty years, a few high-quality Hokkien dictionaries were printed (e.g., Douglas, 1873; Van Nest Talmage, 1894). The city of Xiàmén became an important base for missionary work in general, and for Hokkien language studies in particular. By the beginning of the twentieth century, the alphabetic writing of Hokkien had become associated with European practices of literacy transposed on native communities (cf. Errington, 2008:16). This development is linked to the language planning of a number of missionaries, most importantly, John Van Nest Talmage (1819–1892) of the Reformed Church in America and Carstairs Douglas (1830– 1877) of the Presbyterian Church of England. Van Nest Talmage belonged to the first group of Xiàmén-based missionaries who taught Church Romanization to illiterate Chinese. For this purpose, he composed spelling books and printed them. At the beginning of the twentieth century, Church Romanization had developed into a quasiofficial orthography in missionary primary schools. It has been estimated that between five and six thousand locals were able to read Church Romanization (Pitcher, 1912:209). According to missionary statistics, almost 30,000 Amoy translations of the Scriptures in Church

48

CHAPTER TWO

Romanization were sold between 1890 and 1920 (cf. Stauffer et al., 1922:9). This widespread use of Church Romanization in the Protestant missions of southern China, and, subsequently, the island of Taiwan, had a firm ideological basis. The most outspoken supporter of alphabetic writing was Carstairs Douglas, who, in the preface of his dictionary, dismissed the use of Chinese characters as the ‘most serious defect,’ one reason being that the absence of characters ‘may serve to make manifest the fact that the Vernacular of Amoy is an independent language, which is able to stand alone without the help of the written character’ (1873:viii–ix). Douglas’ position did not remain unchallenged. Taiwan-based William Campbell (1841–1921) of the Presbyterian Church writes in his Hokkien dictionary that ‘no sympathy is shown here for the action of those missionary brethren who push forward Roman letters with the avowed intention of thrusting Chinese methods of writing and printing into the scrap-heap’ (Campbell, 1913: iii). The dispute on the selection of a writing system for teaching a Sinitic vernacular to locals was not restricted to Xiàmén. A very similar situation has been reported for the Basel mission in Hong Kong, where Hakka was taught (for details, see Chappell and Lamarre, 2005:29–34). I. SUMMARY AND CONCLUDING REMARKS Systematic documentation of Hokkien dialects by missionaries started as early as research on Mandarin and the Chinese script. Different objects of analysis—local vernacular vs. linguistic codes of the cultural and political elite—can be associated with two different orders and localities, i.e., the Dominicans in the Philippines and the China-based Jesuit mission. The reasons for the choice of different languages by these two orders can chiefly be found in the specific social contexts of missionary work. The Jesuits in China had no choice but to accommodate to the political and cultural elite, favoring Mandarin, whereas Dominican missionaries in the Philippines, as part of the Spanish colonial order, had interactions with the lower echelons of Chinese society, thus focusing on local varieties. The focus of Jesuit language documentation was a direct continuation of a tradition which had been initiated during the first encounter of Europeans with China. I therefore agree with Schreyer who argues that all claims and hypotheses con-

HOKKIEN DIALECTS IN EUROPEAN SOURCES

49

cerning the Chinese language proposed by the Jesuits can essentially be found in earlier European sources as, for example, Mendoza’s book (cf. Schreyer, 1992:17; for a comprehensive study of the European perception of China prior to the Jesuit mission, I refer to Reichert, 1992). Missionary linguistics in the Philippines, on the other hand, had many facets, most importantly grammar writing and, as will be shown in Chapter 3, dictionary compilation. Chapter 5 analyzes the EMH system of romanization devised by the Dominicans. In the context of this chapter it must be emphasized that the existence of a romanization system by itself does not say anything about practices of literacy. Future historical research will have to examine whether the system was only used for the teaching of newly arriving missionaries, as I assume, or whether its use was expanded to the teaching of illiterate Sangleys. In this respect one must express the hope that the scope of promising research projects aiming at the digitalization of historical sources relating to the Spanish-Chinese history will be expanded in the future (Busquets i Alemany, 2006). On the basis of the available sources, we conclude that it was only with the arrival of Protestant missionaries in southern China during the nineteenth century that Hokkien dialects became object of literacy practices imposed on the Chinese. Finally, it must be asked why the historiography of Chinese linguistics has largely neglected the Dominican documentation of Hokkien dialects. One reason is certainly that within Chinese linguistics, scholarly appreciation of the discipline’s past depends on the extent to which historical studies continue to influence present-day research. As early studies on local languages allegedly contribute little to linguistic research on Mandarin and classical Chinese, their scientific raison d’être has hardly been considered. Occasional references to local language studies typically carry evaluative overtones, judging the quality of early missionary contributions by their degree of convergence with modern scholarship. For example, assessing the value of Varo’s Grammar of the Mandarin Language of 1703, Chen writes that it ‘is the first Chinese grammar worthy of its name’ (2004:1). The reason for his appraisal is the fact that Varo deals with tone sandhi, phonetic motivations for pitch changes, and other related topics, such as underlying vs. surface tones. To be sure, Varo’s analysis is by far more detailed and elaborate than any earlier description of a Chinese dialect I

50

CHAPTER TWO

have seen thus far. I argue, however, that an evaluation of earlier works from the perspective of modern linguistics is misleading. Such an approach narrows the historiographic perspective on missionary linguistics and, as will be shown in the following chapters, impedes access to valuable diachronic data.

CHAPTER THREE

LEXICOGRAPHY A. INTRODUCTION This chapter analyzes the documentation of EMH in dictionaries. The analysis is restricted to titles which can safely be considered as extant. The only exception is the Diccionario de la lengua Chincheo introduced in Section D. The dictionary is now apparently lost, but has been described in some detail in a previous study (Abel-Rémusat, 1825). This chapter is organized according to the chronological order in which the dictionaries were compiled. As the dates of compilation and the names of compilers are in most cases unknown, the dates are estimations based on evidence to be discussed presently. Furthermore, the lexicographic arrangement of each dictionary is analyzed by asking two related questions: Which influences are visible in the lexicographic arrangement of each dictionary? What is the relationship between lexicographic arrangement and language analysis in the dictionaries? These questions are based on two assumptions. First, missionaries did not compile dictionaries from scratch. Instead, they were exposed to some kind of influence, such as a model dictionary prescribing the selection of entries. The analysis of the sources of influence will help us to better understand directions of influence in missionary linguistics of the seventeenth century. Second, the arrangement of a dictionary, including the information found in the entries, has an important analytical dimension. As I have argued before (Klöter, 2007), it was widely held in European scholarship of the seventeenth and eighteenth centuries (and, to a lesser degree, ever since then) that Chinese is a monosyllabic language. This perception has convincingly been refuted in Kennedy’s seminal article The Monosyllabic Myth (1951). The myth owes its existence, at least to some extent, to the perception of language through the lens of traditional dictionaries that essentially list single characters in combination with monosyllabic character readings. As shown in this chapter, EMH dictionaries of the seventeenth

52

CHAPTER THREE

century reflect a linguistic understanding which has little in common with the monosyllabic myth. B. EMH SOURCES: REPORTED TITLES AND EXTANT SOURCES It is not known for certain when contacts between Sangleys and European missionaries first resulted in the compilation of EMH grammars and dictionaries. If historical reports are correct, the first dictionary was written before 1580 by Martín de Rada (see Chapter 2) as part of the Arte y vocabulario de la lengua china (Jiménez, 1998:182; Masini, 2005:183). Zhāng’s (2004) bibliography contains a reference to a Hokkien-Spanish dictionary of 1575 (p. 53). Unfortunately, its original title and present location are not indicated. If these records are correct, this would imply that Chinese missionary linguistics actually started outside China, as de Rada’s Arte y vocabulario would be older than Matteo Ricci and Michele Ruggieri’s Portuguese-Chinese dictionary which was presumably compiled in the 1580s (Weingartner, 1975:224–245; Yang, 1989:210; Witek, 2001:156–158). Van der Loon doubts that de Rada wrote a linguistic work, adding that it is ‘safe to conclude that some anonymous manuscripts were circulating among the missionaries and that they were copied and enlarged as occasion arose’ (1967:96). As far as is known, there is no extant copy of the de Rada dictionary—nor of many other such early works. According to different bibliographies, at least eighteen linguistic sources of Hokkien dialects were compiled between the sixteenth and eighteenth centuries. To these we can add four documents describing Northern Mǐn dialects. All linguistic sources are listed in Table 1. I have thus far only been able to confirm the existence of eight such works, viz. two manuscripts of the Arte, two manuscripts of the Dictionarium Sino Hispanicum, and the Bocabulario de lengua sangleya. The archives of the University of Santo Tomás in Manila hold two EMH dictionaries mentioned by González (1966), although the names are not identical (see Section F). In addition, I have seen an untitled and incomplete wordlist referred to as the lista de palabras in the Dominican archives of Avila (Spain). In two other cases, it seems safe to assume that the sources indeed once existed. Van der Loon (1967:97) traces the journey of the one known copy of the Diccionario de la lengua Chin cheo until its disappearance in 1894, when the library of its last owner, Marquis d’Hervey-Saint-Denys (1822–1892), was sold.

53

LEXICOGRAPHY

Francisco Márquez’ Gramática was kept in the Dominican Provincial Archives in Manila, but it disappeared after the Convent of St. Domingo was bombed by the Japanese in 1941 (González, 1955:153). Riccio’s Gramática chinchea is occasionally referred to in other early missionary sources, but I have thus far not found any conclusive evidence that this grammar actually existed. Whereas the existence of other manuscripts has been claimed on the basis of weak evidence, we may assume that yet again other sources have remained undiscovered and therefore unmentioned in the past centuries. In short, although the dimensions of dialect research by missionaries cannot be ascertained quantitatively in terms of numbers of sources, there is ample evidence in support of the claim advanced in Chapter 2 that early missionary linguistics on Chinese was more diverse and less elite-centered than is generally acknowledged. Table 1: Missionary sources of Mǐn dialects (N. Mǐn = Northern Mǐn)

Compiler

Title

Century

Language

Mentioned / kept in

1.

?

Diccionario chinoespañol

?

Hokkien

González, 1966

2.

?

Diccionario españolchino

?

Hokkien

González, 1966

3.

?

[Una lista de palabras en español y en caracteres chinos]1

?

Hokkien

Archivo del Real Monasterio de Santo Tomás, Ávila, Tomo 38º / Número 14

–––––––– 1

The title only appears in the archive catalogue and not in the document itself.

54

CHAPTER THREE

Table 1 (cont.) Compiler

Title

Century

Language

Mentioned / kept in

4.

?

16th

Hokkien

Zhāng, 2004

5.

Miguel de Benavides

[中文西譯閩 南語西班牙 文對譯字 典]2 Diccionario españolchino vulgar

16th

?

González, 1966

6.

?

Fukien Dialect Dictionary (Spanish)

17th

Hokkien

Zhāng, 2004

7.

Francisco Márquez

Gramática españolachina

17th

Hokkien

González, 1955

8.

Victorio Riccio

Gramática chinchea

17th

Hokkien

González, 1955

9.

Francisco G. de Sampedro

Gramática españolachina del dialecto de Fogán

17th

N. Mǐn

González, 1955

10.

Francisco G. de Sampedro

Vocabularia en las misma lengua

17th

N. Mǐn

González, 1955

–––––––– 2

This is the Chinese translation of the original title, Engl. ‘A Chinese-Southern Mǐn Dictionary.’ The original title is not indicated in Zhāng (2004).

55

LEXICOGRAPHY

Table 1 (cont.) Compiler

Title

Century

Language

Mentioned / kept in

11.

Magino Ventallol

Cabecillas, o léxico del dialecto de Emuy, o del mandarín

17th

Hokkien

González, 1966

12.

?

Gramática China y Española Fokien

17th

Hokkien

Niederehe, 1999

13.

?

Diccionario de la lengua Chin cheo

17th

Hokkien

V. d. Loon, 1967

14.

? signed by M. de Mançano (BMS)

Arte de la lengua chio chiu

17th

Hokkien

- Universitat de Barcelona, Ms. 1027 - British Library, Ms. 25317

15.

P. Petrus Chirino

Dictionarium Sino Hispanicum

17th

Hokkien

- Biblioteca Angelica di Roma, Ms. 60 - National Library of France, Chinois 9276

56

CHAPTER THREE

Table 1 (cont.) Compiler

Title

Century

Language

Mentioned / kept in

16.

?

Bocabulario de lengua sangleya por las letraz de el A.B.C.

17th?

Hokkien

British Library, Ms. 25317

17.

Juan Garcés

Arte foganerolatino

18th

N. Mǐn

González, 1955

18.

Juan Garcés

Dicconario foganerolatino

18th

N. Mǐn

González, 1955

C. THE DICTIONARIUM SINO HISPANICUM 1. Historical Context The oldest extant EMH dictionary was written at the beginning of the seventeenth century. As it was compiled in the Philippines, it may be included in the list of EMH documents. Two manuscripts are extant (Chirino, 1604a). One is kept in the Biblioteca Angelica in Rome (Ms. 60, hereafter: RMS), the other in the National Library of France (Chinois 9276, hereafter: PMS).3 The unnumbered title folio was signed by the Jesuit Pedro [Petrus] Chirino (1557–1635) on 31 March 1604. It also identifies Chirino as the compiler. It reads as follows (original emphasis): Dictionarium, Sino Hispa/nicum quo P. Petrus Chirino / societatis Jesu linguam / sinensium in Filipinis / addiscebat ad convertendos / eos Sinenses qui Filipinas / ipsas incolunt et quadra/ginta millium numerum excedunt. [Chinese-Spanish dictionary, with which Father Petrus Chirino of the Society of Jesus has learned the Chinese language of the Philippines for

–––––––– 3

If no reference to a specific manuscript is made in this study, I refer to the RMS.

LEXICOGRAPHY

57

the conversion of those Chinese who inhabit the Philippines proper and exceed the number of 40,000.]

Chirino was a native of Osuna in Andalucia. He entered the Jesuit Order in 1559, at the age of 22, and first arrived in the Philippines in 1590 (BR 12:175). In 1600, he became the first director of the College of Manila. The college was, in the words of Chirino’s contemporary Francisco Vaez (1543–1619), ‘the leading one’ in the Philippines (Vaez, 1601; transl. BR 11:195). In the early seventeenth century, it was in the first instance a Jesuit residence which also provided instruction to Spaniards and natives (ibid.). Chirino is also known as the author of the Relacion (see Chapter 2), one of the first Western narratives of the history, flora and fauna, customs, and Christianization of the Philippines. The Relacion was published in Rome in 1604, during Chirino’s four-year sojourn in Europe. He returned to Manila in 1606 (BR 12: 176). The following quotation from the Relacion gives a good impression of Chirino’s interest in languages and of his commitment to language learning (Chirino, 1604b; transl. BR 12:234f.). I will say that the facility with which many ministers of the Lord in the four religious orders learned the languages used in their respective missions, even so as to preach and hear confessions in them, seems a gift from heaven. The most tardy student of them, if he applies himself moderately, spends no more than six months; and one of ours, Father Cosme de Flores, learned and mastered this language, so that he could preach and hear confessions, in seventy-four days to the astonishment of our people, as well as of the Indians themselves. The latter, seeing this facility, say that God, without doubt, bestows it upon us, recognizing their needs. In truth these languages are not very difficult, either to learn or to pronounce—and more especially now, since there is a grammar, a vocabulary, and many writings therein. The most difficult is the language of Manila (which they call Tagal)—which, I have already said, Father Martin Henriquez learned in three months; and in three more, he used it fluently. This was the first of the native languages that I learned, to which and to the others I shall profitably devote another chapter.

In the chapter ‘Of the Languages of the Filipinas,’ not a single reference to Hokkien is made. Instead, the chapter exclusively deals with native Philippine languages. On the one hand, this seems hardly surprising, as the rapid growth of the Chinese population in the Philippines had just begun a few decades earlier. Despite the existence of Chinese commercial and residential centers like the Parián (see Chapters 1, 2, 6), the Sangleys were considered a foreign population group, and their language was associated with their Chinese homeland. On

58

CHAPTER THREE

the other hand, however, it seems like an irony in missionary linguistic history that Chirino, as the compiler of one of the first SpanishHokkien dictionaries, does not say anything about this contribution in his Relacion. When writing of Chinese language learning, he merely mentions Father Francisco Almerique (dates unknown), who ‘began the study of the Chinese language, in his zeal to aid in the conversion of the many Chinese who came to Manila and whom we in the Filipinas call Sangleys’ (Chirino, 1604b; BR 12:199f.). In the preface of the English translation of the Relacion, however, it is mentioned that, after 1595, Chirino resided in Cebu. There he ‘undertakes to instruct the Chinese, whose language he soon learns sufficiently for that purpose’ (ibid.:22). Chirino’s command of Chinese is also mentioned by de la Costa who writes that ‘[i]n order to be able to work among them [the 200 Chinese residents of Cebu, H.K.] Chirino applied himself to the study of the language […] under a young Christian Chinese […] sent from Manila to be his tutor’ (1961:166f.). As mentioned above, Chirino signed the first folio of the Dictionarium on 31 March 1604. It was presented to the Augustinian cardinal Angelo Rocca (1545–1620), sacrist of the Pope, well-known bibliophile, and founder of the Biblioteca Angelica, where the manuscript is still kept today (Masini, 2000:61). The date of the signature and the explicit mention of the fact that 40,000 Chinese resided in the Philippines has some historical significance. As pointed out in Chapter 2, the number of Chinese residents in Manila had risen quickly towards the end of the sixteenth century. However, in October 1603, twenty thousand Chinese settlers were killed by the colonial administration. As Chirino had left Manila a year before, van der Loon conjectures that he was unaware of the recent events (1967:98). The first SpanishEMH dictionary was thus signed at a time when political tensions between the Spaniards and the Chinese in Manila had reached new heights. In 1605, two years after the massacre in Manila, the Spanish archbishop of the Philippines, the Dominican Miguel Benavides (1550–1605), described the Sangleys as ‘infidels and idolaters.’ He writes that they are ‘a most pernicious and injurious people to be settled among the Christian natives, newly converted to our holy Catholic faith; for the said infidel Sangleys are most vicious, both with women and in an unnatural manner, and are extremely liberal in spending money for their purposes and desires, and artful and crafty for every form of evil’ (Benavides, 1605; transl. BR 13:271).

LEXICOGRAPHY

59

2. Lexicographic Make-up Chirino’s Dictionarium consists of 84 folios. Each folio contains an average of 13 entries. In accordance with its title (‘Sino-Hispanicum’), the source language of the dictionary is Chinese, or, more specifically, EMH. Characters are the central unit of each entry. This is reflected in the arrangement of an entry: characters are written larger than Roman letters and thus occupy more space, romanized transcriptions and Spanish translations appear as appended units. Moreover, the reading direction of each folio follows the reading of characters, i.e., from top to bottom, from right to left. The inclusion of characters is what distinguishes the RMS from the PMS; otherwise they are largely identical. As the arrangement of romanized transcriptions and Spanish translations in the PMS leaves blank space for characters, I assume that the manuscript was copied from the RMS. It seems likely that Chirino wrote a second copy during his stay in Europe, but he must have been unable to find a Chinese person to add the Chinese characters. The sequence of folios does not follow obvious lexicographic principles. The Dictionarium starts off with five folios which are loosely linked to Chinese lexicographical tradition. The remaining 79 folios contain divisions according to semantic fields, topics, and syntactic patterns. None of these divisions are explicitly marked in the Dictionarium. The first five folios each contain four columns of five characters. The characters on each folio share one graphic component roughly indicating the meaning of the character. Such a graphic element is known as signific or radical (cf. Norman, 1988:68). For example, among the entries on folio 1, we find «銀 | gin | Plata» ‘silver’, «銅 Tang | cobre» ‘copper’, «錫 Sià | Estaño» ‘tin’, and «釘 | Teng | clavo» ‘nail’ (see Figure 1 on p. 60). All these characters share the signific 釒, which is a condensed form of 金 ‘metal, gold’. The characters on folio 2 share the signific 木 ‘wood’, followed by the shared signific 氵 ‘water’ on folio 3, 火 ‘fire’ on folio 4, and 土 ‘earth’ on folio 5. These five significs represent the five elements metal, wood, water, fire, and earth in Chinese philosophy. The folios after f. 5 can be grouped into thematic sections. For example, folios 6–9 list names of animals; folios 18 and 19 contain words for body parts; folios 25 and 26 introduce numbers, measures and weights; and on folios 75–77, we find words for crimes and (for details, I refer to Masini, 2000:65–67). In some sections, the thematic

60

CHAPTER THREE

Figure 1: Dictionarium Sino Hispanicum, f. 1 (Rome, Biblioteca Angelica, Ms. 60, reproduction with permission from the Italian Ministry for Arts and Culture)

61

LEXICOGRAPHY

focus is self-evident. Other sections rather seem like a random collection of words and phrases. Compare, for example, the sequence of entries on folio 10 and on folio 40, in Tables 2 and 3. The thematic coherence of expressions in Table 2 is obvious, as they all relate to natural and geographic phenomena. Some links between the expressions in Table 3 can also be established. The first three entries, arranged in one column in the Dictionarium, are verbal expressions with the complement ‘you’. The following six entries (4–9) can be subdivided into three pairs of antonyms. Despite these links, there is no superordinate concept to which all entries on this folio could be related. Similar lack of thematic coherence can be observed for many other folios. This observation is not intended as criticism of lexicographical shortcomings. It would likewise be wrong to emphasize the provisional character of the Dictionarium. After all, Chirino himself considered it appropriate for the manuscript to be designated a Dictionarium and worthy of being presented to the papal sacrist. The observed idiosyncrasies in the arrangement should rather be taken as evidence that a missionary source labeled Dictionarium does not necessarily fall into modern lexicographic categories. Table 2: Dictionarium, f. 10

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Entry Character 天 日 月 星 雲 雨 霜 雪 露 霧 電 雷 霓 風 地

English Transcription Th˘I Tzit Goe Ch˘ee Hun Hou S˘ng Se Lou Bu Siná Lui Gei Hong Tei

Translation Cielo Sol Dia Luna Estrella Nube llover Nieve Piedra Rocio Niebla Relampago Trueno Granizo Viento Tierra

heaven sun, day moon star cloud rain snow hail dew fog lightning thunder hail wind earth

62

CHAPTER THREE

Table 2 (cont.)

16. 17. 18. 19. 20.

Entry Character 山 海 嶼 石 泥

Transcription Su˘a Hai Su Chio Tou

Translation Monte Mar Isla Piedra Peña Tierra

English mountain sea island stone, rock earth

Table 3: Dictionarium, f. 40

1. 2. 3. 4. 5.

Entry Character 騙尓 度尓 乞尓 虧我 做惡

English Transcription P.hien lu Thou lu Qhi lu Qhui goa Cho oc

Translation Engañô te Dar te Para ti Sin culpa Hazer mal

6. 7. 8. 9. 10. 11.

敢 不敢 定 不定 欺負 作主

ca m̃ ca Tia m̃ tiá Qhi hu Cho chu

12.

只個

Chi gue

atrever se no atreverse seguro No seguro Agravio Agente Hazedor Aqui

I cheat you give you for you without fault do harm, hurt, damage dare dare not sure not sure offense, insult agent, maker here

Its idiosyncrasies notwithstanding, the Dictionarium was certainly not compiled in a lexicographic vacuum. Instead, influence of lexicographic traditions from China and Europe are visible. As pointed out above, entries on the first five folios are selected on the basis of shared significs of the characters. A lexicographic arrangement according to shared significs is quite common in Chinese lexicographic history and goes back to Xǔ Shèn’s 許慎 (ca. 58 CE – ca. 147 CE) well-known Shuōwén jiězì 說文解字 (‘Explanations of Simple Graphs and Analyses of Composite Graphs,’ briefly explained in Norman, 1988:67–69) of 100 CE. The use of only five significs on the first five folios of a dictionary is rather unusual for Chinese standard dictionaries. How-

LEXICOGRAPHY

63

ever, the arrangement of the Dictionarium is reminiscent of the Chinese zázì 雜字 ‘miscellaneous characters’ tradition. A zázì dictionary was traditionally used for teaching characters to children. As Wilkinson points out (2000:47–48): Long before school, they [children, H.K.] usually began with pictographic characters such as ri 日 (sun) or yue 月 (moon) and those with the fewest strokes such as 一 yi (one) or er 二 (two). Later they also used lists of characters for common words in rhyming doggerel arranged by categories such as heaven and earth, trees and flowers, fruits and nuts, animals and birds, pots and pans, farm tools, toponyms, and official titles, and much else besides.

As there are countless different teaching aids of the zázì type, it will be difficult to ascertain whether the Dictionarium was compiled on the basis of one particular work or whether it was inspired by the zázì tradition in general terms. It must be kept in mind that the Dictionarium was compiled in cooperation with a Chinese tutor, as the shape of the characters leaves no doubt that they were written by a native Chinese hand, and not Chirino himself. A very likely scenario is that Chirino’s tutor applied teaching methods that he knew from traditional Chinese elementary education, in which the use of zázì was indispensable. He may have copied sections from different zázì and, in response to the needs of his adult European student, added some parts himself. Thus, the first folios can be seen as an attempt to introduce the principles of character writing. Interestingly, a similar lexicographic tradition exists in Europe as well. Originating in the sixteenth century, European polyglot dictionaries were likewise in the first instance compiled for practical purposes, and generally divided into thematic subsections (Niederehe, 1986:170f.). At the micro level, the Dictionarium has different levels of linguistic analysis. The use of a transcription system indicating the pronunciation of EMH headwords shows that Chirino devoted due attention to linguistic form. The inclusion of polysyllabic expressions and sentences as headwords indicates that Chirino reflected on word formation and sentence structure. In many respects, Chirino’s transcription system is rudimentary. Most importantly, diacritics are used rather scarcely and randomly. As I will argue in Chapter 5, there is no evidence for the systematic distinction of tones in Chirino’s transcription system.

64

CHAPTER THREE

The lexicographic treatment of phrases in the Dictionarium (ff. 42– 53 and ff. 78–83) shows that Chirino’s analytical approach went beyond the listing of character readings. There are several instances in which he tried to render EMH word formation and syntax as closely as possible in the word order of the Spanish translation. To be sure, due to obvious syntactic differences between the source and target languages, this approach had limitations. As can be seen from examining the arrangement of the Spanish translations in Figure 2, each word of the Spanish translation is attached directly to its equivalent in the EMH sentence.

Figure 2: Example sentences of the Dictionarium Sino Hispanicum (left: f. 52, right: f. 53) (Rome, Biblioteca Angelica, ms. 60, reproduction with permission from the Italian Ministry for Arts and Culture)

65

LEXICOGRAPHY

A conversion of Chirino’s vertical arrangement into a horizontal arrangement yields the following: (1)

人 lang hombre man

共 cang con with

我 goa mi me

說 sue habla talk

y and

我 goa mi me

不 m̃ no not

知 chai sabio wise

The sequence of words in line 3, hombre con mi habla y mi no sabio, is not an acceptable sentence in modern Spanish, and was not so in 1604 either. In any case, the added y ‘and’ shows that Chirino’s direct renderings were also supposed to be syntactically intact translations. From the perspective of modern linguistic methodology, Chirino’s Spanish renderings can be considered functional hybrids: They are, so to speak, gloss-translations. The limitations of this method are obvious. As the above example shows, such a rendering conceals the meaning of the EMH sentence, which could be read as ‘I don’t know what the person said to me.’ However, in Chirino’s gloss-translation, the first part of the Spanish sentence hombre con mi habla ‘the person talks to me’ cannot be read as an object of mi no sabio ‘I am not wise.’ In a modern linguistic description, the same EMH sentence would possibly be glossed as follows: (2)

人 lang man

共 cang with

我 goa 1.sg

說 sue talk

我 goa 1.sg

不 m̃ not

知 chai know

‘I don’t know what the person said to me.’

In the translations of other EMH sentences, the glossing feature is not as obvious as in Figure 2 (left column). Overall, Chirino seems to prefer acceptable Spanish to direct glossing, when the sequence of glossed EMH morphemes is in conflict with Spanish syntax. In the case of the sentence in the right column of Figure 2, it is unclear why Chirino chose a translation which hides the original wording. Example 3 is a presentation of Chirino’s entry with a literal English translation of the Spanish translation. Example 4 is my transcription, gloss, and translation of the same EMH sentence. (3) 我 goa yo te lo

親 過 chin cue entregue y

尓 lu tu

手 chiu lo ni

[‘I surrender to you, and you deny it.’]

尓 lu egas

夭 yao

諍 chi

66

CHAPTER THREE

(4)

我 gua2 1.sg

親 cim1 kiss

過 kue3 EXP

尓 lu2 2.sg

手 chiu2 hand

尓 lu2 2.sg

夭 iau2 still

諍 cin3 quarrel

‘I have kissed your hand but you still quarrel.’

The separation of glosses and translations is now a generally accepted principle of linguistic representation and does not need further elaboration here. The hybrid nature of Chirino’s translations can be regarded as an initial attempt at this separation. Chirino had some, albeit rudimentary, understanding of the fact that sentences need to be analyzed in terms of the sequence of morphemes and the meaning of the sentence as a whole.4 This, in turn, shows that the lexicographic approach of EMH dictionaries differed fundamentally from the character dictionaries compiled on the Chinese mainland, which indicate monosyllabic readings of isolated characters. This is not contradicted by the fact that characters are the main unit of an entry in the Dictionarium. Given the didactic purpose of missionary dictionary writing, it seems likely that the use of characters followed communicative considerations: Only characters enabled Chirino’s Chinese tutor to read out the entries for his student. D. A LOST DICTIONARY: THE DICCIONARIO DE LA LENGUA CHINCHEO In the decades following the completion of Chirino’s Dictionarium, EMH lexicography became more sophisticated. Later works are generally more comprehensive and more refined in terms of lexicographic arrangement and linguistic analysis. As Chirino’s Dictionarium is the only extant EMH dictionary of its period providing explicit information on publication date and authorship, the chronological contexts of early EMH lexicography remain vague. Unfortunately, what was presumably a milestone in the lexicographic history of EMH has been lost. Its title, however, is revealing. According to van der Loon (1967:97), the complete title is Diccionario de la lengua Chincheo que contiene los vocablos tambien simples que compuestos, con los caracteres generales y peculiares a questo

––––––––

4 This does not imply that the separation of glosses and translations did not exist when Chirino wrote the Dictionarium; an early example of separating glosses and translations in the analysis of Tamil data has been pointed out by James (2010).

LEXICOGRAPHY

67

dialecto, segun lorden del alfabeto español y las cinco tonadas chineses ‘Dictionary of the Zhāngzhōu Language Containing Both Simple and Compound Words, with General Characters and Those Peculiar to this Dialect, in Spanish Alphabetical Order and with the Five Chinese Tones.’ Van der Loon points out that the dictionary once belonged to the well-known French sinologist Jean-Pierre Abel-Rémusat (1788–1832) who wrote a detailed description of it (Abel-Rémusat, 1825). According to this description, it contained 436 leaves of Chinese paper, each leaf being divided in two columns of 24 lines (van der Loon, 1967:97). Each detail of the title shows that a few years after Chirino’s Dictionarium, linguistic analysis and lexicographic coverage of EMH had undergone considerable refinement. First, vocablos tambien simples que compuestos ‘both simple and compound words’ indicates that different types of word formation were explicitly distinguished. Second, con los caracteres generales y peculiares a questo dialecto ‘general characters and those peculiar to this dialect’ suggests that the compiler was aware of non-standard popular writing conventions for Hokkien dialects (for details, see Klöter, 2005:41–87). Third, segun lorden del alfabeto español ‘in Spanish alphabetical order’ hints at an alphabetical arrangement of entries. According to Abel-Rémusat’s description, this was a bilingual Hokkien-Spanish dictionary. Headwords are represented by Chinese characters, followed by their principal meaning in Spanish. If the arrangement is indeed in alphabetical order, this must be the alphabetical order of the Spanish transcription system indicating the reading of single characters. The analytic scope of this dictionary must have gone beyond the indication of character readings. As Abel-Rémusat emphasizes, ‘[c]e Dictionnaire chinois et espagnol ne paraît pas avoir été fait dans l’intention de donner l’interprétation isolée de chaque caractère en particulier’ (‘this Chinese-Spanish dictionary has apparently not been compiled with the intention to provide an isolated interpretation of each particular character’) (1825:87). Its consequent extension of a single character-oriented analysis is, according to Abel-Rémusat, the most important merit of the dictionary. He writes that (1825:88): Mais ce qui fait le principal mérite de l’ouvrage, et ce qui le distingue de tous ceux du même genre dont j’ai connaissance, c’est l’attention qu’a donnée son auteur aux phrases et aux expressions complexes dans lesquelles peut entrer chaque caractère.

68

CHAPTER THREE

[What is the prime merit of this work, and what distinguishes it from all of the same genre I have consulted, is the attention its author devotes to phrases and complex expressions in which each character can occur.]

Fourth, the reference to cinco tonadas chineses ‘five Chinese tones’ proves that tones were recognized. Note that the number of tones is the same as in the Arte if we follow the categorization of the Arte and do not consider entering tones (see Chapter 5). E. THE BOCABULARIO DE LA LENGUA SANGLEYA The British Library holds a EMH dictionary entitled Bocabulario de la lengua sangleya por las letraz de el A.B.C. ‘Vocabulary of the Language of the Sangleys According to the Letters of the Alphabet’ (Add. 25.317, hereafter: Bocabulario). The Bocabulario is the oldest extant EMH dictionary in which entries are strictly separated into EMH headwords and Spanish translations and information categories. The year of compilation and the author of the Bocabulario are not known. In his meticulous textual and historical analysis, van der Loon narrows the possible period of compilation down to the period 1609–1648 (1967:104). One convincing argument refers to the entry «Bee | 馬» ‘horse’ which contains a lengthy subsection on the twelve animals of the Chinese zodiac (f. 21v). In this subsection, the years 1617 and 1618 are explicitly identified as the Year of the Snake and the Year of the Horse respectively. Other Gregorian years are not related to animals of the zodiac. It seems likely that, in order to illustrate the foreign system, the anonymous compiler chose the years in which he actually wrote the manuscript. Comprising 224 double-sided folios, the Bocabulario has an alphabetic macro structure. Its some 1,500 entries are listed under ca. 300 section headings representing EMH monosyllables. For example, the section heading on folio 18r is followed by the four entries «Ban | poco a poco o despaçio» ‘little by little, slow’, «Ban | diezmil» ‘ten thousand’, «Ban | arrancar» ‘pull out, draw out’, and «Ban | sielo de lo [sic] de la cama» ‘ceiling of bed’. Not all entries are monosyllabic. In some instances, the section heading represents the first syllable of a polysyllabic expression. Subsequent to the section heading on folio 46v, for instance, we find the entries «cam | naranjas» ‘oranges’, followed by «camsia | dar gracias» ‘give thanks’.

LEXICOGRAPHY

69

Characters are hardly used in the Bocabulario. In the few instances where they occur they have no fixed spot in the graphic arrangement of an entry. It cannot be ruled out that they were added by one of the later owners of the manuscript.5 It is in any case obvious that the Bocabulario was not concerned with Chinese writing and lexicography. Instead of listing character readings, it presents words which have a pronunciation and a meaning. This is not to say that the influence of character-based approach to language analysis is absent in the Bocabulario. In view of the fact that ninety percent of the entries are monosyllabic, we must assume that the compiler analyzed EMH expressions as monosyllabic entities. It can be posited that this conception somehow resulted from the recitation of characters under the instruction of native tutors. When looking at the information categories of each entry, however, it becomes obvious that the analytic approach in the Bocabulario went far beyond the listing of monosyllabic readings. Explanations are surprisingly wide-ranging, clearly distinguishing phonological, semantic, and morphological aspects of language analysis. In some instances, an entry even provides detailed cultural information associated with the expression under discussion. A transcript of a typical entry reads as follows (Bocabulario f. 18r): Ban

pronunçia se llanamente, sigca poco a poco o despacio. Redoplase de ordinario. U. g. ban ban sîo lu cheg coa aù lay chia ké chuē. peinsa despaçio tus peçados y despues ven a confesar. ban tâng lay no. Ven poco a poco […]

Ban

Al mismo tono. Diez mil. U. g. cheg ban diez mil no ban Veintemil sa ban Treintamil &a […]

[Ban

It is pronounced plainly, it means ‘little by little’, ‘slow’. It is normally reduplicated. For example, ban ban sio lu cik cua au lai cia ke cue. ‘Think slowly about your sins and then come to confess.’ Bang tang lai no. ‘Come little by little’ […]

Ban

In the same tone. Ten thousand. For example, cik ban ‘ten thousand’; no ban ‘twenty thousand’; sa ban ‘thirty thousand’; etc. […]]

––––––––

5 One of the previous owners was the German orientalist Julius Klaproth (1783– 1835) who allegedly stole the manuscript from the Royal Library in Berlin (cf. van der Loon, 1967:107).

70

CHAPTER THREE

As in the given example, most entries first indicate pronunciation and meaning of the headword. Explanations on pronunciation are often introduced with pronunçiado ‘pronounced’ or pronunçiase (also abbreviated as pse) ‘to be pronounced’. Interestingly, different manners of pronunciation are explained in the entry, but not encoded in the transcription of words. Tone diacritics are used occasionally only. In general, the transcription system is rather unsystematic with regard to tones as well as other phonological distinctions, such as aspiration and nasalization. A whole range of terms are used to describe pronunciation, some of which are listed in Table 4. Table 4: Terms referring to pronunciation in the Bocabulario Spanish a la dentro abierta la boca afecto altillo alto baxa breue

English in the inside mouth open

Spanish detenido en las narizes

English detained in the nose

affected risen high low short

claro con rregano de golpe

clear with a scold with a hit

gangoso llanamente llano melindroso raspado en el pecho reçio riendose seco

nasal plainly flat fussy scratchy in the chest strong laughing dry

The denoted phonetic properties of some terms are quite obvious. Nasality, for example, is described with terms like en las narizes ‘in the nose’ or gangoso ‘nasal’. Different pitches are indicated by terms such as alto ‘high’ or baxa ‘low’. In many cases, however, it is hardly possible to establish clear links between the terms used in the Bocabulario and modern phonological terminology. Tonal differences or correspondences between successive headwords are often indicated, as in the example on p. 69. The entry for the second headword «ban | diez mil» ‘ten thousand’ on folio 18r starts with Al mismo tono ‘in the same tone’, i.e., in the same tone as the preceding headword. The abbreviation sigca for significa ‘means’ precedes the Spanish equivalent of the EMH headword. Sentences exemplifying the usage

LEXICOGRAPHY

71

of headwords are indicated with V.g. abbreviating verbi gratia ‘for example’. Compared with other EMH sources, notably the Arte, the high number of terms for the description of pronunciation is unusual. But then, the Arte—unlike the Bocabulario—uses a complete transcription system rendering all phonological distinctions. It can thus be assumed that the compiler of the Arte had a more systematic understanding of phonological contrasts than the compiler of the Bocabulario, whose usage of pronunciation-related terms is rather unsystematic. A possible conclusion is that around 1620, when both the Bocabulario and the Arte (BMS) were compiled, the phonological contrasts of EMH had been recognized, but this recognition had not yet materialized into a broadly-accepted descriptive terminology and romanization system. The many explanations regarding pronunciation—albeit not in the context of a consistent terminological framework—show that the compiler recognized the importance of analyzing linguistic form. Apart from his treatment of pronunciation, meaning, and usage, he also explains morphological aspects, and in doing this shows a clear understanding of the issues in question. Free and bound morphemes, for instance, are implicitly distinguished in the following entry on folio 8r: Aň

Preposiçion o adueruio que por si solo no se usa. pero allegado con otros sigca por V.g. por donde fue an te lo cû’ por alla an hu te por manila an bin hi la cû’ por aqui an chi te.

[Aň

Preposition or adverb which is not used on its own, but is linked to others. It means ‘by way of’. For example: ‘Where did he go’ an te lo cû’; ‘there’ an hu te; ‘to Manila’ an bin hi la cû’; ‘hither’ an chi te.]

Other entries also show that the compiler of the Bocabulario was obviously aware of word formation patterns. For example, under the section heading (postponed tau), he lists the following entries with the suffix -thau5, lit. ‘head, top’, a very productive Hokkien suffix.

72

CHAPTER THREE

Table 5: Entries with postposed tau Entries Puntau Soantau Tutau cátauh Poutauh Bintau

Translation ca

sig señor sigca ajo legumbre sigca asada o açadon sigca La Rodilla sigca hacha. decortar sigca pan de castilla

mister, gentleman garlic hoe knee axe Castilian bread

The length of entries in the Bocabulario varies considerably. As in Table 5, some entries are confined to providing Spanish equivalents of EMH headwords. Some entries spread across three pages, providing explanations of various example sentences or explaining cultural conventions associated with particular headwords. For example, the entry «Bee | 馬» ‘horse’ not only lists various EMH collocations containing the morpheme be2 ‘horse’, but also introduces the ‘horse’ as a Chinese zodiac sign. The entry reads as follows (ff. 21r–22r): Bee 馬 Como haçiendo destello? despreçiando. Largo bayo u suaue. U.g. be lang el cavallo o Roçien o quartago bebo la yegua be jan Potro chio be subir a caualla Lo be apearse chuan be correr le poa Lobo caer de la cabalgadura. Be uã ensillar el cauallo uã la silla chí/chì be Regalar el cauallo chia be cauallo Rojo o uaio pido o infi / sillar pibe ensillar o aparejar el cauallo be pang lay es el establo estos sangleyes tienen a modo de los sig.nos o planetas que rreinan en el çielo signo de aries signo de pisçis &a. dose signos que cada uno dellos Reyna en un año entero y corren sus dose años ?hasta beluel otra buelta Uorden siruen estos para los hombres qundo nazen y dizen yo nasi en el reino de tal yo en el detal &a V.g. son los siguintes el primero el raton. chu. yasi el que naçio en el ano que reina ua el raton disi gua si siũ chu dize yo tengo el signo del rraton nasi en el signo de raton aquel siu muy gueco sig.ca el tal signo el seg.do es la baca gu: gua siũ gu nasi en el de la boca [sic] y asi sienpre el 3.o hou tigre El 4.o El conejo touh El 5.o La serçierpente leng. alto. El. 6.o La culebra chua que es el que predomina el año de 1617 el 7.o es Predomina el año 1617. El cauallo este bee Reina el año de 1618. El 8.o es la cabra o cabrito yõ muy gueco El 9.o es la mona o mico cau El 10.o La gallina Kei El 11.o El perro cau. El 12.o El puerco tu. [As if emitting sparks? with contempt, long, low, and gentle. For example becang ‘horse’ or ‘nag’ or ‘pony’, bebo ‘mare’, be jan ‘colt’, chio be ‘get on a horse’, lo be ‘shackle a horse’, chuan be ‘ride a horse’, poa lobo ‘fall off the horse’, Be uã ‘saddle the horse’, uã ‘the chair’,

LEXICOGRAPHY

73

chí/chì be ‘give a horse’, chia be ‘red horse or bay’, ? pibe ‘saddle or prepare the horse’, be pang lay is ‘the stable’. These Sangleys use for their signs or planets in Heaven, such as the Aries sign, the Pisces sign, etc. twelve signs, each of which reigns for one whole year, run [over a period of] twelve years until the return of another round or order. They serve men to indicate when they are born by saying I was born in the reign of such and such in the such and such. The examples are the following: The first one is the rat chu. Hence, one who is born in the year reigned by the rat would say gua si siũ chu, meaning ‘I have the sign of the rat, I was born in the sign of the rat.’ This siu [pronounced with] much resonance, means ‘this particular sign.’ The second is the cow gu: gua siũ gu ‘I was born in the one of the [cow].’ The third is hou ‘tiger’. The rabbit touh is the fourth. The fifth is the dragon leng, high [pronunciation]. The sixth is the snake chua. This is the one which predominates in the year 1617. The seventh, the horse—this bee—[is reigning in the] year 1618. The eighth is the goat or the kid yõ, [pronounced] with much resonance. The ninth is the monkey cau. The tenth is the hen kei. The eleventh is the dog cau. The twelfth is the pig tu.]

F. TWO ANONYMOUS AND UNDATED DICTIONARIES IN THE UST ARCHIVES Two bilingual Spanish-EMH dictionaries are kept in the University of Santo Tomás Archives (UST) in Manila. These are the Dictionario Hispanico Sinicum (hereafter: Dictionario) and the Vocabulario de la lengua Española y China N. 1 (hereafter: Vocabulario). Both manuscripts are undated and of unknown authorship. Although the first folios of both manuscripts are missing, their main parts are still extant. The Dictionario contains ca. 20,000 entries on 552 double-sided folios, and the Vocabulario ca. 17,000 entries on 330 double-sided folios. For comparison, the Bocabulario kept in the British Library contains 1,500 entries, and Chirino’s Dictionarium a little more than 1,000 entries. According to González (1966:412), the UST archives must once also have possessed a unidirectional Hokkien-Spanish dictionary of almost 400 double-sided folios. However, the manuscript was missing when I visited the archives in 2006. There is some evidence that the two dictionaries can be placed in the same historical context as the other EMH dictionaries introduced above. The rather cumbersome romanization system is almost the same as the one used in the Arte. There is also some overlap in the use

74

CHAPTER THREE

of linguistic terminology. For example, the Arte contains a list of 79 Hokkien classifiers. These are labeled as ‘proper numerals for counting particular things’ (numerals propios para contar cosas particulares, see Chapter 4). Hokkien classifiers recorded in the Dictionario are likewise labeled as ‘numeral’. Internal evidence in support of the assumption that the dictionaries date back to the first half of the seventeenth century is the EMH translation of ‘China’ which is «大明囯 | dǎy bēng côg'», the name of the Míng dynasty. No reference to the Qīng 清 dynasty (1644–1911 CE) can be found. Moreover, as pointed out in Chapter 2, it is known that the Dominicans shifted their focus of language learning from Hokkien to Mandarin when they were about to enter China in the 1630s. It seems likely that such a comprehensive Hokkien dictionary was compiled before this shift of focus occurred. Taking these pieces of evidence into consideration, it can be assumed that the dictionaries were compiled before or during the 1630s. On the basis of lexical comparison, the target language of both extant manuscripts can easily be identified as Hokkien. Table 6 compares characteristic Hokkien vocabulary as listed in Branner (2000: 95–96) and Norman (1988:231–232) and the corresponding translations in both Manila manuscripts. In order to facilitate comparison, EMH expressions are transliterated on the basis of the analysis in Chapter 5. Table 6: The representation of Hokkien lexicon in the Manila manuscripts Expression ‘cut with scissors’ ‘field’ ‘foot’ ‘house’ ‘leaf’ ‘many’ ‘matter’ ‘mouth’ ‘one’ ‘person’ ‘rain’

Spanish

Amoy 1

cortar

ka

campo pies casa hoja mucho negocio boca uno hombre lluvia

chan5 kha1 chu3 hioh8 cue7 tai7 chui3 cit8 lang5 hoo7

Dictionario Vocabulario ka1

ka1

chan5 kha1 chu3 / chu7 hioh8 cei3 tai3 chui3/chui7 cit8 lang5 hou7

chan5 kha1 chu7 -cei7 tai3 chui7 cit8 lang5 hou7

75

LEXICOGRAPHY

Table 6 (cont.) Expression ‘short’ ‘son’ ‘take off (clothing)’ ‘tasteless’ ‘wear’ ‘wok, cooking pot’ ‘wood’ ‘year of age’

Spanish

Amoy 2

corto hijo quitar

te kiann2 thng3

Dictionario Vocabulario te2 kiann2 thui7

dessabrido chiann2 chiann2 vestir ching7 ching7 caldera tiann2 -Madera año

cha5 he3

cha5 hue7

te2 kiann2 thui7 chiann2 ching7 -cha5 hue7

The arrangement and sequence of entries in both dictionaries are almost identical. Spanish headwords are listed in alphabetical order, followed by EMH equivalents. In both dictionaries, a sequence of headwords commences with a word representing a superordinate concept, followed by sub-headwords representing special meanings of the main headword. In the Dictionario, the first letter of a main headword is capitalized. A sequence of sub-headwords following the main headword ‘disturb’ is shown in Table 7. Table 7: The headword and its sub-headwords (Dictionario) Spanish

EMH

English

Alborotar

擾 xiaù 擾乱 xiaù loǎn 擾攘 xiaù xiàng 草閙 chaùc lâu 變乱 piên loǎn 反乱 huàn loǎn

disturb

alborotar, inquietar alborotar, haçer ruido alborotar, reboluer alborotar, estar todo rebuelto alborotar la tierra

disturb, unease disturb, make noise disturb, upset disturb, being in a total mess shake the earth

76

CHAPTER THREE

Table 7 (cont.) Spanish

EMH

English

alborotar, haçer ruido

嚷嚷 xiàng xiàng 如如嚷嚷 xî xî xiàng xiàng

disturb, make noise

alborotado todo, rebuelto

all messy, in a mess

As can be seen, the general meaning of the main headword is specified in two ways: (1) by indicating the particular meaning of the subheadword with another expression, as ‘disturb, make noise’; and (2) by placing the sub-headword in the context of a Spanish collocation, as ‘shake the earth’. These methods of disambiguating the general meaning of headwords are not applied consistently. Instead, sub-headwords differ considerably in the degree to which headwords are disambiguated. Compare, for instance, the sub-headwords under ‘rough, roughness’ (Table 8) and those under ‘uncle’ (Table 9). In the former case, the semantic distinctions between the eleven Hokkien renderings are not reflected in specifications of the Spanish sub-headwords. In the latter case, sub-headwords are at least partly disambiguated by providing brief explanations of how EMH words for ‘uncle’ are distinguished in terms of features older vs. younger and paternal vs. maternal side. Table 8: The headword and its sub-headwords (Dictionario) Spanish

Hokkien

Rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça rudo, rudeça

惛 hún 訰 tǔn 魯 lòu 拙 chuâr' 愚 gū 魯訰 lòu tǔn 剪 chièn loâr' 愚魯 gū lòu 拙見 chuâr' quiên 拙想 chuâr' siôn

77

LEXICOGRAPHY

Table 8 (cont.) Spanish

Hokkien

rudo, rudeça rudo, rudeça

愚蠢 gū chùnc 訰口 tǔn chǔic

Table 9: The headword and its sub-headwords (Dictionario) Spanish

Hokkien

Tio [uncle] tio, hermano mayor de mi padre [uncle, older brother of my father] tio, hermano mayor de mi padre [uncle, older brother of my father] tio, hermano mayor de mi padre [uncle, older brother of my father] tio, hermano menor de mi padre [uncle, younger brother of my father] tio, hermano menor de mi padre [uncle, younger brother of my father] tio, hermano mayor de mi abuelo [uncle, older brother of my grandfather] tio, hermano de mi abuelo [uncle, brother of my grandfather] tio, casado con hermana de mi padre [uncle, married with my father’s sister] tio de parte di madre [uncle on mother’s side] tio ançiano y distante [old and distant uncle] tio mas moso queyo distante [uncle more than I distant] tio proprio [own uncle]

叔 chêg' 伯 pê:' 大伯 tôa pê:' 外伯 guâ pê:' 二叔 xˇy chêg' 列位叔伯兄弟 lièr' uˇy chêg' pê:' hián tỳ 伯公 pê:' cóng 叔公 chêg' cóng 姑丈 cóu tiǒn 細舅 seỳ cǔ 堂伯 tōng pê:' 堂叔 tōng chêg' 家叔 qué: chêg'

78

CHAPTER THREE

The Dictionario and the Vocabulario not only correspond in their lexicographic macro structure, the selection of headwords and sub-headwords is also largely identical. One difference between the two dictionaries is that in the Dictionario, EMH translations are represented in Chinese characters and romanized transcriptions, whereas the Vocabulairo uses transcriptions only. As shown in Chapter 5, the romanization system is almost the same as in the Arte. In contrast to Chirino’s system, it unambiguously renders seven Hokkien tones. Another difference is that the first part of the Dictionario contains a second column with romanized transcriptions. Many of these transcriptions resemble the Mandarin transcriptions in Varo’s glossaries of the late seventeenth century (in Coblin, 2006). One example is ‘whip, flog’, which is transcribed as in both the added column in the Dictionario and in Varo’s glossary of the Mandarin language. For ‘loosen’, we find in the Dictionario and in Varo’s glossary. It thus seems as if the compiler of the Dictionario— or a later scribe—tried to expand the bilingual Spanish-EMH dictionary to a trilingual Spanish-EMH-Mandarin dictionary. Without obvious reason, this attempt was discontinued on f. 210v. In sum, the Dictionario is more tentative than the Vocabulario. Whereas the former, in addition to the incomplete column with Mandarin transcriptions, is full of corrections, additions, and untranslated headwords, the manuscript of the Vocabulario is coherent, neat, and almost free of corrections. It is therefore obvious that the Vocabulario is a revised version of the Dictionario. How did the missionary compilers select headwords? The answer to this question reveals much about the nature of missionary language research. When analyzing bilingual dictionaries like the Vocabulario and the Dictionario, it must be borne in mind that in the seventeenth century, bilingual Spanish lexicography was still a relatively young discipline. It had received an impetus from Antonio de Nebrija’s Dictionarium ex-hispaniensi in latinum sermonem (also known as Vocabulario de romance en latín), presumably written in 1495 and revised by Nebrija. Nebrija’s dictionary is the first bilingual dictionary with Spanish as source language (Haensch, 1990:1739). Soon after its publication, it became a model dictionary for missionaries compiling bilingual dictionaries of native American languages (ibid.). One famous example for this expansion of the Nebrijan lexicographic tradition into the Americas is Alonso de Molina’s (?1513–1585) Vocabu-

79

LEXICOGRAPHY

lario en lengua castellana y mexicana of 1571, the first part of which is largely based on Nebrija’s dictionary (Hernández ,1996:23). In the light of the fact that, until the end of the seventeenth century, most Spanish missionaries came to Asia via Mexico (Mungello, 2005: 34–35), it seems very likely that the Manila-based missionaries were familiar with the works of Nebrija and Molina. Given that it was common practice to compile new dictionaries on the basis of existing works (Cooper, 1962), it may also be assumed that earlier bilingual Spanish dictionaries inspired the compilers of the Manila manuscripts. This assumption could only be tested on the basis of a large-scale comparison of entries, which would be beyond the scope of this chapter. Table 10, which compares the degree of headword overlap in a sequence of 40 headwords selected from Nebrija (1516/1981), Molina (1571/1970), and the Dictionario, is therefore only intended as a preliminary indication. The sequence of headwords follows the alphabetical order in Nebrija’s dictionary, starting from and ending with .6 Sub-headwords have not been considered. Table 10: Entries in Nebrija (1495?), Molina (1571), and the Dictionario

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Headword

Nebrija

Molina

Dictionario

acabar ‘limit, finish’ a cada ‘at every’ acaecer ‘happen’ acallar ‘calm’ a canales ‘by channels’ acariçiar ‘caress’ acarrear ‘carry’ a caso ‘chance, by chance’ acatar ‘observe’ acaudalar ‘amass’ acaudillar ‘lead’ acceleradamente ‘hurried’ aclarar ‘clarify’

   - -     -

   - -     -

 -- -   --- 

–––––––– 6

The headword ‘caress’, found in the Dictionario only, has been included, as it falls within this alphabetic order. On the other hand, all headwords starting with , which Molina lists between and , have not been considered, as they occur after and in Nebrija’s dictionary.

80

CHAPTER THREE

Table 10 (cont.) 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Headword acocear ‘tread, oppress’ acodar ‘lean on’ acoger en casa ‘receive into a house’ acometer ‘get into, enter upon’ acomodarse ‘accommodate oneself’ acompañar ‘accompany’ aconsejar ‘advise’ acontecer ‘happen’ acordar ‘remember’ acordes en musica ‘harmony in music’ acorrer ‘help, aid’ acorrucado ‘curled up’ acortar ‘shorten’ acossar ‘drive, lead’ acostarse ‘lay down’ acostumbrar ‘accustom’ acrecentar ‘increase’ acreditar ‘guarantee’ acreedor ‘creditor’ acrisolar ‘refine’ acto carnal ‘carnal act’ acudir ‘attend’ aculla ‘there’ acusar ‘accuse’ açacan ‘water-carrier’ açada ‘dibble’ açafran ‘saffron’

Nebrija   

Molina   --

Dictionario ----





--

--

--



 -  

    --

  {} {} --

 --  - ----     

-      - --     --

-- --         ----

Based on the data in Table 10, 77 percent of the headwords occurring in Nebrija (1516) also occur in Molina (1571). This is consistent with the claim that Molina strongly relied on Nebrija’s work. However, the overlap between Nebrija and the Dictionario (28 percent) and between Molina and the Dictionario (47 percent) is much lower. Interestingly,

LEXICOGRAPHY

81

the two headwords ‘happen’ and ‘remember’ do appear in the Dictionario, but they seem to have been inserted later and are not translated. This insertion may represent a belated attempt to harmonize the selection of headwords with a lexicographical model, maybe Nebrija or Molina. All in all, however, the low percentage of shared headwords suggests that headword selection in the Dictionario to a lesser degree followed the track beaten by Nebrija and Molina, but was instead inspired by the compiler’s genuine encounter with the target language. This is consistent with the disambiguation of subheadwords described above. As in the case of ‘uncle’ (Table 9), numerous sub-headwords do not disambiguate polysemous Spanish headwords, but instead paraphrase lexical fields within the target language. This presupposes a considerably deep understanding of the target language by the missionary lexicographer. G. SUMMARY AND CONCLUDING REMARKS The extant EMH dictionaries show that manuscript documentation of Chinese regional varieties was richer and more diverse than previously assumed. Generally speaking, the significance of EMH dictionaries lies in the wealth of recorded EMH language data and the diverse ways the data are recorded. Thus, each title introduced in this chapter provides unique support for my analysis of EMH and its documentation. As will be shown in Chapter 4, the use of hybrid gloss-translations in Chirino’s Dictionarium resurfaces in the Arte. On the basis of example sentence arrangement in the Dictionarium we know that the use of gloss-translations in both sources was no coincidence, but very likely a common device of sentence analysis. Further comparative research is needed to verify this claim. A distinctive feature of the Bocabulario lies in its analysis of morphosyntax and its terminology for the description of tones. As will be shown in Chapter 5, some metalinguistic evidence contained in the Bocabulario is important for the analysis of EMH entering tones. Finally, the Dictionario and the Vocabulario show that in terms of sheer breadth, EMH lexicography reached unprecedented dimensions. The data recorded in these two dictionaries fills many gaps in my analysis of EMH phonology and orthography in Chapter 5. The solving of some of the open questions concerning EMH phonology and research on EMH lexicon will require a complete digitalization of these sources.

82

CHAPTER THREE

As far as I know, the missionaries in the Philippines did not write any documents in which they reflected on the nature of the Chinese language. However, as shown in this chapter, conceptualizations of language do not necessarily have to be explicitly expressed in words, but can also be deduced from the arrangements and contents of dictionaries. A compiler who records polysyllabic Chinese words in the Roman alphabet is unlikely to believe in the monosyllabic myth. He is also unlikely to have thought of Chinese characters as relevant analytic reference points. In any case, the dictionaries left behind by the Philippine-based missionaries show that myths relating to the Chinese language, fostered by European scholars after the sixteenth century, did not accurately reflect the state of affairs in missionary linguistics.

CHAPTER FOUR

LANGUAGE AND METALANGUAGE A. INTRODUCTION In the conclusion of the preceding chapter, I argue that conceptualizations of language can be deduced from the arrangements of dictionaries. This conclusion leads to the more basic question of how language is analyzed in missionary sources, notably the Arte. The rationale underlying this question is that any conclusion about historical language data derived from a document analysis requires a proper understanding of the various explanatory devices employed in a particular source. Such devices could be based on words (e.g., an actual explanation of the idea, expressed in words) or could be independent of words (e.g., a diagram, a table or list, a line or arrow connecting two drawings, bolding, italics, etc). Linguistic explanations are associated with the term metalanguage. My use of the term metalanguage is based on Koerner’s definition, as quoted in Chapter 1. To repeat, according to Koerner, metalanguage is the ‘the use of language for the description of linguistic concepts, ideas or theories of earlier periods’ (1989:31; see also Koerner, 1987). This definition effectively narrows ‘metalanguage’ down to concepts that are expressed in words. One of the claims of this chapter is that this narrow interpretation of metalanguage disregards many explanatory devices that are not bound to explicit wording. It is of minor importance whether non-terminological explanatory devices should also fall under category ‘metalanguage,’ as this is merely a matter of definition. The crucial point is that not all explanatory devices require words. In earlier studies into missionary linguistics, the criticism has been frequently made that missionary metalanguage was heavily influenced by the so-called Greco-Latin grammar model. The first section of this chapter challenges this criticism by arguing that it disregards the multiple roles of the missionary linguist. Examining different explanatory devices employed in the Arte, the second section shows that explanatory devices go beyond the application of existing terms. In the last section, I show that an analysis of explanatory devices is not the only

84

CHAPTER FOUR

path that leads to an understanding of earlier stages of a language. A comparison of language data with sources of the same period that were not compiled by missionaries is necessary if we are to understand epistemological fallacies of metalanguage. B. PARADIGM AND PURPOSE Chapter 1 quotes Even Hovdhaugen’s dictum that missionaries were the first professional linguistic fieldworkers. There can be little doubt that the language documentation by missionaries was derived from activities resembling fieldwork in modern times. But was it indeed conducted in a professional manner? Previous research has asserted that linguistic sources of any kind compiled by missionaries have no value to the modern linguist. Some important points of criticism are addressed in the following quotation by Angulo (1925:98). Anyone who has seen some of the grammars published by the early friars knows that they are pretty nearly worthless from every point of view […] I derived not a little amusement from the fantastic grammars done on a Greco-Latin model by the good priests. I will admit that a good many Zapotecan sounds are rather puzzling to the Spanish ear and present quite a problem in the phonetic transcription. But it had never occurred to me that the easiest way to find out of the difficulty was to write something entirely different.

The first point of Angulo’s criticism pertains to the theoretical framework employed in most missionary grammars. Applying terms and distinctions defined by Antonio de Nebrija (see Chapters 1 and 2), missionary linguists employed the paradigms and methods of traditional European grammar, i.e., the Greco-Latin model. This methodology is generally considered inadequate when the languages described are typologically different from Indo-European languages. Hokkien dialects are a case in point. Like all other Chinese languages, Hokkien dialects have tones and little derivational morphology. Thus, at a first glance, it seems that, for two reasons, the application of Greco-Latin grammatical paradigms would yield distortion. First, the analysis may read something into the described language which does not exist. After all, what insight does a chapter provide that investigates declensions and conjugations in a language without derivations? Second, the Greco-Roman framework may mask linguistic features of the described language which are foreign to Indo-European languages.

LANGUAGE AND METALANGUAGE

85

Linguists today likewise tend to be cautious about the analytic value of missionary sources. This attitude has much in common with Angulo’s criticism quoted above. Chappell, for example, writes (2006: 442f.): Since a Greco-Latin model is one that is evidently designed for highly inflectional languages, it is not surprising to find standard chapters in missionary grammars include the topics of verb conjugations with accompanying abundant descriptions of tenses […]. Consequently, information is not always presented economically or concisely. […] From this it should not be inferred that grammatical analyses from this period are completely inadequate, but rather that there is a distinct mismatch between the structure of the target language and the framework for analyzing it. Hence, in this early phase of grammar writing for Chinese languages, it is reasonable to claim that there was little influence of language type on the design of grammars for Chinese languages.

In China, the well-known linguist Wáng Lì 王力 (1900–1986) once criticized Chinese grammar studies by Westerners as a ‘propositionproving method where one would have to find an explanation for the exceptions’ (1936:94, quoted and translated by Pan and Tham, 2007: 137). Exceptions are, thus, defined and specified on the basis of IndoEuropean languages. In support of this criticism, Pan and Tham argue that this ‘is exactly what missionary grammars were all about a few centuries ago’ (ibid.). Recent scholarship has linked criticism of analytical approaches to discourses on colonialism and politics of language. Gilmour, for instance, argues that ‘[i]n various ways, representations of languages implicitly or explicitly laid out parameters for what constituted appropriate communication between colonizing and colonized people’ (2006:3). In a similar vein, Errington frames colonial linguistics as ‘a nexus of technology (literacy), reason, and faith and as a project of multiple conversion: of pagan to Christian, of speech to writing, and of the alien to the comprehensible’ (2001:21). It is exactly because of this undeniable link to European colonialism that some historians of linguistics seem to deny any significance to missionary sources. For instance, in his seminal study of Western linguistic history, Seuren argues that ‘[t]his kind of early [missionary] grammar writing […] was less inspired by a desire to know more about the languages concerned than by sheer imperialism’ (1998:53). Reduced to its essentials, the criticism comes down to arguing that missionary linguistic analysis is a theoretical construct detached from

86

CHAPTER FOUR

linguistic reality and driven by imperialistic motivations. To be sure, missionary sources are not short of examples supporting this criticism. However, mainly for two reasons, the criticism is itself problematic. First, by taking a hindsight view that is rooted in modern scholarship, it runs the risk of falling into the same epistemological pitfalls as its object of criticism. After all, how meaningful are quality statements which are based on standards other than those relevant for the missionary linguist? Such a perspective inevitably conceals various dimensions of missionary linguistics. Most importantly, missionary sources do not demarcate linguistic analysis, language teaching, and language planning. It is therefore misleading to read analytic paradigms from a purely analytic perspective; one also has to carefully check whether a particular source was part of a language curriculum. Moreover, only an analysis of the aims and intended audience of the language curriculum will enable us to place a particular source in the larger context of missionary language planning. As I will show presently, there is clear evidence that the Arte was used as some kind of teaching aid. It is also obvious that it was used for teaching Hokkien to European missionaries. As was pointed out in Chapter 2, no historical evidence has been presented which would enable us to link the Arte to a missionary language planning program directed at the Sangleys in the Philippines. Second, the criticism seems to be derived from the premise that there is something like theory-free data elicitation that produces an objective representation of a language. This claim has been countered by several experienced fieldworkers. Bowern, for example, explicitly states that theory ‘is inherent in research,’ arguing that ‘[a]s soon as anyone uses a metalanguage for natural language description, they are making choices, categorizing and labeling their data […] There’s no such thing as theory-neutral or atheoretical linguistic description’ (2008:11). Similarly, Crowley writes that ‘[n]o descriptive fieldworker can carry out linguistic analysis of a previously undescribed language in a theoretical vacuum. We therefore need the insights that theoretical linguistics can offer in order to guide our search for new data, and to influence our analysis of the data’ (2007:13). Thus, data elicitation entails theoretical concepts, and is by definition selective. The selective nature of linguistic fieldwork is emphasized by Mithun when she asserts that ‘linguists do more than hold the microphone. They shape the record in obvious ways, such as selecting certain lines of research

LANGUAGE AND METALANGUAGE

87

and methodology’ (2001:51). Selection is not restricted to theoretical and methodological considerations, but also pertains to the question as to which informants may be considered representative of a language. In this respect, Branner argues that ‘[i]t is the field-worker who creates the primary material, which is merely one particular sample of the ideal dialect’ (2000:37). Thus, reading missionary sources obviously requires an approach that goes beyond pointing to isolated examples of discrepancies between the Greco-Latin paradigm and the analyzed language. Instead, we must carefully examine how a particular metalinguistic paradigm, a particular term, or an explanatory device not bound to language— the explanans—is juxtaposed with the explanandum, the language data. Ideally, this juxtaposition is complemented by linguistic data selected from contemporary non-missionary sources. Data from the Arte, for example, can be compared with data recorded in early editions of Hokkien stage plays. C. EXPLANATORY DEVICES: FROM NON-TERMINOLOGY TO NEW TERMINOLOGY As explained in Chapter 1, I examine influences manifest in the Arte on the basis of the criteria proposed by Koerner (1989), i.e., (1) the intellectual background of the author, (2) textual parallels, and (3) direct references to previous works. When applied to the Arte, the first criterion falls out of the scope of the present investigation, as authorship cannot be established with certainty. References to previous scholarship are also barely made. Therefore, the criterion that may be applied are textual parallels and, to a lesser degree, the third criterion. In very general terms, the Arte employs linguistic terms that are clearly part of a Nebrija-inspired tradition of missionary linguistics. The label Arte alone links the document to a particular intellectual tradition. According to Breitenbach, a designation ‘as “arte” denotes the grammatical arts and refers to a linguistic tradition which distinguishes ars, the grammatical arts, from scientia, science’ (2000:xxiv; the dichotomy ars vs. scientia is also discussed by Bossong, 1990: 191ff.). Referring to Verburg (1981:208–209), she explains that (2000: xxxiv): The trivial artes revealed a certain Humanist trait […] At times, rhetoric played a more influential role than grammar. Rhetoric, which had, in the

88

CHAPTER FOUR

opinion of Verburg, always been the most dynamic part of the sermonical artes of the Trivium, aimed at an efficient degree of eloquence as the optimum of speech. Since the beginning of humanism a central focus was to value ‘skills’ above ‘knowledge,’ ‘action’ above ‘abstraction,’ i.e., to value the practical and applicable above ‘abstract thinking,’ ‘competence’ above ‘contemplation,’ and ‘content’ above ‘doctrine.’

Such a conception of grammar certainly cannot be attributed to one specific scholar. Breitenbach points to the important position of the Dutch humanist Gerardus Joannes Vossius (1577–1649), who in turn was influenced by the Italian Lorenzo Valla (1406–1457). In other words, when the Arte was composed in the seventeenth century, linguistic thinking had been shaped by a number of thinkers, and it would be wrong to place too much emphasis on Nebrija alone. We also have to take into consideration that grammatical analysis aiming at practical applicability met the fundamental needs of missionaries. A missionary simply had no other choice but to learn the language of his immediate surroundings. It is thus no coincidence that the achievement of language skills is a characteristic feature of missionary linguistics. Whereas some ‘humanist traits’ in grammar writing are visible in the Arte, it also has to be emphasized that missionary linguistics in general and the Arte in particular represent a dissociation from humanist traditions. As pointed out in Chapter 2, the choice of vernacular languages as an object of description is obviously not compatible with the humanist ideal of revitalizing spoken Greek and Latin (Kramer, 2000). Against this background, the extended scope of grammar writing represented by missionary linguistics must be considered a tradition by itself, and not a mere relocation of one particular intellectual tradition into non-European contexts. I therefore concur with Percival, who argues that ‘the age-old curricular subject grammatical, which up to the fifteenth century had confined its attention exclusively to Latin, spawned not only vernacular grammatical traditions, but also a new branch of knowledge, subsequently called linguistics’ (2007:111; see also Percival, 1975/2004, 1976/2004, 1983/2004). The analytical approach in most passages of the Arte is not original, as it fits into a pattern found in several missionary grammars. Most of its explanatory sections are brief and sketchy in style. They are followed by rather long sections of examples of concrete usage. Chapters 8 (‘On the Ways of Asking and Responding’) and 9 (‘On the Mode of

LANGUAGE AND METALANGUAGE

89

Composition’) specifically focus on pragmatic and stylistic competence in EMH. Note that ‘composition’ here does not refer to the writing of texts, but to the formation of spoken sentences. This approach is, for example, reminiscent of that taken in the Tarascan grammar of the Franciscan friar Maturino Gilberti (1558/2004; Monzón, 2005, 2007),1 which closely resembles that in the Arte. In this respect, it is important to reiterate that in the early seventeenth century, many approaches in missionary linguistics, including the adaptation and permutation of linguistic terminology, had already been shaped in the Americas (see Chapter 2). Thus, like the field of missionary lexicography (see Chapter 3), missionary grammar writing in Asia owed many of its concepts to the work that had previously been done in the Americas. Therefore this body of missionary linguistic literature must be considered as a separate source of influence that shaped the analytic approach of later missionary grammarians. Concretely speaking, when the compiler of the Arte distinguishes between linguistic form and meaning by using terms like vos ‘word’ and significaçion ‘meaning’, this may represent terminological influence from Nebrija. It is, however, equally possible that he uses the terms because he had read another’s missionary use of Nebrijan terminology. As I have pointed out in Chapter 3, many missionaries reached Asia via Mexico. This strongly supports the notion that missionary linguists reaching Asia were familiar with the work by their confrères in the Americas. In sum, reference to Nebrija alone is insufficient when it comes to understanding the various explanatory devices. As shown in the following sections, missionary grammarians use a number of explanatory devices which would not be classed as the application of Greco-Latin grammatical paradigms. These devices include: (1) the use of transcriptions and the arrangement of examples, which are not related to terminology in any way; (2) the avoidance of terminology; (3) the questioning of existing terminology by denying its relevance; (4) the redefinition of existing terminology; and (5) terminological innovation. The redefinition of existing terminology is not necessarily explicit; in many cases the use of a particular term

–––––––– 1

Tarascan is spoken in western Mexico; the modern language is also known as P’urhépecha (Monzón, 2005:65).

90

CHAPTER FOUR

with reference to a particular syntactic construction implies terminological redefinition even if this remains unexplained. 1. Non-Terminological Devices 1.1. Macrostructure The macrostructure of a source provides important clues to understanding how the compiler organized his analysis and what influences are manifest in this arrangement. Although the contents of the nine chapters in the BMS and the LMS are almost identical, there are a few differences. The BMS is longer (33 folios) than the LMS (23 folios). Additional folios of the former contain five untitled sections on respectively numerals, measure words, units of measurement, days of the week, months, years, and monetary units. Apart from explanatory notes on measure words and three examples, all these sections are missing in the LMS. The BMS has the following sequence of chapters: On the mode of pronouncing this language (f. 1r) On declensions (f. 2v) On the conjugation of the verb (f. 5v) On adverbs (f. 8v) On other adverbial particles (f. 10r) On conjunctions (f. 12r) On negations (f. 12v) On the ways of asking and responding (f. 13r) On modes of composition (f. 14r) [numerals, f. 16v] [measure words, including units of measurement f. 23r] [time, f. 30r] [money, f. 32v] [units of measurement f. 33r] From the sequence of chapters it can be seen that the Arte considers pronunciation more fundamental than declension or conjugation—the chapter ‘On the Mode of Pronouncing this Language’ is the first one. Analyzing a similar sequence of chapters in Varo’s Mandarin grammar, Breitenbach (2005:68) observes that such a treatment of pronunciation is not in accordance with the grammar model set by Nebrija, which treats pronunciation after declension and conjugation. It must

LANGUAGE AND METALANGUAGE

91

be noted, however, that numerous missionary grammars of the same period likewise start with sections on pronunciation and orthography. One example is John Eliot’s grammar of the Massachusett language (Eliot, 1666; cf. Swiggers, 2007). We may therefore assume that such an arrangement arose from the missionary linguistic tradition itself, obviously as a response to the difficulty beginning learners encounter in pronouncing the words of an unknown language. After all, as correct pronunciation is a prerequisite for other parts of the language curriculum, such as forming negations, asking questions, etc., this arrangement clearly had a didactic purpose. As the sequence of chapters shows, almost one half of the Arte is devoted to counting in a broad sense. The incomplete list of EMH numerals is followed by sections on classifiers (see Section 4.2. below) and rather detailed comparisons of weights, measures, chronology, monetary units, etc.2 In other words, the second half of the Arte is essentially an attempt to inventorize words and phrases related to one semantic field. The fact that this sort of thematic word and phrase inventory occurs in a manuscript entitled Arte shows that the boundaries between different genres of linguistic works were at times fluid. As pointed out in Chapter 3, in EMH dictionaries like Chirino’s Dictionarium or the Bocabulario, there are also entries devoted to the analysis of EMH sentence structures and to aspects of Chinese culture. In contrast, a grammatical work like the Arte may likewise contain lengthy sections which would, due to their ‘inventorizing nature’ and lack of overt analysis, fit just as well into a lexicographic work. The lack of overt analysis does not mean that the second part of the Arte lacks analytical dimensions. The analytical task associated with the compilation of conversion lists, for example, of monetary units and units of weight, is of a different nature from that of the analysis of EMH pronunciation and sentence structure in the first part of the Arte. As I will point out presently, reference to traditional Chinese phonology is entirely missing in the sections on pronunciation. Incorporation of native Chinese concepts in the field of morphosyntax can also be ruled out; as the native approaches to morphosyntax were rather rudimentary at the time when the Arte was written.3 Hence, the analytical

–––––––– 2

On the last folio of the list of numerals, EMH transcriptions appear without diacritics and Chinese characters and Arabic numerals are missing. 3 For a comprehensive study on grammatical approaches in premodern China, I refer to Sūn (2005).

92

CHAPTER FOUR

approach in the first part of the Arte does not reflect how the object of analysis, EMH, was conceptualized in native Chinese analyses. The translation of modes of counting into Spanish, by contrast, involves an understanding of how time and amounts are configured in the Chinese system of counting. This, in turn, presupposes precise observation of the way the Sangleys operated with weights. The following quotation by the Franciscan Juan Francisco de San Antonio (1682–1744) gives a good impression of how Spaniards observed unknown weights and established equivalents with their own units. The quotation is cited from his Cronicas which deal with native customs of the Philippines in general, and not specifically with the Sangleys. He writes (1738; BR 40:362): In order to weigh bulkier things, such as wax, silk, meat, etc., they had steelyards, which they called sinantan, which was equivalent to ten cates, of twenty onzas [i.e., ounce—explanation added by BR] apiece. The half of that they called banàl, which was five cates; and the half of the cate they called soco. Consequently, these old weights having been adjusted to the Spanish weights by the regulations of the year 1727, one cate is equivalent to one libra, six onzas; one chinanta to thirteen libras, and twelve onzas; hence one quintal, of eighty of the old cates, corresponds to four arrobas and ten libras of our weight. A pico of one hundred cates is equivalent to five arrobas, twelve and one-half libras, in the new arrangement. As in the case of gold, one tahèl must weigh one and one-fourth onzas in our weight.

The quotation hints at one particular difficulty associated with the translation of the second part of the Arte. As both Chinese and Spanish units of measurement are subject to regional and chronological differences, it is in many cases very difficult, if not impossible, to assess the quality of the conversion provided in the Arte. For details, I refer to the notes on particular units in Part II of this study. I am not aware of another Arte devoting so much space to issues of counting. As I argue below, in light of the enormous significance of Manila as an international trade port, this focus cannot be coincidental. 1.2. Written Representation of Data The way linguistic data are represented in written form has important analytical, methodological, and didactic dimensions. One characteristic of the transcription system used in the Arte is its consistent use of tone diacritics. The same diacritics are used in the Dictionario and the Vocabulario; the LMS, by contrast, scarcely uses

LANGUAGE AND METALANGUAGE

93

tone diacritics. A detailed analysis of EMH tones and tone diacritics is provided in Chapter 5. In both Arte manuscripts, explanations of the properties of tones are rather short, and make no reference to traditional Chinese terminology. It seems hardly conceivable that explanations like ‘higher or with a sharper voice’ (mas alto o con mas agudeza en la voz) or ‘held up in a way that it does not drop’ (sustenido de suerte que no se dexe caer cosa alguna) could convey precisely how the tones were actually pronounced. However, as discussed below, the Arte was chiefly written for teaching purposes, and it can be assumed that all examples were recited by native speakers to be imitated by the learner. Such a procedure must have been practiced, instead of a more detailed description of tones. It may thus be concluded that tones are recognized in the Arte, but that the recognition does not result in a systematic descriptive conceptualization. However, the consistent application of a set of diacritics must be seen as an indication that the contrastive function of tones was at least well understood. The analytical dimensions of the written representation of language data go beyond the indication of tones. Today and in the past, any linguist who writes about Chinese has to decide in which script(s) the data should be represented. The compilers of the LMS and BMS came to different decisions: Apart from a few exceptions, the former uses romanized transcriptions only, whereas the latter uses romanized transcriptions in combination with Chinese characters throughout the whole document. These two modes of data representation—romanization only or a mix of romanization and characters—reflect common practice in modern linguistic literature out of China. The choice of which of the two modes to use may arise from different considerations. Chappell, for instance, argues that, given ‘the relatively high proportion of homophones in Sinitic languages,’ a digraphic representation ‘increases comprehensibility for Chinese-literate readers, and confers cultural authenticity on the data’ (2001a:xiii). Wiedenhof, on the other hand, uses Chinese characters in combination with romanized transcriptions ‘only when an example line is quoted from a written Chinese source’ (1995:8). Today, any linguist opting for a digraphic representation of data benefits from the fact that a horizontal arrangement of characters has replaced the traditional vertical arrangement. In other words, the direction of Western and Chinese writing is the same. As this was not the case in the seventeenth century, the compiler of the Arte had to inte-

94

CHAPTER FOUR

grate two different directions of written representation. This technical problem was reinforced by the fact that the Chinese paper on which the Arte is written had an imprinted grid for writing Chinese characters. This grid separates each folio into two big columns, each of which is subdivided into three columns. Faced with this graphical specification, the compiler found an inventive solution by using the three-column subdivision for the separation of three levels of data representation. As shown in Figure 1, the left column contains EMH expressions in a romanized transcription, followed by Chinese characters in the middle column and a Spanish representation in the right column.

Figure 1: Arte, f. 7r As Figure 1 also shows, the arrangement of characters within one cell of the document grid is horizontal, thus following Western writing traditions. As longer syntactic units do not fit into one cell, the compiler arranged the continuation of interrupted sentences vertically. A vertical line in the left margin indicates that the contents of the cells are read vertically as one syntactic or lexical unit. Compare, for example, the first two rows in Figure 2 with the last three rows.

LANGUAGE AND METALANGUAGE

95

Figure 2: Arte, f. 5r The first two rows do not have a vertical line in the left margin; there is therefore no syntactic link between the items in row one («cheî | 多 | munchos» ‘many’) and row two («cǎc cheî | 可多 | mas munchos» ‘many more’). By contrast, the items in rows 3–5 represent a syntactic unit, the reading direction running vertically within one column. The section reads as follows: Left column: týcn ě: cǎc chêi hò sím Middle column: 天下可多好心 Right column: debaxo el çielo los mas son buenos The sequence in the right column obviously does not qualify as a good Spanish sentence. The Arte contains many more Spanish renderings of this kind which must be judged, from the perspective of Spanish syntax, as unacceptable. The reason for using ‘incorrect’ Spanish in the right column follows methodological considerations. As in Chirino’s dictionary introduced in Chapter 3, the Spanish renderings are obviously not intended as a translation in a strict sense. Instead, they are a functional hybrid, attempting to provide both a direct rendering of the EMH sentence structure and an indication of the meaning of the EMH sentence as a whole. It was the obvious difficulty in combining

96

CHAPTER FOUR

the two functions in one sentence that ultimately led to the functional division of glossing and translating in modern linguistics. My English translations in Part II of the present study attempt to capture both analytical levels of the Spanish columns. Whenever a Spanish sentence structure seems to be modified in order to attain a glossing function, I provide two separate English renderings, the first being a one-to-one representation of the Spanish gloss-translation, the second being an acceptable English sentence rendering the meaning of the EMH sentence. Both English renderings are separated by a rightpointing triangle. Thus, the sentence in Figure 2 is translated as follows. under heaven the most are kind  Most men under heaven are kind. A detailed look at the various examples of vertical sentence arrangement shows that the compiler was keenly aware of the comparative function offered by the document grid. In most cases, there is a precise match of the contents of the cells in one row. Thus, the romanized transcriptions in the left column exactly match the characters in the middle column, and the right cell provides Spanish equivalents of the two EMH cells. This arrangement shows that the Arte contains more detailed syntactic analysis than the short explanatory sections suggest. This analysis, which can be regarded as a rudimentary form of Chinese-Spanish contrastive syntax, is implicit in the layout of the example sentences. It can be assumed that the comparative function of the grid was born out of a specific classroom situation. As pointed out above, the representation of one Chinese example sentence in two scripts may arise from different considerations. In the case of the Arte, the digraphic arrangement obviously had didactic functions. On f. 2v we find the explicit instruction that the learner should be ‘given a Sangley for exercise.’ Without doubt, the Chinese tutor would have been more familiar with Chinese characters than with the Roman alphabet. It also seems that the Spanish learner received instruction in reading and writing the Roman transcription and not Chinese characters. This assumption is supported by the fact that most religious documents, like the Doctrina Christiana mentioned in Chapter 2, use transcription only. Moreover, the Arte does not contain a section on Chinese characters or character writing, which would suggest that character writing

LANGUAGE AND METALANGUAGE

97

was not part of the curriculum. We may therefore imagine a classroom situation with a Chinese tutor reading the middle column and Spanish student(s) linking the perceived pronunciation with the romanized transcriptions and the Spanish gloss-translations. 2. Avoiding Terminology This explanatory device refers to instances in which a syntactic pattern is recognized and described, but no terminological innovation or adaptation derives from the analysis. For example, f. 12r has the following section: The particle tioh8 means that one succeeded in an action, liah8 tioh8 lang5 ‘succeeding in catching the one to catch.’ A negation placed between the tioh8 and the verb says one has failed and not caught the one to catch, for example: liah8 tioh8 lang5 扐着人 he caught and he succeeded with the one he had to catch  ‘He succeeded in catching the one he had to catch.’ liah8 m3 tioh8 lang5 扐不着人 ‘He failed in catching the one he had to catch.’

The Arte does not use a term for the ‘particle tioh8’ that is referred to as ‘resultative verb suffix’ in Bodman’s Amoy Hokkien textbook. Interestingly, Bodman’s explanation largely resembles that in the Arte. He writes (1987:326; original emphasis): Many other verbs are followed by tioh8 which may be roughly translated as ‘accomplish, do successfully’. These are but two types of the large class of forms which are compounded with verbs and modify the verbal meanings in a particular way so that the emphasis is on the result of the action.

Bodman does not offer an explanation for the negation of tioh8 with m7. In the textbook of the Taiwanese Maryknoll series, the following explanation can be found: ‘m3-tioh8, frequently used after a verb to modify a noun indicates that in the performance of the action an error was made in respect to the noun’ (Maryknoll, 1990:113). Taiwanese examples include (ibid.; transcription modernized, my gloss, original translation):

98

CHAPTER FOUR

Gua2 ciah8 m7-tioh8 ioh8-a2. 1.sg eat not-correct medicine ‘I took the wrong medicine.’ ue7. Gua2 kong2 m7-tioh8 1.sg speak not-correct speech ‘I said the wrong thing.’ On the basis of this explanation, the EMH sentence liah8 m3 tioh8 lang5 can be understood as describing a situation in which a person has been caught, but it turns out that it was the wrong person. The analysis in the Arte goes beyond the introduction of correct sentence structures. Some sections also include statements about the semantic and stylistic connotations of syntactic variation. One example can be found on f. 14r: As for verbs of giving and taking, what is given or what is taken can be placed before or after the verb, although it is more elegant to place it before the verb. lu2 thou7 pe3-lo1 ci3 no2 ge2 cinn5 汝度 呶只二个銭 ‘You give Pedro these two coins.’ ci3 no2 ge2 cinn5 lu2 thou7 pe3-lo1 只二个銭汝度 呶 ‘These two coins—you give them to Pedro.’

In modern Hokkien and other Sinitic varieties, this sentence pattern is very common, and the use of a topicalized object preceding a clause is not restricted to verbs of giving. From the description in the Arte it is not possible to judge whether this pattern also occurs with other EMH verbs. It is interesting to note, however, that the topicalized object in the second example sentence is judged as being more elegant. Again, the construction with a topicalized object preceding the clause is recognized, albeit without terminological specification. 3. Questioning of Existing Terminology The chapters on declension and conjugation seem to exemplify the inappropriate application of the Greco-Latin model for the analysis of an isolating language. However, a closer look at the explanatory sections

LANGUAGE AND METALANGUAGE

99

preceding the paradigms shows that the author swiftly circumvented the analytical dilemma by frankly denying the relevance of parts of his model. The opening sentences of these two chapters briefly capture what Hokkien does not have with regard to the paradigm under discussion. In Chapter 2 of the Arte, the statement that ‘this language’ does not have declensions is immediately followed by a brief inventory of grammatical particles distinguishing cases (ff. 2v–3r): This language does not have different nouns and terms for distinguishing cases. But there are particles which distinguish cases. The genitive is distinguished by a postponed particle ge2. The dative is preceded by the particle khit4; the ablative by one of the particles kang7, kap4, tang5. The nominative and the accusative and the vocative do not have particles. The nominative stands for persons who do something and the accusative for persons who suffer.

In Chapter 3, the applicability of the European model is likewise denied in the very first sentence. Interestingly, however, although the author states that ‘this language’ does not have different verb endings, a conjugation table for the verb si3 ‘be’ follows directly after the statement: On the verb sun est fui in this language. This particle si3 是 be is sun and est and fui. It has no more endings and neither distinguishes persons or moods unless it occurs with particles mentioned later. gua2 si3 ho2-sim1 我是好心 ‘I am kind.’ lu2 si3 ho2-sim1 汝是好心 ‘You are kind.’ i1 si3 ho2-sim1 伊是好心 ‘He is kind.’ guan2 si3 ho2-sim1 阮是好心 ‘We are kind.’ lun3 si3 ho2-sim1 恁是好心 ‘You are kind.’ in1 si3 ho2-sim1 因是好心 ‘They are kind.’

The claim that Greco-Latin paradigms are unsuitable for languages with an isolating morphology is obviously true and needs no further elaboration. This is, however, not the crucial point. Instead, we should ask why the compiler employed Greco-Latin paradigms despite their obvious inappropriateness. Here the didactic dimension of missionary grammar writing must again be taken into consideration. As explained before, the Arte was not written for Sangleys, but for newly arriving missionaries who had to learn Hokkien from friars and Sangleys in situ. From a didactic perspective, the compiler had to address the intellectual background of his students. As the linguistic mindset of any missionary arriving in Asia during the seventeenth was shaped by the

100

CHAPTER FOUR

Greco-Latin grammar model, it was a convenient didactic device to address this particular background. In the context of the explanation in the first sentence, the table should thus not be seen as an attempt to enforce an inappropriate grammar model on EMH, but rather as an exemplification of missing verb endings. Thus, by denying the applicability of conjugation and declination paradigms, the compiler questioned the relevance of his model as the sentences introducing declensions and conjugations clearly mention the absence of these paradigms from Hokkien grammar. For example, the Arte introduces Hokkien ‘declensions’ by stating correctly that ‘this language does not have different nouns and endings for distinguishing cases.’ In other words, the conjugation and declension paradigms are used as a negative example, so to speak. From the perspective of the missionary language teacher, the introductory sentence reads like an advice to the student to dismiss from his mind his European ideas about language structure. Until today, this process of unlearning the familiar remains a challenge to both language learners and linguistic fieldworkers. According to Gil (2001:104f.): [W]hen confronted with a new language, it is sometimes easier to recognize the presence of exotic, unexpected, and hitherto-unknown items than to come to grips with the absence of familiar, commonplace, and presumed-to-be universal entities. Our native language imposes a straitjacket from which it is often difficult to break free, in order to realize that certain grammatical categories, obligatory in our own language, may be absent in the language under investigation. Moreover, if our native language is European, then this straitjacket is likely to be reinforced by the weight of Eurocentric linguistic traditions, which either implicitly presuppose or else explicitly assert that certain grammatical categories are universal. Thus fieldwork involves not only the learning of new items, but also the unlearning of old and familiar ones.

4. Terminological Innovation 4.1. Modes of Pronouncing Tones are represented as two of the so-called ‘modes of pronouncing’ (modo de pronunçiar) Hokkien. In the BMS, the term ‘mode’ (modo) occurs in the heading and in the first paragraph of Chapter 1. In the subsequent paragraphs, each mode is referred to as ‘difference’ (difirençia). The LMS uses ‘mode’ in each subheading preceding a particular mode of pronunciation. A comparison of both manuscripts shows that the category ‘modes of pronouncing’ does not refer to a

LANGUAGE AND METALANGUAGE

101

fixed analytic framework for the analysis of foreign speech sounds. The modes and their sequences of presentation in both manuscripts are listed in Table 1. Table 1: ‘Modes of pronouncing’ in the BMS and LMS BMS 1 2 3 4 5 6 7 ----

LMS 1 2 5 3 7 9 -4 6 8

8

10

Feature non-entering tones vowel ee aspirated initial entering tones non-aspirated initial, open nasal final aspirated initial, open nasal final velar nasal final, no medial vowel aspirated initial, entering tone aspirated initial, vowel ee, glottal stop -h unaspirated initial, no medial vowel, velar nasal final aspirated initial, no medial vowel, velar nasal final

As shown in the table, numbers for pronunciation modes do not correspond to the same phonological properties. The LMS to a greater extent summarizes combinations of features as individual modes (in 4, 6, 8). There is a certain redundancy in this approach, as the features represented by mode 4 (aspirated initial plus entering tone) are also covered by modes 3 and 5. It can therefore be argued that the BMS is more refined in its analytic approach, as its distinction of modes is less redundant than that of the LMS. The individual features described in both manuscripts, however, are identical, viz. tone, aspiration, vowel ee, open nasal final, and velar nasal final without preceding vowel. Obviously, these are all features which do not exist in Spanish phonology. Hence, the distinction of ‘modes of pronouncing’ is by no means an attempt to address EMH pronunciation in its completeness. Instead, ‘modes’ only include unfamiliar sounds. In other words, the notion ‘modes of pronunciation’ as such is designed to address the unfamiliar. I have thus far not found a similar use of the category ‘mode of pronouncing’ in another missionary grammar. The Arte explicitly introduces these modes in terms of contrastivity by writing that the same Chinese word can have different meanings only distinguished by

102

CHAPTER FOUR

different modes of pronunciation: ‘high or low, with or without aspiration, nasal, etc.’ (f. 1r). As pointed out, tones are represented by two modes. It is noteworthy that the five tone diacritics appear at the very beginning of the first chapter, which suggests that the compiler attributes special importance to them. Also, at the end of the first chapter, he stresses the importance of learning tones to perfection. In this context it is, however, not entirely clear whether the term ‘tones’ refers to tone in the modern linguistic sense or whether it also includes non-tonal features listed as modes of pronunciation. Irrespective of this detail, in the broader context of missionary linguistics, it is remarkable that a clear understanding of tone is manifest in a document of 1620. In Chapter 3 I have argued that EMH sources which were very likely written before the Arte do not employ a consistent system of tone notation, although it is obvious that documents of the period 1605–1621 reflect a growing awareness of the importance of tones. This leads to the question of whether the understanding of the linguistic feature ‘tone’ was a genuine discovery of the Manilabased missionaries or whether it was inspired by some kind of external influence. In the following sections I argue that the conceptualization of tones in the Arte must have been influenced by the work of the China-based Jesuits, although the exact nature of this influence has thus far not been documented. Interestingly, in New Spain, where missionary linguistic research had started much earlier than in Asia, missionaries were rather indifferent towards tones. As Smith-Stark writes (2005:24): Of all the phonological characteristics of the languages of New Spain, tone, which is particularly common in the languages of the Otomanguean family, was the one that caused the most difficulty for the early linguists. In many cases, the tonal nature of a language was not really recognized until the 20th century.

This is in obvious contrast to the situation in Asia. Bossong suggests that the first description of the ‘phonemic role of tone in Far Eastern languages’ is attested in a description of Vietnam of 1633, written by the Jesuit Cristoforo Borri (1583–1632) (2007:129–130; on Borri, see also Dror and Taylor, 2006). Referring to Jacques (2002) Bossong also points out that Francisco de Pina (seventeenth century, dates unknown) mentions toadas ‘tonalities’ in Vietnamese as early as 1622 or 1623 (ibid.).

LANGUAGE AND METALANGUAGE

103

Although possible links between missionary linguistics in Vietnam and the Philippines are yet to be investigated, I assume that these Vietnamese sources are irrelevant for the Arte. Anyone familiar with missionary linguistic sources of the seventeenth century will immediately recognize that the set of tone diacritics introduced in Chapter 5 was exactly the same set that was used by the China-based Jesuits at the same time. The same tone diacritics occur in various sources, for instance in the famous Xīrú ěrmù zī 西儒耳目資 ‘Aid to the Ear and Eye of Western Scholars’ of the Jesuit Nicolas Trigault (1626) (see Chapter 2), Varo’s Mandarin grammar (1703/2000), Martini’s grammar (1689), and Prospero Intorcetta’s (1626–1696) treatise on the Chinese script (ca. 1660, facsimile in Lundbæk, 1988).4 In the first version of their bilingual Portuguese-Chinese Dictionary (compiled in the 1580s, see Chapter 3), Ricci and Ruggieri neither provide tone descriptions nor use tone diacritics (Yang, 1989:218). In a later version, however, which Ricci supposedly compiled together with Lazzaro Cattaneo (Guō Jūjìng 郭居靜, 1560–1640) in 1598–1599, tone diacritics are used for the first time. Brockey points out that it was very likely Cattaneo and not Ricci himself who laid the foundations for the systematic notation of tones (2007:249f.; see also Takata, 2004). The documentation of Cattaneo’s work, which apparently made use of European methods of musical notation, awaits detailed analysis.5 With regard to China, Smith-Stark conjectures that ‘it seems reasonable to suppose that traditional Chinese language studies may have provided an important stimulus which made these early and sophisticated analyses of tone systems possible in Asia’ (2005:24, fn. 39). This is certainly true in the case of the Jesuit linguistic tradition initiated in China. As regards EMH documentation, however, the intellectual encounter between traditional Chinese philology and the European missionary linguist was less intense. This certainly has to do with

–––––––– 4

The manuscript does not contain information about authorship. My attribution of authorship to Intorcetta follows Lundbæk (1988:42). 5 This does not imply that the Jesuits used a fully uniform system of romanization once tone diacritics had been introduced. Trigault, in his Xīrú ěrmù zī, uses a revised version of Ricci’s transcription system. This is referred to as the ‘Ricci-Trigault System of Transcription’ by Chiao and Kriegeskorte (2000:996). It is, however, known that prior to Trigault, in addition to Cattaneo, the missionary Diego de Pantoja (Páng Díwǒ 龐迪我, 1571–1618) and the Chinese convert Zhōng Míngrén 鍾鳴仁 (Sébastien Fernandes Tchong, 1562–1621) were also involved in improving the first version of Ricci’s system (cf. Vande Walle, 2001:865–868).

104

CHAPTER FOUR

the social contexts of missionary work. As pointed out in Chapter 2, Manila-based missionaries had much less social intercourse with Chinese scholars than the Jesuits in China. This can certainly explain why the Jesuits in China were more receptive to traditional Chinese philology. The Arte refers to an unidentified Chinese vocabulary (f. 1r; see also Chapter 5). Although this reference to a previous work suggests some kind of influence, the complete lack of reference to phonological terms of the Chinese tradition indicates that the missionaries in the Philippines were to a greater extent exposed to the spoken language of their tutors than to the written characters in Chinese books. How, then, exactly, did the Manila-based missionaries come to use a set of diacritics for the tones they heard in EMH speech? Thus far, no research has been done that examines the possible exchange of linguistic concepts across the geographical and congregational boundaries separating missions in China and the Philippines. There can, however, be no doubt that some kind of exchange must have taken place, as the use of an identical set of tone diacritics cannot be coincidental. I assume that the missionaries in the Philippines received the set of tone diacritics used in the Arte, the Dictionario, and the Vocabulario from the Jesuits in China, and not vice-versa. As pointed out, there is solid historical evidence that in China, the tone diacritics were used in the late sixteenth century already. EMH sources written at the same time, most notably Chirino’s Dictionarium (1605), do not use these diacritics. 4.2. Classifiers as ‘Proper Numerals’ One widespread typological feature of Hokkien and other Asian languages is the use of so-called measure words or classifiers. In modern Mandarin, for example, classifiers occur between a numeral or a determinative and a noun. In the phrase yì běn shū ‘one book’ (yì ‘one’, běn ‘volume’, shū ‘book’), běn ‘volume’ is the classifier. In one of the untitled final sections of the Arte there is a surprisingly long list of 79 EMH classifiers (ff. 23r–30r). The explanatory introduction to this list runs as follows: In this language there are, apart from common numerals, other proper numerals for counting particular things. These are placed between the common numeral and the noun. For example, for ‘one snake’ you say cik8 bue2 cua5. Cik8 is the common numeral, bue2 the proper [numeral] and cua5 ‘snake’, as you will see in the following examples.

LANGUAGE AND METALANGUAGE

105

2

(1) bue 尾: for things with a tail si2 bue2 cua5 四尾蛇 ‘four snakes’ si2 bue2 gok4-hu5 四尾鱷魚 ‘four crocodiles’ si2 bue2 cui2-be2 四尾水馬 ‘four seahorses’

Here the compiler shows some understanding of the fact that the use of the same classifier for different nouns can be explained with a feature shared by the nouns. Instead of just listing snakes, crocodiles, and seahorses as nouns used with the classifier bue2 (literally ‘tail’), the ‘tail’ is identified as the common feature of nouns sharing one classifier. To be sure, given the literal meaning ‘tail’ of the classifier bue2, this particular example does not require exceptional analytical talent. The remaining examples list classifier-noun combinations without analyzing common features, as in example 46 (f. 26v–27r): (46) tionn1 張: for sheets of paper, for musical instruments, for beds and tables si2 tionn1 khim5 四張琴 ‘four lutes’ si2 tionn1 phio7 四張票 ‘four documents or licenses’ si2 tionn1 phei1 四張批 ‘four letters’ si2 tionn1 cua2 四張帋 ‘four sheets of paper’ si2 tionn1 ji5 四張字 ‘four written sheets’ si2 tionn1 an3 四張案 ‘four lawsuits’ si2 tionn1 toh4 四張槕 ‘four tables’ si2 tionn1 chng5 四張床 ‘four beds’

106

CHAPTER FOUR

The sheer number of examples indicates that classifiers had been correctly recognized as a characteristic EMH feature. This recognition resulted in a terminological distinction between common numerals and proper numerals. However, as common features of nouns sharing a classifier were not identified systematically, the recognition of classifiers lacked analytic depth. Interestingly, whereas Varo’s Mandarin grammar of 1703 labels classifiers as ‘particles’ (particulas), the terminology in the Mandarin grammar of the Protestant missionary Robert Morrison (Mǎ Lǐxùn 馬禮遜, 1782–1834), written in 1815, is reminiscent of the Arte. Morrison writes that ‘it is proper to notice a class of words called Numerals, which generally precede or follow the Noun. From their name, it is apparent, that they are used in numbering’ (1815:37). As the China-based Jesuits paid little attention to sentence structure in the first half of the seventeenth century, it seems likely that the analysis of classifiers originated in EMH documentation. One particular fact which can be adduced to support this assumption is the importance of trade in the Spanish-Chinese encounter. As the exchange of goods between Sangleys and Spaniards was an essential feature of everyday life, the Spaniards were required to have a level of linguistic competence that enabled them to conduct this trade. As the counting of goods is an essential feature of any direct trade interaction, an analysis of the linguistic dimension of counting was an obvious thing to do. In this respect it is important to point out that most of the examples in the list of classifiers involve goods that were traded in Manila, such as various kinds of fabrics, sugar, rice, and precious metals. D. CONSTRUCTED LANGUAGE? The previous sections have shown that explanatory devices represented in the Arte go beyond the application of linguistic terms. However, if the interpretation of historical language data relies solely on the analysis of metalanguage, then epistemological pitfalls cannot be avoided. It is, for example, known that missionaries manipulated data in order to facilitate linguistic standardization. Another possible reason for manipulation is that data had to be trimmed in order to fit the theoretical linguistic paradigms. For example, in some cases, one cannot escape the impression that the compiler of the Arte desperately tried to satisfy his desire for

LANGUAGE AND METALANGUAGE

107

structural parallels between Hokkien and his native Spanish language. Take, for example, the section on job and office titles (f. 5r). In Spanish, words for professions and offices are suffixed by -ero. In Hokkien, according to the Arte, job and office titles are expressed by nominal phrases headed by ‘person’, preceded by the subordinative particle ge2 (e5 in modern Hokkien dialects) and the subordination specifying the profession, for example ‘sell fish’. Thus, in both Hokkien and Spanish, the element expressing ‘the person who’ follows the element specifying the office. Although this structural parallel is not explained explicitly, it is nicely highlighted by the arrangement of the examples, as shown below (f. 5r): bêi hū guè lāng

賣魚 個人

el pesca dero

pà'c gūin guè lāng

打艮 個人

el pla tero

chô chìo' gùe lāng

作石 個人

cante ro

pà'c gūin sáy hû

打艮 師阜

plate ro

In names of crafts or names often referring to a craft with verbs meaning ‘make’, the verb is placed at the beginning, followed by the thing denoting the craft, the third [element] is the particle ge2 and the last lang5. bei7 hu5 ge2 lang5 賣魚個人 ‘fishmonger’ phah8 gin5ge2 lang5 打艮個人 ‘silversmith’ co7 cioh8 ge2 lang5 作石個人 ‘stonecutter’ These nouns are also formed with sai1-hu7 at the end, which is ‘workman’. It replaces lang5, for example: phah8 gin5 sai1-hu7 打艮師阜 ‘stonecutter’

As stated above, the compiler of the Arte seemed to be well aware of the typological similarities between the Hokkien and Spanish expressions. The juxtaposition of forms indicating ‘the person who’ in the presentation of job and office titles is one example. Obviously, his interest in structural parallels shaped the way he presented his data. However, in the context of the section quoted above it is important to point out that the EMH forms are not attested in the Dictionario,

108

CHAPTER FOUR

which has phah8 cioh8 lang5 ‘stonecutter’, tho2 hu5 lang5 ‘fishmonger’, and phah8 gin5 ‘silversmith’. Another example of possibly constructed language is the second person plural pronoun, transcribed as in the Arte. The compiler explains the in in terms of a quasi-plural suffix (f. 3r): The plural of the three primitive pronouns ego, tu, ille is different from the singular. It is formed by adding n to the singular and is declined with the aforementioned particles. Singular lu2 汝 ‘you’ Plural lun3 恁 ‘you’ Singular i1 伊 ‘he’ Plural in1 因 ‘they’

It is difficult to judge whether the search for European-looking language structures in Hokkien is behind this paradigm. As will be pointed out in Chapter 6, I have not seen any Hokkien data in which the final of the second person singular is -u and the medial of the second person plural is -u-. We are thus left to assume that the compiler may indeed have ‘manipulated’ his data in order to make it look more European. Another problematic instance of data presentation and analysis can be found in the treatment of the interrogatives si3-cui7 ‘who’ and si3mih8 ‘what’.6 In all example sentences of the Arte, the first element of the two interrogatives is si3. Some examples are the following: si3-cui7 lai5 是誰來 ‘who is coming’ (f. 3v) si3-cui7 ge2 是誰個 ‘of whom’ (f. 4r) lu2 phah8 si3-cui7 汝打是誰 ‘Whom do you hit?’ (f. 4r) si3-cui7 lai5 是誰來 ‘Who came?’ (f. 4r) si3-cui7 khu3 是誰去 ‘Who is going?’ (f. 4r) si3-mih8 lang5 是乜人 ‘What person?’ (f. 4r) si3-mih8 geh8 是乜月 ‘What month?’ (f. 4r) si3-mih8 pit8 是乜筆 ‘What quill?’ (f. 4r)

–––––––– 6

Note that the tone of si3 is noted as si7 in other studies on Hokkien. The possible reasons for this particular tone notation will be discussed in Chapter 5.

LANGUAGE AND METALANGUAGE

109

The same element si3 是 is introduced as a copula ‘be’ in the first sections of Chapter 3 of the Arte (ff. 5v–6r). Examples for si3 是 ‘be’ include the following: lu2 si3 ho2-sim1 汝是好心 ‘You are kind.’ (f. 5v) tio5-si3 ciah4 si3 gau5 僚氏即是賢 ‘God is very wise.’ (f. 6r) The use of the same element si3 是 as part of interrogatives ‘who’ and ‘what’ and as copular ‘be’ may either be explained as a case of homonymy or polysemy. If we treat it as homonymy, the formal congruence is coincidental and no further analysis would be required. As studies by Lien (2006, 2009) have pointed out, however, there are good reasons for a polysemic analysis of si3 in early Hokkien. Lien’s analyses are based on data in printed versions of stage plays in local Hokkien vernaculars (Wú, 2001a–d, 2002a, b) and they therefore offer the advantage of allowing us to compare EMH examples selected from missionary sources with those selected from contemporary sources not associated with the missionary tradition. Lien (2009) points out that si7 ‘be’ is also used as a focus marker, as in the following example (example, glosses, and translation are cited from Lien): Kiann3 si7 li2 phah4 phua3 鏡是你打破 mirror FM you strike broken (FM = focus marker) ‘It is you who broke the mirror.’ If we accept the interpretation of si7 as copular ‘be’ > focus marker, Lien’s following interpretation is plausible (Lien, 2009:752): The sequence si7 mih8 是乜 is open to two interpretations: (1) si7 是 ‘be’ and mih8 乜 ‘what’ are separate constituents forming a construction, or (2) si7 mih8 是乜 as a whole denotes ‘what’ and the two elements have been reanalyzed as a grammatical function word, viz., what whword, where the phonetic shape si7 是 has lost its meaning as well as its independent status.

One of Lien’s conclusions is that in the late sixteenth century, ‘some instances of si7-mih8 had already begun to show signs of lexicalization’ (2009:760). The claim that the feature ‘focus marker’ is lost in many examples of si7 mih8 ‘what’ is supported by the fact that the

110

CHAPTER FOUR

mih8 ‘what’ without preceding si7 occurs less frequently than si7-mih8 in the data examined by Lien (2006:803). Whereas the treatment of si3-mih8 as an ‘unfocused’ interrogative ‘what’ in the Arte is therefore supported by Hokkien data in other sources, the case of si3-cui7 is different. In contrast to mih8 ‘what’, cui7 ‘who’ without preceding si3 is a very productive interrogative in the Lìjìng jì (see Chapter 1), where it occurs much more often than si3-cui7 ‘who’ (Lien, 2006:803). We can therefore assume that si3 in si3-cui7 was to a much lesser degree, if at all, lexicalized and that it hence still retained its function as a focus marker or simply as a copula. On the basis of this assumption, I propose the following reinterpretations of the sentences on p. 108 (marked by ): si3-cui7 ge2 是誰個 ‘of whom’ (f. 4r)  ‘Whose is it?’ lu2 phah8 si3-cui7 汝打是誰 ‘Whom do you hit?’ (f. 4r)  ‘Who is it that you hit?’ si3-cui7 lai5 是誰來 ‘Who came?’ (f. 4r)  ‘Who is it that came?’ si3-cui7 khu3 是誰去 ‘Who is going?’ (f. 4r)  ‘Who is it that is going?’ It seems unlikely that the frequently occurring interrogative cui7 ‘who’ would have been simply missed by the Spanish friars. I suppose that the fact that only si3-cui7 is introduced in the Arte is attributable to its structural parallel with si3-mih8. After all, any kind of structural parallelism in the target language fits better to an approach that looks for rules and regularities. E. SUMMARY AND CONCLUDING REMARKS A recurrent theme in studies on missionary linguistics is the conceptualization of ‘unfamiliar’ linguistic features. As shown in this chapter, the Arte contains examples of different analytical responses to the unfamiliar. It is argued that the use of linguistic terms associated with the work of the Spanish grammarian Nebrija is not per se indicative of a direct influence. Instead, the structure, the style, and the terminology

LANGUAGE AND METALANGUAGE

111

of the Arte were presumably influenced by linguistic treatises of other missionaries, notably those based in the Americas during the sixteenth and early seventeenth century. Some of the terms used in the Arte, such as ‘mode of pronouncing’ and the distinction between ‘common numerals’ and ‘proper numerals’ specifically address EMH features which are not recognized in the Greco-Latin model of grammar. The former attempts to capture features in EMH pronunciation which are foreign to Spanish phonology, such as tones and aspiration; the latter distinguishes numerals and measure words. The explanatory devices employed in the Arte go beyond the use of linguistic terminology. The consistent application of tone diacritics reflects the compiler’s understanding of tone contrasts; the arrangement of example sentences is closely linked to syntactic analysis. Some features, such as resultative verbs, are well described without the use of specific terms. These explanatory devices point to different sources of influence. Whereas the possibility that tones were recognized on the basis of Chinese rhyme books cannot be ruled out, an influence of earlier studies by the China-based Jesuits seems more likely. From the broader perspective of linguistic historiography it is important to emphasize the significance of the discovery of tones by missionaries in China and the Philippines in the first decades of the seventeenth century. This discovery must be considered a milestone in the research on tone languages in the history of European linguistics. In short, the Arte by no means conceals linguistic data by the use an inappropriate model. Instead, the compiler uses the model rather flexibly by denying its relevance in certain parts and by treating linguistic features foreign to Indo-European languages. The examination of missionary metalanguage and its object of analysis in this chapter is restricted to the Arte. Further conclusions can only be drawn from additional study of currently available material, or from the discovery and study of as yet undiscovered material. The search for these is among the exciting tasks in the study of Chinese missionary linguistics.

CHAPTER FIVE

PHONOLOGY AND ORTHOGRAPHY A. INTRODUCTION This chapter analyzes the romanization system used to spell EMH in the Arte. The purpose of the analysis is threefold. First, it aims at providing an overview of EMH phonology, i.e., an inventory of tones, consonants, and vowels. Second, on the basis of a comparison with spelling conventions in other EMH sources, it will document orthographic diversity in EMH romanization during the seventeenth century. Third, it introduces the conventions used for the retranscription of EMH expressions in the present study. These conventions are largely based on the romanization system known as Taiwan Language Phonetic Alphabet (TLPA), which is especially convenient due to its ability to render quite a large number of regional Hokkien varieties and the lack of diacritics. My use of the TLPA system is not a one-toone conversion of the letters and diacritics used in the Spanish system. Instead, it reflects assumptions concerning EMH phonology to be explained presently. Methodologically, my analysis will draw on two areas of previous research: Spanish missionary orthography studies and Hokkien phonology. Not only do I expand on both these areas, but, crucially, I will examine the interplay between the two areas. This will reveal aspects which have hitherto not been considered in earlier research. My examination of previous studies on missionary orthographies will not be restricted to studies on specific languages, and I will consider both studies on Sinitic languages, e.g., research on Francisco Varo (1703/ 2000), and non-Sinitic languages. Comparisons with spelling conventions in other missionary sources of the Spanish tradition will help us to determine possible sound-letter correspondences in the Arte system. The examination of previous research on synchronic and diachronic Hokkien phonology, in turn, helps to assess the validity of conclusions drawn from the analysis of the orthography. For example, f. 12v of the Arte has the EMH expression for ‘Holy Spirit’. An analysis of the orthography alone would probably lead to the conclu-

PHONOLOGY AND ORTHOGRAPHY

113

sion that EMH had a syllable initial consonant cluster sp-. Yet research on Hokkien phonology tells us that this assumption is certainly wrong. We may thus conclude that was a loan from the Spanish Espíritu Santo which was strongly hispanicized when pronounced in EMH speech. Thus far, the only study which addresses issues related to EMH phonology and orthography is van der Loon (1967:108–116). His analysis is based on the Bocabulario (see Chapter 3) and on relatively short extracts from religious sources. The sections which follow therefore do not aim at proposing an entirely new analysis of EMH phonology and orthography. Instead, I will test van der Loon’s hypotheses against EMH data not considered by him, notably the BMS of the Arte, Chirino’s Dictionarium (RMS, see Chapter 3), the Dictionario, and the Vocabulario (see Chapter 3). There are two reasons for doing this. First, the orthographic comparison will show that romanization systems used in different sources were by no means uniform. Second, a comparison of different spellings will allow for more precise claims concerning some phonological details. My consultation of the Dictionario and the Vocabulario has been restricted to the search for examples illustrating contrasts in EMH. Unlike van der Loon, I do not intend to propose a complete inventory of EMH syllable finals. As such an undertaking would require a complete examination of the data recorded in the Dictionario and the Vocabulario, it is beyond the scope of the present study. B. TONE DIACRITICS AND TONES 1. Tone Diacritics One striking feature of the transcription system used in the Arte is its extensive use of diacritics. No less than nine diacritics are used by themselves, and four diacritics occur only in combination with another diacritic, giving a total of thirteen different units. As discussed in the following sections, five diacritics and two combined diacritics indicate tones, one diacritic contrasts aspirated and non-aspirated initials, three diacritics pertain to syllable final properties, and two combined diacritics represent properties of syllable initials and finals in combination. My treatment of combined diacritics as units follows the Arte,

114

CHAPTER FIVE

which categorizes the properties of these combinations as distinct ‘modes of pronouncing’ (see Chapter 4). This section analyzes tone diacritics only. Other diacritics will be introduced in connection with the consonants and vowels they modify. In total, seven diacritics represent EMH tones. They are used both by themselves and in combination. Terminologically, the Arte simply refers to them as a ‘line’ (uirgula). In the BMS, marks on entering tones are labeled ‘straight stroke’ (raya derecha); but this term is not used in the LMS. Further terminological distinctions are not made in either manuscript. In this study, the handwritten tone diacritics are represented by the acute accent , macron , grave accent , háček , circumflex , and vertical stroke vertical stroke ‘slide’ (row six in Table 1) represents the lower register, the second,

and

‘venerate’ ‘permit’ ‘handsome’ ‘bone’ ‘ship’ ‘when’ ‘slide’

TLPA cun1 cun2 cun3 kut4 cun5 cun7 kut8

C. CONSONANTS 1. EMH Consonant Spellings and Hokkien Consonants The transcription system used in the Arte makes use of seventeen different consonants and digraphs, viz. . One diacritic, the spiritus asper , is used for the indication of aspiration. From previous research on Hokkien phonology, we may expect an inventory of thirteen or fourteen consonant phonemes. In the early nineteenth century, Quánzhōu presumably had fourteen consonant phonemes (cf. Ang, 1996). One hundred years later, Tung attributes phonemic status to thirteen consonants (1959: 792); the same number is reported by Lín for modern Quánzhōu (2008: 34). Fourteen consonants have been reported for Zhāngzhōu of the nineteenth and twentieth centuries (Tung, 1959:849–850; Klöter, 2006). Chén (2007:9) lists seventeen consonants for modern Zhāngzhōu. However, his inventory does not consider [m, n, ŋ] to be allo-

131

PHONOLOGY AND ORTHOGRAPHY

phones of /b, l, g/: the former occurring before nasalized vowels, the latter before oral vowels. Thus, the number of phonemes can in effect be reduced to fourteen as well. An inventory of fourteen Hokkien consonant phonemes is provided in Table 13. In many Hokkien dialects, /d͡z/ has merged with /l/, yielding thirteen consonants. Table 13: Hokkien consonants voiced plosive unaspirated voiceless plosive aspirated voiceless plosive unaspirated voiceless affricate aspirated voiceless affricate unaspirated voiced affricate voiceless fricative voiced lateral

Bilabial /b/

Dental

Alveolar

Velar /g/

/p/

/t/

/k/

/ph/

/th/

/kh/

Glottal

/t͡s/ /t͡sh/ /d͡z/ /s/

/h/

/l/

The following sections analyze sound-letter correspondences in the Arte. Although the main purpose is the decoding of the Arte transcription system, orthographic variation across different EMH sources will also be touched upon. My comparison is between the Arte and the four dictionaries introduced in Chapter 3. Retranscriptions in TLPA are also listed. In the tables below, the sources are marked by Roman numerals, as follows:

132

CHAPTER FIVE

I II III IV V VI

Arte Chirino, Dictionarium Bocabulario Dictionario español-chino Vocabulario de la lengua española-china Retranscription in TLPA

As the Arte is of primary relevance for the present study, it is listed first. The remaining sources are listed according to the assumed chronology of compilation, starting with Chirino’s dictionary of 1605. 2. Initial and Final Consonants All letters and digraphs for consonants occur in the initial position. The following consonants also occur in the final position: . The letters and

can be interpreted as representing the unreleased bilabial stop [p¬] which is written as -p in TLPA. In most cases,

is used, except for ‘with’. The fact that

may also occur after a medial [a] indicates that the use of vs.

in final position does not follow an obvious principle. The same goes for the distribution of the letters and , which both stand for an unreleased velar plosive [k¬] (TLPA: -k). In the Arte, may occur after a number of different vowels (including [a]), whereas occurs only after [a], e.g., in ‘strength’. I concur with van der Loon and assume that both and stand for an unreleased alveolar plosive [t̚] (TLPA: -t) and that the orthographic alteration is phonemically insignificant. The letters , , and will be analyzed in later sections. 2.1. Initial and Final As shown in Table 14, initial is used consistently in all EMH sources.3 The letter represents the voiced bilabial plosive [b] (van der Loon, 1967:113). Following the TLPA system, I transcribe this sound as b-. The use of final has been discussed in C-2 above.

––––––––

3 Tables 14 through to 29 cite the entries of the dictionaries. Thus, if the original dictionary entry does not indicate a character, no character is shown in the table. The original Spanish translations are listed below each table for reference.

133

PHONOLOGY AND ORTHOGRAPHY

Table 14: Initial 1. 2. 3.

I bâ' 肉 ‘meat’ bè 馬 ‘horse’ bèi 買 ‘buy’

II ba 肉 ‘meat’ bee 馬 ‘horse’ bei 買 ‘buy’

III ba ‘meat’ bee ‘horse’ bei ‘buy’

IV bâ' 肉 ‘meat’ bè: 馬 ‘horse’ beỳ 買 ‘buy’

V bâ' ‘meat’ bè: ‘horse’ beỳ ‘buy’

VI bah4 be2 bei2

meat = carne; horse = cauallo (I) ~ cavallo (II, III) ~ caballo (IV, V); buy = comprar

2.2. Initial and Final As shown in Table 15, the EMH sources are fairly consistent in the use of initial before [a] and [u]. My data support van der Loon’s claim that represents the aspirated and unaspirated velar plosives [k] and [kh], spelled as k- and kh- in the TLPA system. Table 15 lists examples of unaspirated initials. In row four of column I, 根 ‘root’, also occurs before [i]. As this is the only instance of a spelling or in the Arte and also not matched by the spellings in the other EMH sources, I do not attribute any significance to this orthographic inconsistency. The use of final has been discussed in Section C-2 above. Distinction of aspirated [kh-] from unaspirated [k-] is made by a spiritus asper , thus vs. (third mode of pronunciation in the BMS, fifth mode in the LMS; see Chapter 4). As Braselmann (1991:196) points out, most Spanish varieties had lost aspiration by 1500. The treatment of aspiration as a ‘mode,’ i.e., non-Spanish feature (see Chapter 4), is therefore consistent with the analytical framework of the Arte. Table 15: Initial 1. 2.

I cám 柑 ‘orange’ càu 九 ‘nine’

II cam 柑 ‘orange’ cau 九 ‘nine’

III cam ‘orange’ --

IV cám 柑 ‘orange’ caù 九 ‘nine’

V cám ‘orange’ caù ‘nine’

VI kam1 kau2

134

CHAPTER FIVE

Table 15 (cont.) 3. 4.

I cúa 瓜 ‘watermelon’ cyń 根 ‘root’

II cue 瓜 ‘watermelon’ quin 根 ‘root’

III -kin ‘root’

IV cuá 瓜 ‘watermelon’ quiń 根 ‘root’

V cuá ‘watermelon’ qín ‘root’

VI kua1 kin1

orange = naranjas (I) ~ naranja; nine = nueue; watermelon = sandia; root = rais (I) ~ raiz (II, IV, V) ~ vena o raiz (III)

The Arte explicitly describes syllables transcribed with a spiritus asper as being pronounced ‘with aspiration’ (con asperaçion). The contrastive function of aspiration in EMH is shown by the examples in Table 16. The examples ‘sail’ and ‘winter’ are selected from the Dictionario. In the TLPA transcription, aspiration is indicated by h in second position of the digraphs kh-, ch-, and th-. Table 16: Spiritus asper for aspiration 1. 2. 3. 4. 5. 6. 7. 8.

EMH ‘sentence’ ‘go to’ ‘sin’ ‘search’ ‘room’ ‘sail’ ‘winter’ ‘allowed’

Douglas (1873) ku3 ‘sentence’ khu3 ‘go to’ cue3 ‘sin’ chue3 ‘search’ pang5 ‘room’ phang5 ‘sail’ tang1 ‘winter’ thang1 ‘allowed’

As shown in Table 17, the use of the spiritus asper for aspiration is attested in the Arte (BMS), the Dictionario, and the Vocabulario. The LMS (not represented in Table 17) does not indicate aspiration outside the section on aspiration, nor does the Bocabulario. Chirino’s transcription uses a letter following the initial letter.

135

PHONOLOGY AND ORTHOGRAPHY

Table 17: Indication of aspiration in EMH sources I 1. cǔc 去 ‘go to’ 2. kît'c 乞 ‘for’

II qhu 去 ‘go to’ qhi 乞 ‘for’

III cu ‘go to’ kir dative

3. pūec 皮 ‘skin’ 4. tángc 可 ‘be possible’ 5. tè'c 提 ‘take’

p.hue 皮 ‘skin’ thang 可 ‘be possible’ the 提 ‘bring’

IV cũc 去 ‘go to’ quît’c 乞 ‘for’, dative c pūec 皮 pue ‘skin’ ‘skin’ tángc 可 tangc ‘can’ ‘be possible’ the tè'c/ těc 提 ‘bring’ ‘take’

V cûc ‘go to’ qît’c ‘for’, dative pūec ‘skin’ tāng ‘be possible’ těc ‘take’

VI khu3 khu7 khit4 phue5 thang1 the2 the3

go to = hir (I) ~ ir (II, III) ~ yr (IV, V); for = para (I, II) ~ determina el datiuo (III) ~ para datiuo (IV, V); skin = cueros (I) ~ piel (II), pellejo (III, V), pellego (IV), be possible ~ can = es liçito (I) ~ es licito (II) ~ poder (iii) poderliçito (IV, V); take ~ bring = toma (I) ~ traer (II, III) ~ tomar (IV, V)

2.3. Initial Initial stands for both the aspirated and the unaspirated voiceless affricates [t͡s] and [t͡sh]. As in Section C-2.2 above, the distinction between aspirated and unaspirated is represented by the spiritus asper in the former case. The examples in Table 18 are all unaspirated. TLPA spellings are c- for the unaspirated initial and ch- for the aspirated initial. Table 18: Initial I 1. chàp' 十 ‘ten’ 2. chêi 多 ‘many’

II chap 十 ‘ten’ chei 多 ‘many’

çhi 銭 3. chīn 銭 ‘money’ ‘money’

III chap ‘ten’ chey ‘many’ chīn ‘money’

IV chàp' 十 ‘ten’ chêy / chěy 多 ‘many’ ch¯yn 銭 ‘money’

V chàp' ‘ten’ chěy ‘many’

VI cap8 cei3 cei7

ci5 chīn ‘money’

136

CHAPTER FIVE

Table 18 (cont.) 4. 5.

I chô 做 ‘make’ chù 煮 ‘cook’

II çho 做 ‘make’ --

III cho ‘make’ chu ‘cook’

IV chô 做 ‘make’ chù 煮 ‘cook’

V chô ‘make’ chù ‘cook’

VI co7 cu2

ten = 10 (I, II) ~ numeral diz (III) ~ diez (IV) ~ dies (V); many = munchos (I) ~ mucho (II) ~ mucgo (III) ~ mucho, muchas cosas (IV, V); money = dinero; make = haçer (I, IV) ~ hazer (II, III, V); cook = coser (I) ~ cozer (III) ~ coçer (IV) ~ coser como comida (V)

Van der Loon notes that ‘we do not know the sound values of these Hokkien consonants in the early seventeenth century; the Castilian ch, though suggesting an alveopalatal, may merely have been used for lack of a better alternative’ (1967:114). To this we can add that the digraph is attested for a post-alveolar affricate [tʃ] in Spanish missionary sources of the sixteenth and seventeenth century. This usage is described in, among others, Monzón’s (2005) analysis of the orthography of Tarascan (see Chapter 4) and Hovdhaugen’s (2005) account of Mochica (spoken at the north-western cost of Peru) orthography. 2.4. Initial and Final The initial represents a voiced velar plosive [g], spelled as g- in TLPA. As regards , two usages must be distinguished. If is the final letter of the syllable, it stands for the vowel [u] (see below). If followed by or , as in ‘hot’ (f. 90v) and ‘like, as’. I therefore claim that phonemically stands for a single voiced affricate, which has either the value [d͡z] or [d͡ʒ]. Its TLPA spelling is j-. Table 29: Initial I 1. xī 二 ‘two’ (bound)

II tzi 二 ‘two’ (bound)

2. xī 字 ‘letter, charcacter’ 3. xìt' 日 ‘day’ 4. xiōng 絨 ‘silk’

tzi 字 ‘letter charcacter’ tzit 日 ‘sun, day’ tziong 絨 ‘light silk’

III IV s yi xˉy/ ‘two’ xˇy 二 ‘two’ (bound) -xˇy 字 ‘letter’ ysit ‘sun’

xìt' 日 ‘day’ xiōng 絨 ‘silk’

V xˇy ‘two’ (bound)

VI ji5 ji3

xˇy ‘letter’

ji5 ji3

xìt' ‘day’ xiōng ‘silk’

jit8 jiong5

two = 2 (I, II) ~ dos (III, IV, V); letter, character = deduced from «sià xī | 寫 字» ‘write’ in (I) and «Escrevir | Sa tzi | 寫字] in (II) ~ letra (IV, V); day = dia (I) ~ sol, dia (II) ~ dia (IV, V), sun = el sol (III); silk = seda ~ seda floxa (II)

2.17. Initial As discussed in the previous paragraph, van der Loon refers to an initial with superscript . As shown in Table 24 above, however, these are apparently orthographic alterations used in the Bocabulario only and therefore not relevant for the Arte (BMS). As I argue below (Section D-2), initial in the Arte stands for the vowel [i].

144

CHAPTER FIVE

3. EMH Consonant Phonemes The seventeen consonant spellings (including diagraphs) analyzed in Sections C-2.1 to C-2.17, some of which are used in combination with the spiritus asper (C-2.2), can be interpreted as representing fourteen consonant phonemes in EMH. Examples of these phonemes and the allophones represented in spelling are given in Table 30. Table 30: Examples of EMH consonants 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Phone /b/ [b] /b/ [m] /p/ /ph/ /l/ [l] /l/ [n] /t/ /th/ /d͡z/ /t͡s/ /t͡sh/ /g/ [g] /g/ [ŋ] /k/ /kh/ /s/ /h/

Arte b m p p+c lnt t+c xchch- + c g, gung c, k, quc-, k-, qu- + c sh-

Example ‘horse’

‘month’ stand for gueh8 or for geh8? Both forms are attested in modern Hokkien dialects. On the basis of the available EMH data, it is impossible to find a satisfactory answer to this question. All in all, however, the phonemic status of [u] as well as [a] and [i] is unambiguously evidenced by minimal pairs like can3 ‘help’ vs. cun3 ‘handsome’ and cun3 ‘handsome’ vs. cin3 ‘last’. The phoneme /i/ is also spelled as . The transcription system used in the Arte does not systematically distinguish and . For example, the same morphemes ti2 ‘which’ and lai5 ‘come’ are spelled as ~ and ~ respectively. Parallels can be found in the spelling of Spanish words, for example, of ~ ‘there is’. 3. Interpretation of The value and the phonemic status of these vowels are unclear due to a lack of comparable examples. In the Arte, the only occurrence of is in ‘place’ (in ‘this place, here’ and ‘that place, there’). Examples in the Dictionario are also scarce, te2 ‘short’ and ce3 ‘sit’ being among the few. In most cases, co-occurs with either two dots or in combination with the letter . Examples of include ‘crawl’, ‘under’, ‘father’, and ‘yesterday’ (occurring four times) and ‘pigeon’ (occurring once). As the use of a typographic convention without prior explanation does not fit into the approach of the Arte, I assume that both examples represent scribal mistakes. Moreover, both examples occur without two dots in the Dictionario and in the Vocabulario. In the case of ‘pigeon’, it must moreover be taken into consideration that syllables ending in an unreleased final -p by definition carry a vertical stroke. Therefore, the transcription and phah8 ≈ phah4 ‘strike, beat’

200

ARTE DE LA LENGUA CHIO CHIU

f. 4r, right column [LMS: 319v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

responde pà' kǐnnìa 打呱仔 c

açote al muchacho =======\\=======\\======= De ordinario sichui no sirue sino para honbres {y cosas in telectuales} ________\\________\\________ lù pà'c 汝打 tu sacudes te sǐ chûi 是誰 a quien =======\\=======\\======= simi sirue para todas las cosas ut. ________\\________\\________ sǐ mì' lāng 是乜人 que hombre sǐ mì' gùe' 是乜月 que mes sǐ mì' pìt' 是乜筆 que pluma =======\\=======\\======= Diferencia quando se pregunta con sichui noselepone despues el nombre que se pregunta pero al simi mi añadesele ut. ________\\________\\________ sǐ chûi lāi 是誰來 quien bie ne sǐ chûi cǔc 是誰去 quien fue

TRANSCRIPT AND TRANSLATION

201

TRANSLATION [1–4]

Answer phah8 kin3-nia2 打呱仔 ‘I hit the child.’

[5–8] [9–11]

[12–13] [14–17]

Normally si3-cui7 only serves persons and intellectual things: lu2 phah8 si3-cui7 汝打是誰 you hit whom  ‘Whom do you hit?’ Si3-mih8 serves all things, for example: si3-mih8 lang5 是乜人 ‘What person?’ si3-mih8 geh8 是乜月 ‘What month?’ si3-mih8 pit8 是乜筆 ‘What quill?’

[18–22]

[23–25]

A difference between asking with si3-cui7 and si3-mih8 is that the noun asked for is not placed after si3-cui7, but in the case of si3-mih8, mih8 is attached to it, for example: si3-cui7 lai5 是誰來 ‘Who came?’ si3-cui7 khu3 是誰去 ‘Who is going?’

Notes 2, etc. 16, etc.

kin3-nia2 ≈ kin2-na2 ‘child’ po2-pueh4 ≈ po2-pue3 ‘precious things’ EMH diminutives are analyzed in Chappell (2000). The tones of and are not indicated. These diminutives are not attested in Douglas (1873).

212

ARTE DE LA LENGUA CHIO CHIU

f. 5v, right column [LMS: 322r] 1 2 3 4 5 6 7 8 9 10 11 12 13

cap. iii. de la conjuga çion de los ueruos Del ueruo sun est .fui. en esta len gua esta particula——

šy

是

ser

=======\\=======\\===== es sun es fui y no tiene mas termi naçion ni di diferencias de personas o modos si no es con unas particu las que despues se diran. ________\\________\\______ guà šy 我是 yo soi hò sím 好心 bueno lù šy hò sím

汝是 好心

tu eres bueno

17

ý šy hò sím

伊是 好心

aquel es bueno

18

guàn ši

阮是

20

hò sím

好心

nosotro[s] somos buenos

21

lǔn šy

恁是

hò sím

好心

14 15 16

19

22 23

vosotros sois buenos

TRANSCRIPT AND TRANSLATION

213

TRANSLATION [1–2]

Chapter 3: On the Conjugation of Verbs

[3–11]

On the verb sun est fui in this language. This particle si3 是 be is sun and est and fui. It has no more endings and neither distinguishes persons or moods unless it occurs with particles mentioned later.

[12–23]

gua2 si3 ho2-sim1 我是好心 ‘I am kind.’ lu2 si3 ho2-sim1 汝是好心 ‘You are kind.’ i1 si3 ho2-sim1 伊是好心 ‘He is kind.’ guan2 si3 ho2-sim1 阮是好心 ‘We are kind.’ lun3 si3 ho2-sim1 恁是好心 ‘You are kind.’

Notes 1–23 3

This section is discussed on pp. 99–100. In missionary grammars, the meaning ‘to be’ is often indicated by the use of Latin sum esse fui (also sum est fui or sum es fui; cf. Zwartjes, forthcoming:20). The use of sun est fui and sun es fui (f. 6r, left column, line 8) is unusual and inconsistent.

214

ARTE DE LA LENGUA CHIO CHIU

f. 6r, left column [LMS: 322r–322v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

ýn šy 因是 aquellos son hò sím 好心 buenos =======\\=======\\======= La variaçion de los demas mo dos y tienpos es con forme la de los demas ueruos que luego se pon dra açerca deste ueruo sun est fui se a denotar en esta lengua sun es fui no sinifica estar en lugar ni te ner como en la latina munchas veçes se suple este ber {b}o quando signifia ser açeden te de offiçio se suple coneste beruo chó que es açer ut. ________\\________\\________ guà chô 我做 yo soi hōng těi 皇帝 Rey =======\\=======\\======= Suplese tanbien sun es fui con esta particula .pe. que sinifica ygual dad. ut. =======\\=======\\======= lǔn lāng 恁人 vosotros g[xxx] pēn pēn gāu 並並賢 soys igual mente sa uios ________\\________\\________ suplese con adgetiuos ________\\________\\________ lù hò sím 汝好心 {tu eres} bueno {tu eres bueno} ________\\________\\________

TRANSCRIPT AND TRANSLATION

215

TRANSLATION [1–3]

[4–16]

[17–19]

[20–23]

[24–29]

[30–31] [32–34]

in1 si3 ho2-sim1 因是好心 ‘They are kind.’ Variation associated with other moods and times matches other verbs that will be mentioned later. With regard to what is expressed by the verb sun est fui in this language: sun es fui neither means ‘be at a place’, nor ‘have’ as in Latin. Many times this verb is substituted. When it means ‘hold an office’ it is substituted with the verb co7 which means ‘to do’. gua2 co7 hong5-tei3 我做皇帝 ‘I am king.’ Sun es fui is also replaced by the particle penn5 which signifies equality, thus: lun3 lang5 penn5-penn5 gau5 恁人並並賢 ‘You are equally wise.’ It is substituted with adjectives. lu2 ho2-sim1 汝好心 ‘You are kind.’

216

ARTE DE LA LENGUA CHIO CHIU

f. 6r, right column [LMS: 322v–323r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 margin 12 [right]

suplese con esta particula bu kieg. ut. ________\\________\\________ dīo ši 僚氏 dios es bū kìg' tôa 無極大 infinita mente gran de ________\\________\\________ suplese con conparatiuo {y con sustantiuos ut} ________\\________\\________ dīo ši 僚氏 dios es chîa' [x] ši 即是賢 muy sauio gāu =======\\=======\\======= en esta lengua una mesma terminaçion y vos sirue para todos las modos tienpos y personas barianse con algun nonbre o ad ueruio que sinifique tienpo pre sente o pasado o poruenir de ordi nario se conoçe la diferençia de tienpo de lo que se ua tratando el presente munchas veces si cono çe sin particula con la bos sola. ut. =======\\=======\\======= presente de indicatiuo gùa lāy 我來 yo uengo lù lāy 汝來 tu bienes ý lāy 伊來 aquel biene guà cǎc hò lù

我可 好你

yo soi mejor que tu

TRANSCRIPT AND TRANSLATION

217

TRANSLATION [1–3] [4–8]

[9–11]

[12–15]

It is substituted with the particle bu5-kik8, for example: tio5-si3 bu5-kik8 tua7 僚氏無極大 ‘God is infinitely great.’

It is substituted with comparatives {and with substantives, thus:} tio5-si3 ciah4 si3 gau5 僚氏即是賢 ‘God is very wise.’

[16–26]

In this language, the same term and word serves all moods, times, and persons. These change with some nouns or adverbs signifying present tense or future. Ordinarily the time difference is known. What is expressed in the present is known without particle, only by the word.

[27–30]

Present of the indicative gua2 lai5 我來 ‘I come.’ lu2 lai5 汝來 ‘You come.’ i1 lai5 伊來 ‘He comes.’

margin

gua2 kha3 ho2 lu2 我可好你 ‘I am better than you.’

Note 12

The LMS does not have the example added in the margin.

218

ARTE DE LA LENGUA CHIO CHIU

f. 6v, left column [LMS: 323r–323v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

guàn lāy

nosotros be n{i}mos lǔn lāy 恁來 vosotro benis ín lāy 因來 aquelos bienen =======\\=======\\======= quando se le añade alguna par ticula que sinifica presente sin x quibocaçion se conoçe ser indica tiuo ut. =======\\=======\\======= tán chîa' lāy 今即來 agora uin{o} bàc' chēng 目前 ahora luego [x] lāy x來 vendra ki{´}nīlāy 今年來 este ano + kin ni lai bino n quín t[x]ǒa 今旦日來 oy uino {x}ìt lāy [tón] kím lāy 當今來 agora uiene tán lāi 今來 agora uiene bà' chè[x]¯ lāi 目前來 agora [vendra?] cheng + [xxx] =======\\=======\\======= Para el preterito inperfecto no ay particula conoçer sea ser pre terito inperfecto por lo que [se] açe ut. ua [açertando] segun al sentido que ________\\________\\________ 阮來

margin: 20 [left] lay, 27 [left] cheng

219

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–7]

guan2 lai5 阮來 ‘We come.’ lun3 lai5 恁來 ‘You come.’ in1 lai5 因來 ‘They come.’

[8–12]

[13–28]

When a particle unequivocally signifying present is added, [the verb] is recognized to be indicative, thus: tann1 ciah4 lai5 今即來 just he has come  ‘He has just come.’ bak8-cing5 lai5 目前來 now/soon he will come  ‘He will come soon.’ kin1-ni5 lai5 今年來 this year he came  ‘He came this year.’ kin1-toann3-jit8 lai5 今旦日來 today he comes  ‘He comes today.’ ton1-kim1 lai5 當今來 now he comes  ‘He comes now.’ tann1 lai5 今來 now he comes  ‘He comes now.’ bak8-cing5 lai5 目前來 now he comes  ‘He comes now.’

[29–34]

For the preterite imperfect, there is no particle. You will know that it is preterite imperfect by what is spoken about according to the sense it makes, for example: Notes on p. 221

220

ARTE DE LA LENGUA CHIO CHIU

f. 6v, right column [LMS: 323v–324r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

chá: bôu {x}ìt' lù lāi

查暮日 ayer 汝來 quando beniste hù [tán] sī 許今時 entonçes pálè 巴礼 el padre liâm kéng reçiua =======\\=======\\======= donde se conose claramente que el ueruo liam es preterito inperfecto el preterito perfecto se forma con estas particulas =======\\=======\\======= là liàu là 了 liàu 了 n n tiá mǔa 今晏 =======\\=======\\======= y con nombres de tienpo que sini fican tienpo passado. ut. ________\\________\\________ cû nī lāy 年來 el ano pasa do bino chá: bôu 查暮 ahir

TRANSCRIPT AND TRANSLATION

221

TRANSLATION [1–7]

ca:1-bou7-jit8 lu2 lai5 hu2-tann1-si5 pa1-le2 liam7 king1 查暮日汝來許今時巴礼 yesterday when you came then the father was praying  ‘The father was praying when you came yesterday.’

[8–12]

Here it is clearly known that the verb liam7 is preterite imperfect. The preterite perfect is formed with these particles:

[13–17]

[18–20]

[21–23]

la2 liau2 la2 了 2 liau 了 tiann1 mua3 今晏 The preterite perfect is also formed with time nouns indicating passed time. ku7-ni5 lai5 年來 last year he came  ‘He came last year.’ ca:1-bou7 [jit8 lai5] 查暮[日來] yesterday [he came]  ‘He came yesterday.’

Notes f. 6v, left column 21 *ton1 ≈ tong1 (cf. LMS f. 314v), Douglas: tong1-kim1 ‘now’ f. 6v, right column 1 ca1 ≈ ca7, the spelling is on p. 147. 6 =念 =經 6 13, 14 =朥 21 =舊

222

ARTE DE LA LENGUA CHIO CHIU

f. 7r, left column [LMS: 324r–324v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

xìt' lāy 日來 bino ________\\________\\________ exemplo del preterito perfecto ________\\________\\________ guà lāy 我來 yo bi[x]ne là tía mǔan lāy 今旦來 denantes uino lù lāy là 你來 tu beniste ý lāi là 伊來 aquel bino guàn lāy là 阮來 nosotros venimos lǔn lāy là 恁來 vosotros benistes ín lāy là 因來 aquellos uinieron =======\\=======\\======= y ansi se conjuga con las demas particulas o diferençias de tiem po de año o dia passado. con este beruo u. que es tener se suele açer prete rito como. ut. =======\\=======\\======= lù ù 你有 tu tienes le pà'c ý 打伊 sacudido ā [bō] 亜無 o no ________\\________\\________

TRANSCRIPT AND TRANSLATION

223

TRANSLATION [1–2]

[3–4] [5–16]

[ca:1-bou7] jit8 lai5 [查暮]日來 [yesterday] he came  ‘He came yesterday.’ Example of the preterite perfect: gua2 lai5 la2 我來 ‘I came.’ tiann1-mua3 lai5 今旦來 a moment ago he came  ‘He came a moment ago.’ lu2 lai5 la2 你來 ‘You came.’ i1 lai5 la2 伊來 ‘He came.’ guan2 lai5 la2 阮來 ‘We came.’ lun3 lai5 la2 恁來 ‘You came.’ in1 lai5 la2 因來 ‘They came.’

[17–22]

[23–27]

And like this one conjugates with other particles or differences of time, year, or passed days; and with this verb u2, which is ‘have’, we usually have the preterite, as for example: lu2 u2 phah8 i1 a5 bo5 你有打伊亜無 you have him hit or not  ‘Have you hit him?’

Note 23, etc.

u2 ≈ u7 ‘have, exist’

224

ARTE DE LA LENGUA CHIO CHIU

f. 7r, right column [LMS: 324v–325r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 margin under 27

el preterito plus quamperfecto no tiene particula determinada para se conoçer conoçer sea como el preterito inperfecto deloque antecedo por el sentido que açe y [{x}untando?] al ueruo de plusquam perfecto esta particula liaula. ut. =======\\=======\\======= chá: bôu xìt' 查暮日 ayer lù lāy 你來 quando tu benistes guà liâm 我 yo xia abia quéng liàu 了 rrecado là =======\\=======\\======= el futuro se açe con estas particulas siguintes ________\\________\\________ bûe' 欲 querer luego aǔ 後 despues hià cxuà 歇 仔 de aqui a un poco aǔ lāy 後來 despues +{tia}[xxx] qu[a] 霎也 idem bīnoa {x}ìt' 向日 mañana mēhui 昏 a la tarde a prima noche ________\\________\\________ {+tiap yà' qùua}

TRANSCRIPT AND TRANSLATION

225

TRANSLATION [1-8]

[9–15]

[16–18] [19–28]

The preterite pluperfect does not have a specific particle by which it can be recognized. It can be recognized like the preterite imperfect by what preceded and the sense it makes; the particle liau2-la2 is joined to verbs in the pluperfect. ca:1-bou7-jit8 lu2 lai5 gua2 liam7 king1 liau2-la2 查暮日你來我 了 yesterday when you came I had prayed  ‘I had already prayed when you came yesterday.’ The future is formed with the following particles: bueh4 au3 hia2 ku-a2 au3-lai5 tia-ia-ku bin5-oa-jit8 me5-hui

欲 後 歇 仔 後來 霎也 向日 昏

‘wish, soon’ ‘later’ ‘almost ready to’ ‘later’ idem ‘tomorrow’ ‘in the evening, after dark’

Notes 3 6 20, etc. 21 21–22

25 margin

In the LMS, the word is not repeated. The repetition makes sense if we posit a sentence boundary after the first . The first word is difficult to decipher due to later corrections. It reads in the LMS; Ishizaki (2006:168) posits the reading as a variant spelling for . au3 ≈ au7 ‘behind’ =久 hia2 ku-a2 ≈ hiah4 ku2-a2 ‘quite lately.’ The Spanish translation de aqui a un poco seems to be of Biblical origin, cf. Exodus 17/4: ‘Entonces clamó Moisés a Jehová, diciendo: ¿Qué haré con este pueblo? De aquí a un poco me apedrearán.’ (Moses cried to Yahweh, saying, ‘What shall I do with these people? They are almost ready to stone me.’; trans. World English Bible). =冥 ) and 15 (); line 14 corresponds to the modern reading boh8. 14, etc. tann7 ≈ tann3 ‘speak’ 24, etc. ji5 ≈ ji7 ‘character’; = 寫 f. 8r, left column = 袂; bei3 ≈ bei7 ‘cannot’ 2, etc. Notes continued on p. 235

234

ARTE DE LA LENGUA CHIO CHIU

f. 8v, left column [LMS: 326v–327r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

con gue. se forma la pasiua ponien do primero la persona que padesse luego sun es fui luego la persona que [pa+de]{çe} luego el ueruo ulti mamente la particula .gue. ut. ________\\________\\________ actiua Dīo ši 僚氏 Dios cn hǔa tí têi 化天地 crio çielo y tierra Paçiua tícn têi 天地 çielo y tierra ši dīo ši 是僚氏 por dios hǔa gùe 化个 fueron criados =======\\=======\\======= ay una diferençia entre estas dos particulas dichas y modo de paçiuas que el kit. de ordinario sirua pa ra beruos cuya açion presto pasa pero el gue. para los que tiene e fecto. permanente como consta de los exenplos puestos aun[que] alguna ues se puede poner que para cossas que cito transeunt. ut. ________\\________\\________ chǐ ǒa šy 只話是 esta palabra es n gùa tâ gùe 我詚个 por mi abla da =======\\=======\\=======

margin 4 [left]

+ açe

TRANSCRIPT AND TRANSLATION

235

TRANSLATION [1–6]

Passives with ge2 are formed by placing first the person who suffers and then sun es fui and then the person doing the act and then the verb and finally the particle ge2.

[7–15]

Active tio5-si3 hua3 thinn1-tei7 僚氏化天地 ‘God created heaven and earth.’ Passive thinn1-tei7 si3 tio5-si3 hua3 ge2 天地是僚氏化个 heaven and earth by God were created  ‘Heaven and earth were created by God.’

[16–25]

[26–30]

There is a difference between the two mentioned particles and the mode of passive they express, namely, the khit4 normally serves verbs whose action quickly passes. The ge2 is for those having a permanent effect as in the given examples. Sometimes, however, ge2 can be placed with things which pass quickly. ci3 ua3 si3 gua2 tann7 ge2 只話是我詚个 this word is by me said  ‘This word was said by me.’

Notes f. 8r, left column 4, etc. ei3 ≈ ei7 ‘can’ 19–20 The correction corresponds to the LMS, cf. f. 326r: . f. 8r, right column 6ff. These passive constructions are analyzed in Chappell (2000). f. 8v, left column 9, etc. tei7 ≈ tei3 (te3 ‘earth’) 26 ua3 ≈ ua7 ‘words, speech’

236

ARTE DE LA LENGUA CHIO CHIU

f. 8v, right column [LMS: 327r–327v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

cap. iiii De los adueruios estos çinco adueruios que se sigue son adueruios de tienpo =======\\=======\\======= tì s−y 時 quando chûn [siche] 時[ ] quando {sīchê} hù sī 許時 entonçes hù tán sī 許今時 entonçes hù chèg' sī 許一時 entonçes =======\\=======\\======= el segundo que es .chun. es corela tiuo del primero ques . ti si. y los tres ultimos son correlatiuo del segundo ques chun. el ti si es para preguntar y se antepone al ueruo y se rresponde con el chun posponiendo lo a la oraçion. ut. =======\\=======\\======= Pregu.ta lù tì sī 你 時 tu quando lāy 來 beniste responde

margin 7 8

left margin: left margin, under :

TRANSCRIPT AND TRANSLATION

237

TRANSLATION [1–4]

Chapter 4: On Adverbs The following five adverbs are adverbs of time.

[6–11]

ti2-si5 si5-ce7 hu2-si5 hu2 tann1-si5 hu2 cik8-si5

時 時 許時 許今時 許一時

‘when’ ‘when’ ‘then’ ‘then’ ‘then’

[12–19]

The second, which is cun7, correlates to the first, which is ti2si5. The last three correlate with the second which is cun7. The ti2-si5 is for asking and is preposed to the verb. The answer is with cun7 postposed to the clause.

[20–23]

Question lu2 ti2-si5 lai5 你 時來 you when came  ‘When did you come?’ Answer

Notes 7 13

= 郎. Note that the modern Hokkien readings lang1 ~ nng5 ~ long5 bear no resemblance to . Note that cun7 was originally listed as the second ‘adverb of time’ in line 7.

238

ARTE DE LA LENGUA CHIO CHIU

f. 9r, left column [LMS: 327v–328r] 1

lù chìa'

你食

2

quando tu comias

20

pûin chûn =======\\=======\\======= quando el chun no corresponde al tisi como relatiuo que presede al chun. corresponden como rrela tiuos alguno de los tres ultimos ut. ________\\________\\________ lù chìa' 你食 quando tu co n {pûi } chûn [xxx] mias [xxx] hù tán sī 許當時 entonçes guà lāi 我來 yo biene lù liâm 你 quando tu reça kéng ch{û}n uas hù sī 許時 entonçes guà laī 我來 yo biene

21

lù lāy

你來

chûn hù chèg' sī

許一時

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

22 23 24

quando tu be niste entonçes

TRANSCRIPT AND TRANSLATION

239

TRANSLATION [1–4]

[5–9]

[10–24]

lu2 ciah8 puinn7 cun7 你食 when you ate  ‘When you were eating.’ When the cun7 does not correspond to ti2-si5 as a relative preceding cun7, then one of the last three [expressions listed on f. 8v, right column, lines 9–11] corresponds as relative, for example: lu2 ciah8 puinn7 cun7 hu2 tann1-si5 gua2 lai5 你食 許當時我來 when you ate then I came  ‘I came when you were eating.’ lu2 liam7 king1 cun7 hu2 si5 gua2 lai5 許時我來 你 when you prayed then I came  ‘I came when you were praying.’ lu2 lai5 cun7 hu2 cik8 si5 [gua2 khun3] 你來 許一時[我困] when you came then I slept  ‘I was sleeping when you came.’

Note 3, 10

=飯

240

ARTE DE LA LENGUA CHIO CHIU

f. 9r, right column [LMS: 328r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

guà cǔnc 我困 yo dormia =======\\=======\\======= estos adueruios que se siguen son adueruios de lugar. =======\\=======\\======= čhi tě 只處 aqui hù tě 許處 aquella ǎn čhi tě 按只處 por aqui ǎn hù tě 按許處 por aquella čhi pēng 只砰 desta uanda hǐon pēng 向砰 de la otra banda câu 到 asta ǎn 按 por tě 處 locusinquo =======\\=======\\======= como se aya de usar de los aduer uios dichos costa[i]a de la oraciones siguintes. =======\\=======\\======= pálè 巴礼 el padre tù chǐ tě 在只處 esta aqui

TRANSCRIPT AND TRANSLATION

241

TRANSLATION [1–2]

[3–5] [6–16]

[17–20]

[21–22]

[lu2 lai5 cun7 hu2 cik8 si5] gua2 khun3 [你來 許一時]我困 when you came then I slept  ‘I was sleeping when you came.’ The following adverbs are adverbs of place. ci3 te3 hu2 te3 an3 ci3 te3 an3 hu2 te3 ci3 ping5 hionn3 ping5 kau7 an3 te3

只處 許處 按只處 按許處 只砰 向砰 到 按 處

‘here’ ‘there’ ‘from here’ ‘from there’ ‘this side’ ‘that side’ ‘to’ from ‘locus in quo / the place in question’

How the given adverbs can be used may become obvious from the following sentences. pa1-le2 tu2 ci3 te3 巴礼在只處 ‘The Father is here.’

Note 8–12

There is a mismatch between the examples and the Spanish translations; my translations are based on the EMH examples. In modern Hokkien dialects, the preposition an3 means ‘from.’ The EMH examples ci3 ping5 ‘this side, here’ and hionn3 ping5 ‘that side, there’ are not preceded by a preposition ‘from’, as suggested by the Spanish translation.

242

ARTE DE LA LENGUA CHIO CHIU

f. 9v, left column [LMS: 328r–328v] 1 2 3

Pá lè tù hù tě

4 5 6 7

Pá lè ǎn čhi tě lāy

8 9 10 11

Pá lè ǎn hù tě lāy

12 13 14

chîo pēng càng

15 16 17

n

hiǒ pēng càng

hute 巴礼 在許處 anchite 巴礼 按只處 來 anhute 巴礼 按許處 來 chi{o}peng 障砰港 hiopeng 向砰港

el padre asta aquella el padre por aqui bino el padre por aculla uino desta uanda del rrio de la otro uan da del rrio

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–3]

hu2 te3 pa1-le2 tu2 hu2 te3 巴礼在許處 ‘The Father is there.’

[4–7]

an3 ci3 te3 pa1-le2 an3 ci3 te3 lai5 巴礼按只處來 Father to this place came  ‘The Father came here.’

[8–11]

an3 hu2 te3 pa1-le2 an3 hu2 te3 lai5 巴礼按許處來 Father to that place came  ‘The Father came to that place.’

[12–14]

cionn7 ping5 cionn7 ping5 kang2 障砰港 ‘on this side of the river’

[15–17]

hionn3 ping5 hionn3 ping5 kang2 向砰港 ‘on the other side of the river’

243

244

ARTE DE LA LENGUA CHIO CHIU

f. 9v, right column [LMS: 328v] 1 2 3 4

lāi câu minilà' bâr' câu chǐn

5 6 7 8

bâr' kéng câu chǐn

cau. 來到 uino asta 民希臘 manila 識到盡 saue asta el fin 識 saue el reco 到盡 todo asta el fin

f. 10r, left column [LMS: 328v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

margin

cap. v.{i.} de otras par ticula o adueruios Ay algunas Particulas que tie nen particular dificultad y an si sepondran aqui todas juntas para que se sepa como se a de usar dellas. estos particulas .sa. y. sio. son reciprocas pero el . sio es lengua de chio chio no el sa. ________\\________\\________ hù nò lāng 許二人 aquellos dos honbres sió pà'c 相打 se sacuden lǔn sío sîo' 恁相惜 vosotr{os} amaos

bottom margin, under left and middle column: chi no lang siang ay estos dos se aman bottom margin, under right column: [X]n[xxx] mugeres

TRANSCRIPT AND TRANSLATION

245

TRANSLATION f 9v, right column [1–8]

kau7 lai5 kau7 mi-ni-lah8 來到民希臘 ‘He came to Manila.’ bat4 kau7 cin3 識到盡 know to the end  ‘know thoroughly’ bat4 king1 kau7 cin3 識 到盡 know the prayer to the end  ‘know the prayers thoroughly’ f 10r, left column

[1–2]

Chapter 5: On Other Particles or Adverbs

[4–11]

Some particles are particularly difficult. Thus, we will list them all here in order to know how to use them. The particles sann1 and sio1 are reciprocal. The sio1 belongs to the Chio Chiu language, but not the sann1.

[13–17]

hu2 no2 lang5 sio1 phah8 許二人相打 ‘Those two men hit each other.’ lun3 sio1 sioh4 恁相惜 ‘you love [each other]’

margin under left and middle column: chi no lang siang ay ‘These two persons love each other’ under right column: ? women Notes on p. 247

246

ARTE DE LA LENGUA CHIO CHIU

f. 10r, right column [LMS: 328v–329r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

lǔn sío nàu

vosotros os aboreçeis =======\\=======\\======= La negatiua se açe anteponiendo a la particula la negaçion. ut. ________\\________\\________ {hù} nò guè lāng 許二个人 aquellos dos honbres mˇ sío sîo' 不相惜 no se aman =======\\=======\\======= en las demas oraçiones negatiuos la negaçion se pone antes del ber uo en mediate. ________\\________\\________ tou 恁相僥

27

esta particula .tou. sinifica en todo y por todo y ansi aumenta la sini ficaçion de los veruos. gua tou m bar l{u} yo en ningun modo te co nosco lu tou m bar lay de nin gun modo bienes. =======\\=======\\======= guà tóu 我都 yo de ningun modo [te co mˇ bâr' lāi 不識來 nosco] {bine}

28

lù tóu

汝都

mˇ bâr' lāi lù tóu si h{o+sī}m

不識來 汝都 是好心

25 26

29 30 31 32

margin under 32

sǐ hò + sím

tu de ningun modo bienes entodo eres buenos

TRANSCRIPT AND TRANSLATION

247

TRANSLATION lun3 sio1 nau2 恁相僥 ‘You annoy each other.’

[1–3]

[4–6]

The negative is rendered by preposing the negation to the particle, thus: hu2 no2 ge2 lang5 m3 sio1 sioh4 許二个人不相惜 ‘Those two men do not love each other.’

[7–10]

[11–14]

In the other negative phrases, the negation is placed immediately before the verb.

[15–23]

tou1 This particle means ‘in all’, ‘for all’, thus it increases the meaning of the verbs. Gua2 tou1 m3 bat4 lu2 ‘I by no means know you’, lu2 tou1 m3 bat4 lai5 ‘by no means you come.’

[24–32]

gua2 tou1 m3 bat4 lai5 我都不識來 I by no means came.  ‘I have never been here.’ lu2 tou1 m3 bat4 lai5 汝都不識來 you by no means come  ‘You have never been here.’ lu2 tou1 si3 ho2-sim1 汝都是好心 in all you are good  ‘You are kind in everything.’

margin

[‘are

kind’]

Notes f. 10r, left column 9–11 ≈ sann1 particle expressing mutual or reciprocal action f. 10r, right column 1 nau2 ≈ nau7 ‘noise or confusion’ Notes continued on p. 249

248

ARTE DE LA LENGUA CHIO CHIU

f. 10v, left column [LMS: 329r–329v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

pe esta particula. pe. sinifica ygual dad a la cossa aque se ajunta pone se despues del sugeto antes del p[re] dicado y si sera duplica significa grande ygualdady puedes po ner con esta particula. cang. y si n ella como constara de los exen plos siguientes =======\\=======\\======= lù câng gùa 汝共我 tu y {y}o n pē tôa 並大 somos ygual mente grandes dīo ši pè: câng dīo ši kǐan tóu pēn pēn tôa

僚氏父 共僚氏 子都 並並大

Dios padre con Dios hijo entodo son ygual mente gran de

lǔn nò 恁二人 vosotros lāng dos tó{u} pēn kiâp'c 都並怯 Entodo[x] ygual mente sois vellacos =======\\=======\\======= na esta particula. na. es condiçional sinifica si. Y otras beçes si no po ne se inmediatemente despues

TRANSCRIPT AND TRANSLATION

249

TRANSLATION [1–11]

penn5 This particle penn5 expresses equality with the thing with which it combines. It is placed after the subject and before the predicate. When duplicated, it signifies great equality. It can be placed with the particle kang7 and without it, as will be obvious from the following examples.

[12–14]

lu2 kang7 gua2 penn5 tua7 汝共我並大 ‘You and I are equally big.’

[15–20]

tio5-si3 pee2 kang7 tio5-si3 kiann3 tou1 penn5-penn5 tua7 僚氏父共僚氏子都並並大 God the Father and God the Son both are equally great  ‘God the Father and God the Son are equally great.’

[21–26]

lun3 no2 lang5 tou1 penn5 khiap4 恁二人都並怯 you two in everything are equally wicked  ‘You two are equally wicked.’

[27–30]

na7 The particle na7 is conditional and means ‘if’, and sometimes ‘if not’. It is placed immediately after

Notes f. 10r, right column 9, etc. m3 ≈ m7 ‘not’ 15–30 My translation ‘have never been here’ is based on the assumption that the EMH phrase tou1 m3 bat4 lai5 implies experientiality expressed by bat4 in modern Hokkien dialects (cf. Chappell 2001c). This is not clearly rendered by the Spanish translations.

250

ARTE DE LA LENGUA CHIO CHIU

f. 10v, right column [LMS 329v–330r] 1 2 3 4 5 6 7

del nominatiuo {quando sinifi ca si}. =======\\=======\\======= lù nâ 汝那 si tu mˇ lāy 不來 no bienes c guà běi cǔ 我 去 yo no puedo hir

25

bō lāng lāy 無人來 no ubo hon bre que binie sse nâ sǐ 那是 si no es pá lè 巴礼 el padre tîan tîan 定定 tan solam.te =======\\=======\\======= cang esta particula. cang. tiene particu lar dificultad por no penetrar su significaçion unas vezes sini fica con. i sirue para ablatiuo co mo queda dicho otras vezes es ueruo y sinifica ayudar. ut. =======\\=======\\======= lù câng gùa 汝共我 tu ayuda me bèi hū 買魚 a comprar pes cado

26

lù bèi

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

汝買

conprame

TRANSCRIPT AND TRANSLATION

251

TRANSLATION [1–3]

the nominative {when it means if}.

[4–7]

lu2 na7 m3 lai5 gua2 bei3 khu3 汝那不來我 去 if you do not come I cannot go  ‘I cannot go if you don’t come.’

[8–14]

bo5 lang5 lai5 na7 si3 pa1-le2 tiann7-tiann7 無人來那是巴礼定定 ‘No one came, unless it was the father alone.’

[15–22]

kang7 This particle kang7 is particularly difficult, due to its impenetrable meaning. Sometimes it means ‘with’ and serves the ablative, as said before. Sometimes it is a verb meaning ‘help’, thus:

[23–26]

lu2 kang7 gua2 bei2 hu5 汝共我買魚 ‘You help me to buy fish.’ lu2 bei2 [gua2 hu5] 汝買 [我魚] buy me [my fish]  ‘Buy fish for me.’

Note 8–14

My English translation is based on the Spanish gloss. The EMH sentence is ungrammatical when judged from the perspective of modern Hokkien grammar. It seems to be constructed on the basis of the Spanish sentence structure.

252

ARTE DE LA LENGUA CHIO CHIU

f. 11r, left column [LMS 330r–330v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

gùa hū lù câng gùa chù pûin

mi pescado tu ayudame a coser mo risqueta =======\\=======\\======= Pero es de aduirtir que son equi bocas estas oraçiones puede açer sentido que me ayude en conprar pescado de otros para mi o que el pescado es mio que lo conpre tan bien açe sentido equeboco en uer uos de [conprar] {prestar ut} =======\\=======\\======= lù câng gùa 汝共我 tu ayuda {me} chîo' chīn 借銭 aque me p. ste dinero otro senti do lù câng gùa 汝共我 tu me pedis te pres n chîo' chī 借銭 tado dinero hēng gùa 還我 pagalo =======\\=======\\======= en esta que se sigue tanbien sinifi ca aiudar el arroz es mio y digo lo uenda a otros =======\\=======\\======= 我魚 汝共我 煑

TRANSCRIPT AND TRANSLATION

253

TRANSLATION [1–5]

[lu2 bei2] gua2 hu5 [汝買]我魚 [buy me] my fish  ‘Buy fish for me.’ ‘Buy my fish.’ lu2 kang7 gua2 cu2 puinn7 汝共我煑 ‘You help me to cook rice.’

[6–13]

[14–23]

But it has to be noted that these clauses are equivocal. One possible sense is ‘someone helps me buying fish from others for me’ or ‘the fish is mine and he buys it.’ It also has an equivocal sense with verbs of lending, thus: lu2 kang7 gua2 cioh4 cinn5 汝共我借銭 you help me that money is borrowed to me  ‘Please help me to borrow some money on my behalf.’ Other sense: lu2 kang7 gua2 cioh4 cinn5 hing5 gua2 汝共我借銭還我 you borrowed money from me return it  ‘Please return the money you have borrowed from me.’

[24–27]

In the following sentence it also means ‘help.’ The rice is mine and I say: Sell it to others.

254

ARTE DE LA LENGUA CHIO CHIU

f. 11r, right column [LMS 330v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

lù câng gùa tìoc bì

汝共我 tu ayuda me 糶米 a uender a roz =======\\=======\\======= cang esta particula açe tanbien di f[er]ente sentido con sue. y .ta. que sinifican ablar. gua ta. lu. es desir m{o}rmuxro de ti pero con cang es abla{x ut.} gua. cang. lu ta. di go te lo. ut. =======\\=======\\======= gùa tân lù 我詚汝 yo mormu ro de ti gùa câng 我共 yo lù tân 汝詚 te lo digo gùa câng lù 我共汝 yo a ti sùe' chǐ 說只 digo este kìan sû 件事 negoçio =======\\=======\\======= De suerte que el cang. junto con uer uos. de conprar o bender o prestar o ablar varia notablemente la sinificaçion como esta chio =======\\=======\\======= chio

TRANSCRIPT AND TRANSLATION

255

TRANSLATION [1–4]

[5–12]

lu2 kang7 gua2 thio2 bi2 汝共我糶米 ‘You help me to sell the rice.’ kang7 This particle also has different senses. For example, with sueh8 and tann7, which mean ‘speak’: gua2 tann7 lu2 ‘I gossip about you’; but with kang7 it means ‘talk’, for example, gua2 kang7 lu2 tann7 ‘I tell it to you.’

[13–20]

gua2 tann7 lu2 我詚汝 ‘I gossip about you.’ gua2 kang7 lu2 tann7 我共汝詚 ‘I tell it to you.’ gua2 kang7 lu2 sueh8 ci3 kiann2 su7 我共汝說只件事 I to you tell this affair  ‘I tell you about this.’

[21–25]

Luckily the kang7 is joined to verbs of buying, selling, lending, or talking, varying noticeably in meaning, as this cioh4.

[26]

cioh4

Notes 2 18 5–19

thio2 ≈ thio3 ‘sell (rice)’ tih8 ≈ thih4 ‘iron’ =管

266

ARTE DE LA LENGUA CHIO CHIU

f. 12v, right column [LMS 332v–333r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

cap. vii de los negaçiones en esta lengua ay diuersas ne gaçiones para decir no es liçito açer esto. hurtar. matar. eta. se di çe ansi =======\\=======\\======= 1 mˇ tángc 不可 no es liçito c c táu tè' 偷提 hurtar mˇ tángc 不可 no es liçito pà'c lāng 打人 sacudir a o los honbres =======\\=======\\======= Para deçir. no tengo. como no ten go raçon o dinero et.a se dice. =======\\=======\\======= 2 bō chīn 無銭 no ay din.o bō lì 無理 no ay raçon =======\\=======\\======= para mandar prohiniendo es. bo. distinto del passado. ut. =======\\=======\\======= 3 bò' cǔc no 莫去郎 no bayas bò' tân no 莫詚郎 no hablas =======\\=======\\======= Bue. sinifica a u no pero suele denotar que falta poco. ut. =======\\=======\\======= 4 chìa' pûin 食 comiste

TRANSCRIPT AND TRANSLATION

267

TRANSLATION [1–6]

Chapter 7: On Negations In this language there are different negations for saying ‘it is not right to do this’, such as stealing, killing, etc. These are expressed like this:

[7–12]

(1) m3 thang1 thau1-theh8 不可偷提 ‘It is not allowed to steal.’ m3 thang1 phah8 lang5 不可打人 ‘It is not allowed to hit persons.’

[13–15]

[16–18]

For saying ‘not have’, as ‘I’m not right’, ‘I don’t have money’, etc., you say: (2) bo5 cinn5 無銭 ‘There is no money.’ bo5 li2 無理 ‘There is no reason.’

[19–21]

For giving a prohibition, the negation is boh8, which is distinct from the previous bo5.

[22–24]

(3) boh8 khu3 no 莫去郎 ‘don’t go’ boh8 tann7 no 莫詚郎 ‘don’t talk’

[25–27]

Bue7 means ‘has he or has he not’, but usually indicates ‘hardly’ or ‘a moment ago’.

[28]

(4) ciah8 puinn7 [a5 bue7] 食 [亜未] ‘Have you eaten [or not?’] Notes on p. 269

268

ARTE DE LA LENGUA CHIO CHIU

f. 13r, left column [LMS 333r–333v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

ā bûe

亜未 responde de [x] dos maneras n bûe chìa' pûi 未食 tanbien res ponde con solo el bue bûe 未 =======\\=======\\======= 5 bei sinifica no poder correspon + ei de le .{e}i.+ ques poder. ut. ________\\________\\________ lù ěi cǔc [xxx]ˇc 汝會去 ā běi 亜 běi chô tît' 做得

16 17 18 19 20 21 22 23 24 25 26 27 28 29

o no

auno e comido

aun no come

puesdes hir o no no se pue de açer

=======\\=======\\======= La sexta y ultima es. m. que es la ordinaria para todos los de mas modos de negar. =======\\=======\\======= lù sǐn 汝信 c{r}es lo ā mˇ sǐn 亜不信 o no =======\\=======\\=======

cap. viii del modo de

preguntar y rresponder en esta lengua se pregunta dife rentemente que la española en la espa[ñola] solo se diçe yçiste [esto]

269

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–9]

[(4) ciah8 puinn7] a5 bue7 [食 [‘Have you eaten ] or not?’

]亜未

There are two manners of responding: bue7 ciah8 puinn7 未食 ‘I have not eaten yet.’ You can also answer with bue7 only: bue7 未 ‘not yet eaten.’ [10–12]

[13–17]

(5) Bei3 means ‘cannot’ and corresponds to ei3 which is ‘can’, thus: lu2 ei3 khu3 a5 bei3 汝會去亜 ‘can you go or not’ bei3 co7 tit4 做得 ‘not able to do’

[18–21]

[22–24]

[25–29]

The sixth and last one is m3, which is the normal one for all other modes of negation. lu2 sin3 a5 m3 sin3 汝信亜不信 ‘Do you believe it or not?’

Chapter 8: On the Mode of Asking and Responding The way of asking questions in this language is different from Spanish. In Spanish you only say yçiste esto ‘did you do this’,

Notes f. 12v, right column 19 = Notes on p. 273

270

ARTE DE LA LENGUA CHIO CHIU

f. 13r, right column [LMS 333v–334r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

el sanglei dice yçiste esto o no fuiste o no el rresponder es de dos maneras la una es con toda [la] oraçion deziendo ya fui: ya. reçe. quiero hir a manila et.a o con solo el ueruo quiero. o fui et.a -ut.=======\\=======\\======= lù ǎy 汝 quier[es] chìa' pûin 食 comer ā. ˇm. 亜不 o no responde ǎy chìa' 食 quiero pûin comer 2o modo de responder ǎy quiero =======\\=======\\======= si pregunta de perterito o [rrespon] de con toda la oraçion o con solo liau la. que es ordinario modo de responder. ut. =======\\=======\\======= lù liâm 汝 tu recaste kéng liàu 了 ya ā bûe 亜未 o no

TRANSCRIPT AND TRANSLATION

271

TRANSLATION [1–7]

[8–17]

the Sangley says ‘did you do this or not’, ‘were you or not’. There are two manners of responding. One is by using the whole former clause ‘have already been’, ‘have already prayed’, ‘want to go to Manila’, etc.; or only with the verb for ‘want’, ‘was’, etc. lu2 ai3 ciah8 puinn7 a5 m3 汝 食 亜不 you want to eat or not  ‘Do you want to eat?’ Response: ai3 ciah8 puinn7 食 ‘I want to eat.’ Second mode of responding: ai3

[18–22]

[23–25]

‘I want.’

If a question in the preterite is asked, the response is either with the whole clause or only with liau2 la2, which is the ordinary way of responding, for example: lu2 liam7 king1 liau2 a5 bue7 汝 了亜未 have you prayed already or not  ‘Have you already prayed?’

Notes f. 12v, right column 25ff. EMH bue7 corresponds to Amoy Hokkien be7 ‘yet’. According to Bodman, ‘[w]hen it occurs finally in questions, it usually has the unstressed be. Such questions mean “Has such-and-such happened yet?”. They are inquires as to whether a certain action has been completed or not’ (1987:46, original emphasis). Notes on p. 273

272

ARTE DE LA LENGUA CHIO CHIU

13v, left column [LMS 334r–334v] 1 2 3

liâm kéng liàu

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

liàu là

responde 經 了 2o modo de responder 了

ya re çe

ya acaue de rreçar =======\\=======\\======= estos dos ueruos ei y bei ques poder i no poder sienpre andan juntos en las preguntas y res pondeçe con el uno como puede se açer o no se puede açer. =======\\=======\\======= čhi sû 只事 este negoçio ěichôtît' 會做得 puedes açer a běi 亜 o no responde ěichôtît' 會做得 puedese açer 2o modo RP. ěi 會 puedese açer =======\\=======\\=======

TRANSCRIPT AND TRANSLATION

273

TRANSLATION [1–8]

Response: liam7 king1 liau2 經了 ‘I prayed.’ Second mode of responding: liau2 la2 了 ‘I have already finished praying.’

[9–14]

[15–17]

[18–20]

The two verbs ei3 and bei3, meaning ‘can’ and ‘cannot’, always occur together in questions. You answer with one, like ‘can do it’ or ‘cannot do it.’ ci3 su7 ei3 co7 tit4 a5 bei3 只事會做得亜 this affair can it be done or not  ‘Can it be done?’ Response: ei3 co7 tit4 會做得 ‘It can be done.’

[21–25]

Second mode of responding: ei3 會 ‘can be done’

Notes f. 13r, left column 13 The word is a misspelling, the LMS has . f. 13r, right column 8 =愛 =愛 12, 16

274

ARTE DE LA LENGUA CHIO CHIU

f. 13v, right column [LMS 334v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Ase de notar que se a de rrespon der con la negaçion que se pre gunta y o no con otra ut. =======\\=======\\======= lù ù chīn 汝有銭 tu tienes dinero ābō 亜無 o no tienes responde bō 無 no tengo =======\\=======\\======= y no {res}pondera con m. o bue. o. bei. y si por {b}ei. si preguta respondera por {b}ei eta. algunas ueçes se pregunta sin poner negaçion y entonçes se rres ponde con la açion que açe. ut. =======\\=======\\======= lù chá: bôu 汝查暮 tu aier xìt' chô 日做 hi{s}iste sǐmì' mì' 是乜物 que cossa responde sìa xī 字 escreui lù ǎy 汝 a do quie res

TRANSCRIPT AND TRANSLATION

275

TRANSLATION [1–4]

[5–10]

It must be noted that one has to respond with the negation without adding anything. lu2 u2 cinn5 a5 bo5 汝有銭亜無 do you have money or do you not have  ‘Do you have money?’ Response: bo5 無 ‘I don’t have [money].’

[11–17]

[18–20]

[21–24]

And for ‘no’ the answer is m3 or bue or bei3, and for ‘yes’ ei3. In case of a question, answer with ei3, etc. Sometimes a question without a negation is being asked. Then you answer with the action which is performed. lu2 ca:1-bou7-jit8 co7 si3-mih8 mih8 汝查暮日做是乜物 you yesterday did what  ‘What did you do yesterday?’ Response: sia2 ji5 字 ‘I wrote.’ lu2 ai3 [te3-loh8 khu3] 汝愛 [ 落去] ‘Where do you want [to go]?’

Note 12

The spelling instead of is according to the original.

276

ARTE DE LA LENGUA CHIO CHIU

f. 14r, left column [LMS 334v–335r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

tě lò' cǔc

落去 hir responde ǎi cǔc 去 quiero hir minilà' 民希臘 a manila =======\\=======\\=======

cap. ix del modo de conponer

De lo que esta dicho en los pre[{ceden}] tes capitulos constara el modo de conposiçion de esta lengua – La oraçion se açe poniendo lo pri mero la perssona que açe y des pues el ueruo y luego la persona q. padese y despues las demas par ticulas segun las rreglas dadas =======\\=======\\======= guà liâm 我 yo ya kéng liàu 經了 reçe gùa cǔc minilà'

我去 民希臘

yo fui a manila

guà chīn 我銭 [pre]stete chîo' l{ù} 借汝 mi dinero =======\\=======\\=======

TRANSCRIPT AND TRANSLATION

277

TRANSLATION [1]

[2–5]

[lu2 ai3] te3-loh8 khu3 [汝 ] 落去 ‘[Where do you want] to go?’ Response: ai3 khu3 mi-ni-lah8 去民希臘 ‘I want to go to Manila.’

[6–15]

Chapter 9: On the Mode of Composing From what has been said in the preceding chapters, the mode of composition in this language will become evident. The clause is made by placing first the person doing the act and then the verb and then the person who suffers and then the other particles according to the given rules.

[16–22]

gua2 liam7 king1 liau2 我 經了 ‘I have already prayed.’ gua2 khu3 mi-ni-lah8 我去民希臘 ‘I went to Manila.’ gua2 cinn5 cioh4 lu2 我銭借汝 ‘I lent you my money.’

278

ARTE DE LA LENGUA CHIO CHIU

f. 14r, right column [LMS 335r–335v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

en ueruos de dar e tomar lo que se da o se toma se puede anteponer posponer al ueruo aunque con mas elegançia {se antepone ut} =======\\=======\\======= lù tôuc 汝度 tu da pêló 呶 a pedro čhi nò gùe 只二个銭 estos dos chīn tostones čhi nò 只二 estos dos n gùe chī 个銭 tostones lù tôuc 汝度 tu da los pêló 呶 a pedro =======\\=======\\======= Por enperatiuo se dize destos dos modos siguientes. ut. =======\\=======\\======= čhi nò 只二 estos dos gùe chīn 个銭 tostones tè'c lāy no 提來郎 toma los y {t}rai los

TRANSCRIPT AND TRANSLATION

279

TRANSLATION [1–6]

[7–15]

As for verbs of giving and taking, what is given or what is taken can be placed before or after the verb, although it is more elegant to place it before the verb. lu2 thou7 pe3-lo1 ci3 no2 ge2 cinn5 汝度 呶只二个銭 ‘You give Pedro these two coins.’ ci3 no2 ge2 cinn5 lu2 thou7 pe3-lo1 只二个銭汝度 呶 ‘These two coins—you give them to Pedro.’

[16–18] [19–22]

The imperative is expressed by the following two modes. ci3 no2 ge2 cinn5 theh8 lai5 no 只二个銭提來郎 these two coins take them and bring them over  ‘These two coins—take them over!’

Notes 1–14 8

This section is discussed on p. 198. =弊

280

ARTE DE LA LENGUA CHIO CHIU

f. 14v, left column [LMS 335v–336r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

otro modo c lù tè' 汝提 tu toma chǐ nò 只二 estos dos n gùe chī 个銭 tostones lāy no 來郎 trai los =======\\=======\\======= este ueruo. tè'c 提 es muy universal para lle uar o traher o tomar o sacar an da de ordinario junto con los uer uos de mouimiento. lāy y. cūc. ut. pero açe de poner primero el uer uo te. y despues el ueruo de mo uimiento. cu. o lay. =======\\=======\\======= [x]hù ch{ì}g' 許一 aquel pùn chêg' 本冊 libro c tè' lāi 提來 trahe lo chǐ chèg'

只一

este

TRANSCRIPT AND TRANSLATION

281

TRANSLATION [1–7]

Other mode: lu2 theh8 ci3 no2 ge2 cinn5 lai5 no 汝提只二个銭來郎 you take these two coins bring them over  ‘You take these two coins over !’

[8–17]

[18–21]

This verb theh8 提 is very universal for taking, bringing, carrying, drawing. It normally occurs with verbs of movement, for example, lai5 and khu3. You must first place the verb theh8 and then the verb of movement, khu3 or lai5. hu2 cik8 pun2 cik4 theh8 lai5 許一本冊提來 ‘That book—bring it over!’ ci3 cik8 [ge2 cinn5 theh8 khu3] 只一[个銭提去] ‘This [coin—take it over!’]

Note 8–17

In Amoy Hokkien, according to Bodman, khi3~khu3 ‘go to’ and lai5 ‘come’ ‘are used with verbs of motion to express direction away from or towards the speaker respectively’ (1987:271).

282

ARTE DE LA LENGUA CHIO CHIU

f. 14v, right column [LMS 336r] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 margin under 23

gùe chīn 个銭 c c tè' cǔ [x] 提去

toston lleua lo

=======\\=======\\======= fuera de para lleuar traher me ter. sacar. para quien sirue. te ay para otros particulares mo dos de traher otros particulares ueruos. ut. =======\\=======\\=======  pòc lāi 抱來 traher en bracos chèg' gùe 一个 este kǐnnìa 蕳仔 mucha cho =======\\=======\\======= el instrumento se pone en la o raçion antes del beruo. ut. =======\\=======\\======= huǎlāng 化人筆 con pluma pìt' despañol sìa xī 字 escreui ________\\________\\________  xhìt' gùe kǐnnìa pòc lāy es mejor

TRANSCRIPT AND TRANSLATION

283

TRANSLATION [1–3]

[8–13]

[15–20]

[21–23]

[ci3 cik8] ge2 cinn5 theh8 khu3 [只一]个銭提去 [‘This] coin—take it over!’ Except for ‘take’, ‘bring’, ‘put’, and ‘get’ expressed by theh8, there are other particular verbs for other particular modes of taking, for example:  pho2 lai5 cik8 ge2 kin3-nia2 抱來一个蕳仔 hold in one’s arm this child  ‘hold this child in one’s arms’ In a clause, the instrument is placed before the verb, for example:

[24–27]

hua3 lang5 pit8 sia2 ji5 化人筆 字 with a Spanish quill I wrote  ‘I wrote with a Spanish quill.’

margin

 hit8 ge2 kin3-nia2 pho2 lai5 is better.

Notes 15, etc. margin

pho2 ≈ pho7 ‘embrace’ It is better to say: hit8 ge2 kin3-nia2 pho2 lai5 [that piece child embrace come ‘Bring this child over.’] gok4-hu5 ≈ gok8-hu5 ‘crocodile’ The character is not attested in Chinese dictionaries. In printed editions of Hokkien stage plays, the classifier khu1 is written with the character 坵 (Chinfa Lien, p. c.).

292

ARTE DE LA LENGUA CHIO CHIU

f. 23r, right column [LMS 336v] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

cǔ

句

sì cǔ 四句話 o{ǎ} pòan cǔ 半句話 o{ ǎ }

para pala bras 4 pala bras media pa labra

4

5

cùnc

綑

p.a fardos o enbolto rios

sì cùmc pǒu sì cùnc toân sì cùnc līn sì cùnc pànc sý

四綑布

4 fardos de mantas 4 fardos de de d{a}mascos 4 fardos de chamelote 4 fardos de damas quillos

cūn

群

四綑緞 四綑綾 四綑紡系

p.a manada conpanias tropas

6

TRANSCRIPT AND TRANSLATION

293

TRANSLATION [1–6]

(4) ku3 句: for sentences si2 ku3 ua3 四句話 ‘four sentences’ puann2 ku3 ua3 半句話 ‘half a sentence’

[8–20]

(5) khun2 綑: for bales or bundles si2 khun2 pou3 四綑布 ‘four bales of cloth’ si2 khun2 tuan7 四綑緞 ‘four bales of damask’ si2 khun2 lin5 四綑綾 ‘four bales of camlet’ si2 khun2 phan2-si1 四綑紡系 ‘four bales of damassin’

[22–24]

(6) kun5 群: for herds, companies, and troops

Notes 8, etc.

10 12 13, etc.

It is not clear whether is used for ‘bundle’ in general or as a specific unit of measurement. In a footnote to Antonio Álvarez de Abreu’s account of commerce between the Philippines and New Spain in the eighteenth century, Blair and Robertson write that ‘[f]ormerly the fardo was 1 1/3 varas long, 3/4 vara high, and 3/4 vara less one pulgada (nearly one English inch) wide; but for a long time previous to 1726 the bale of this size had not been used’ (BR 45:55). On vara, see also p. 347. This is the last line of the LMS. is a misspelling of . Douglas (1873:378): poo3 ‘cotton or linen cloth’. The Spanish word manta means ‘blanket’, in Mexico also ‘coarse cotton cloth’. Fang points out that in the Manila trade of the seventeenth century, manta is used generically for cotton cloth (2006:32f.).

294

ARTE DE LA LENGUA CHIO CHIU

f. 23v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

sì cūn pánc sì cūn chiàu sì cūn gū sì cūn hū sì cūn lāng châi

sì châi [ſs]óa sì châi chìo' sì châi húe sì châi tōuc

四群蜂 4 enxan bres de a uexar 四群鳥 4 manadas de paxaros 四群牛 4 manadas de carabas 四群魚 4 manados de pescados 四群人 4 companias de honbres 載 para car gas por a gua

7

四載沙 quatro cha[n]pa nes de arena 四載石 4 chanpanes de piedra 四載灰 4 chanpanes de cal 四載泥 4 chanpanes de tierra

chàm 站 p.a jornadas sì chàm 四站路 4 xornadas lôu de [a 6.] le guas

8

TRANSCRIPT AND TRANSLATION

295

TRANSLATION [1–11]

si2 kun5 phan1 四群蜂 ‘four swarms of bees’ si2 kun5 ciau2 四群鳥 ‘four flocks of birds’ si2 kun5 gu5 四群牛 ‘four herds of water buffalos’ si2 kun5 hu5 四群魚 ‘four shoals of fish’ si2 kun5 lang5 四群人 ‘four groups of men’

[12–23]

(7) cai7 載: for loads, for water si2 cai7 sua1 四載沙 ‘four loads of sand’ si2 cai7 cioh8 四載石 ‘four loads of stones’ si2 cai7 hue1 四載灰 ‘four loads of lime’ si2 cai7 thou5 四載泥 ‘four loads of earth’

[25–28]

(8) cam2 站: for journeys si2 cam2 lou7 四站路 ‘four journeys of six leagues’ Notes on p. 297

296

ARTE DE LA LENGUA CHIO CHIU

f. 23v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

pa raçones parrafos

châr' sì châr' tǒ lỳ

四 道理

una raçon[x] pa ordenes de piedras co mo en pared

chān

sì chān chìo'

四 石

sì chān túic

四 梯

chāng

欉

sì chāng chǐuc

四欉樹

sì chāng ſsùn sì chān chǎic

四欉笋 四欉菜

9

10

4 ordenes de piedras una ençima de otra cuatro es calones pa arboles or 11 talicas que su prinçipal par te no es la caue ça para ellas est gue. 4 arboles para renue bos 4 renuebos de arbol quatro plan tas de ortali ca

TRANSCRIPT AND TRANSLATION

297

TRANSLATION [2–6]

(9) cat4 : for reasons, paragraphs si2 cat4 to3-li2 四 道理 ‘one reason’

[8–16]

(10) can5 : for layers of stones, as of a wall si2 can5 cioh8 四 石 ‘four layers of stones, one above the other’ si2 can5 thui1 四 梯 ‘four steps’

[18–31]

(11) cang5 欉 : for trees, vegetables, the principal part of which is not its top, for those it is ge2. si2 cang5 chiu2 四欉樹 ‘four trees’ for sprouts si2 cang5 sun2 四欉笋 ‘four shoots of trees’ si2 cang5 chai3 四欉菜 ‘four vegetable plants’

Notes f. 23v, left column 7 = ‘water buffalos’; Hokkien gu5 means ‘cow, ox’. As carabaos were commonly used in farming in the Philippines, the word is used as a generic term here. 25–28 cam2 ≈ cam7 ‘a stage on a journey’ (Douglas, 1873:575). The Spanish translation legua usually refers to a distance traveled in one hour (DLE, 2001). According to Clarke, a Spanish legua equals 6.78 km, an Argentinean legua is 0.52 km, and a Portuguese legoa corresponds to 6.17 km (1891:91). As this example shows, the Spanish units of measurement do not exactly equal the Chinese units. Moreover, due to regional and historiNotes continued on p. 299

298

ARTE DE LA LENGUA CHIO CHIU

f. 24r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

p.a nauios y para anima les grandes

chîa'

sì chîa' gū sì chîa' chūn sì chîa' tú sì chîa' bè

四 牛 四 船 四 四 馬

sì chín bỳ

四 米

四 厝

13

quarto gan tas de arroz p.a ringle ras de cosas en orden

chûe

sì chue chùc

quatro ca rabaos quatro na uios quatro puer cos quatro ca ballos p.a gantas

chín

12

quatro or denes de ca sas

14

TRANSCRIPT AND TRANSLATION

299

TRANSLATION [2–13]

(12) ciah4 : for ships and big animals si2 ciah4 gu5 四 牛 ‘four water buffalos’ si2 ciah4 cun5 四 船 ‘four ships’ si2 ciah4 tu1 四 ‘four pigs’ si2 ciah4 be2 四 馬 ‘four horses’

[15–18]

(13) cin1

: for ganta

si2 cin1 bi2 四 米 ‘four ganta of rice’

[20–26]

(14) cue7

: for rows of arranged things

si2 cue7 chu2 四 厝 ‘four rows of houses’ Notes f. 23v, left column 25–28 cal differences, one and the same Chinese or Spanish term can actually represent various different contents (cf. Morse, 1908:172; Wilkinson, 2000: 239). f. 23v, right column 5 should be ‘four reasons’ (cf. EMH si2 ‘four’) 6 to3-li2 ≈ to7-li2 8 =層 26 chiu3 ≈ chiu7 ‘tree’ 29 is a misspelling of . Notes continued on p. 301

300

ARTE DE LA LENGUA CHIO CHIU

f. 24r, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

sì chue tiǎm sì chui pāng

四 店 四 房

4 ordenes de tiendas 4 ordenes de aposen tos

chùicn

串

pa ensar tados

sì chùicn 四串 珠 chín chú sì chùicn 四串魚 hū

4 sartas de alxofar 4 sartas de pezcado

chǔic

口

pa bocados de arros agua et.a

sì chǔic pûin

四口

4 bocadose de morisq[ue]ta

hǔ

付

sì hǔ pōan

四付盤

pa atados de escudillas o platos de a 10 platos 4 atados de platos cada uno de diez platos

15

16

17

TRANSCRIPT AND TRANSLATION

301

TRANSLATION [1–5]

si2 cue7 tiam3 四 店 ‘four rows of shops’ si2 cue7 pang5 四 房 ‘four rows of rooms’

[7–13]

(15) chuinn2 串: for strings si2 chuinn2 cin1-cu1 四串 珠 ‘four strings of pearls’ si2 chuinn2 hu5 四串魚 ‘four strings of fish’

[15–20]

(16) chui3 口: for mouthfuls of rice, water, etc. si2 chui3 puinn7 四口 ‘four mouthfuls of rice’

[22–29]

(17) hu3 付: for bundles of bowls and plates of up to ten plates si2 hu3 puann5 四付盤 ‘four bundles of plates each one with ten plates’

Notes f. 24r, left column =隻 2, etc. 10 =猪 15, 17 =升 15–17 According to Morse (1908:172), one cin1 (Mnd.: shēng) equals a tenth of a dǒu 斗 (= 629 cubic inches or 10.31 liters). The Spanish word ganta is Philippine Spanish, equaling three liters (BR 3:253; DLE, 2001). 20, 24 The character character is neither attested in Chinese dictionaries, nor in other Hokkien sources. 24 chu2 ≈ chu3 f. 24r, right column =真 10

302

ARTE DE LA LENGUA CHIO CHIU

f. 24v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

sỳ hǔ òan

四付碗

4 atados de escudillas

gùe

箇

chèg' gùe 一箇日 xìt' sỳ gùe 四箇鳥 chiàu

este es comun 18 para muchas cosas por la esperiencia seuera p.a dine ro dias sema nas et.a 4 tostones de dineros {un} quatro me ses {un mes} quatro dias {un dia} quatro pa xaros eta.

ě:

下

para açotes

sỳ ě:

四下

pà'c pê:' ě: pà'c í cùi' ě:

打百下

quatro aço tes da le çien a çotes da le quantos a çotes

sì gùe 四箇銭 chīn chèg' gùe 一箇月 gùe'

打伊 下

19

TRANSCRIPT AND TRANSLATION

303

TRANSLATION [1–2]

[4–20]

si2 hu3 uann2 四付碗 ‘four bundles of bowls’ (18) ge2 箇: this is common for many things, for severe experience, for money, days, weeks, etc. si2 ge2 cinn5 四箇銭 ‘four coins of money’ cik8 ge2 geh8 一箇月 ‘one month’ cik8 ge2 jit8 一箇日 ‘one day’ si2 ge2 ciau2 四箇鳥 ‘four birds, etc.’

[22–30]

(19) ee3 下: for blows si2 ee3 四下 ‘four blows’ phah8 peeh4 ee3 打百下 ‘give him a hundred blows’ phah8 i1 kuih8 ee3 打伊 下 ‘give him some blows’

Note 29

= 幾 phit8 ≈ phit4 ‘cl.: pieces of cloth’ The spelling is according to the original. pho2 ≈ po7 ‘clump, tuft’ The tone of is not indicated. The transcription in the Dictionario is phou3, which was later changed to phou7; Douglas (1873: 400): phou3 ‘a measure of ten lí; on the great roads, esp. in Cn. [Chin-chew = Quánzhōu], a league of about three miles; in the country parts, esp. in C. [Chang-chew = Zhāngzhōu], often four miles or more’. The Spanish translation also appears on f. 23v, where it translates EMH cam2 which is different from phou3.

320

ARTE DE LA LENGUA CHIO CHIU

f. 26v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

pùn

本

p.a libros

sỳ pùn chê'c

四本冊

4 libros

sỳn

扇

p.a puertas 42 y uentanas

sỳ sỳn mūy sì sỳn tángc yà

四扇門

4 puertas

四扇 窓仔

4 ventanas

siáng

双

para pares

sì siáng {ēi} sì siáng bùe'

四双鞋

4 pares de zapatos 4 pares de calças

四双

41

43

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–4]

(41) pun2 本: for books si2 pun2 cheh4 四本冊 ‘four books’

[6–12]

(42) sinn2 扇: for doors and windows si2 sinn2 mui5 四扇門 ‘four doors’ si2 sinn2 thang1-ia2 四扇窓仔 ‘four windows’

[14–19]

(43) siang1 双: for pairs si2 siang1 ei5 四双鞋 ‘four pairs of shoes’ si2 siang1 bueh8 四双 ‘four pairs of stockings’

Notes 12 18

thang1-ia2 ≈ thang1-a2 =襪

321

322

ARTE DE LA LENGUA CHIO CHIU

f. 27r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

pa bracas

siām sì siām toân sì siām pǒu sì siām lēng sì siām ſsô' sì siām tǒa

四

tiàm

點

sỳ tiàm chióng sì tiàm chécn

四點鍾 4 oras

tión

張

四 布 四 綾 四 四

44

4 braças de damasco 4 bracas de manta 4 braças de razillo 4 brasas de soga 4 braças de liston p.a estrellas para oras

45

四點星 4 estrellas p.a pliegos 46 de papel p.a instrumen tos muçicos p.a camas mesas.

TRANSCRIPT AND TRANSLATION

323

TRANSLATION [1–12]

(44) siam5

: for brazas

si2 siam5 tuan7 四 ‘four brazas of damask’ si2 siam5 pou3 四 布 ‘four brazas of cloth’ si2 siam5 ling5 四 綾 ‘four brazas of satin’ si2 siam5 soh4 四 ‘four brazas of rope’ si2 siam5 tua3 四 ‘four brazas of ribbon’

[14–20]

(45) tiam2 點: for stars, for hours si2 tiam2 ciong1 四點鍾 ‘four hours’ si2 tiam2 chenn1 四點星 ‘four stars’

[22–27]

(46) tionn1 張: for sheets of paper, for musical instruments, for beds and tables

Notes 1, etc. =撏 1ff. Douglas does not indicate a unit of measurement for Hokkien siam5 ‘to embrace in the arms’ (1873:427). The Spanish measure braza equals 1.6718 meters according to DLE (2001) and one English fathom (1.8 meters or 6 feet), according to BR (33:308). 9 =索 11 =帶

324

ARTE DE LA LENGUA CHIO CHIU

f. 27r, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

sỳ tión kīmc sỳ tión pîoc sì tión péyc sì tión chòa sì tión xī sì tión ǎn sì tión tô' sì tión chūncnn

四張琴 四張票 四張

4 vigue las 4 cedulas o liçençias 4 cartas

四張案

4 pliegos de papel 4 pliegos escritos 4 proçesos

四張

4 messas

四張床

4 camas

tǔin

断

p.a pedaços 47 de cossa que es tube entero

sì tǔin chāc sì tǔin hū

四断

sì tiǒn bâ'

四断肉

quatro tro cos de leña quatro pe daços de pes cado quatro peda ços de carne

四張帋 四張字

四断魚

325

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–16]

si2 tionn1 khim5 四張琴 ‘four lutes’ si2 tionn1 phio7 四張票 ‘four documents or licenses’ si2 tionn1 phei1 四張 ‘four letters’ si2 tionn1 cua2 四張帋 ‘four sheets of paper’ si2 tionn1 ji5 四張字 ‘four written sheets’ si2 tionn1 an3 四張案 ‘four lawsuits’ si2 tionn1 toh4 四張 ‘four tables’ si2 tionn1 chng5 四張床 ‘four beds’

[18–28]

(47) tuinn3 断: for pieces of a thing which was complete si2 tuinn3 cha5 四断 ‘four pieces of firewood’ si2 tuinn3 hu5 四断魚 ‘four pieces of fish’ si2 tuinn3 bah4 四断肉 ‘four pieces of meat’

Notes on p. 327

326

ARTE DE LA LENGUA CHIO CHIU

f. 27v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

tûin

para comi das

sỳ tûin pùi'

四

quatro co midas

sýn

身

cuerpo

sì sýn ún chīon

四身 衣裳

quatro vestidos en teros

tân

担

p.a cargas de uno solo p.a picos

sì tân ſsóa sì tânn chùi sì tânn tōuc sì tânn tūngcnn

四担沙 quatro car gas de arena 四担水 quatro car gas de agua 四担泥 quatro car gas de tierra 四担糖 quatro car gas de chan cacas

48

49

50

TRANSCRIPT AND TRANSLATION

327

TRANSLATION [1–5]

(48) tuinn7

: for meals

si2 tuinn7 puih8 四 ‘four meals’

[7–11]

(49) sin1 身: body si2 sin1 un1-cionn5 四身衣裳 ‘four complete dresses’

[13–25]

(50) tann7 担: for a single person’s load, for picos si2 tann7 sua1 四担沙 ‘four loads of sand’ si2 tann7 cui2 四担水 ‘four loads of water’ si2 tann7 thou5 四担泥 ‘four loads of earth’ si2 tann7 thng5 四担糖 ‘four loads of brown sugar’

Notes f. 27r, right column 4 phio7 ≈ phio3 ‘warrant’ = 批; the character 批 is widely attested in early written Hokkien for 5 phue1 ‘letter’ (cf. Klöter, 2005:265). 13 =槕 18, etc. tuinn3 ≈ tuinn7 ‘piece’ 27 is presumably a misspelling of tuinn3 ≈ tuinn7 ‘piece’ f. 27v, left column = 頓, tuinn7 ≈ tuinn3 ‘cl.: meals’ 1, 4 Notes continued on p. 329

328

ARTE DE LA LENGUA CHIO CHIU

f. 27v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

sì tân

四担

tūngcnn ſsúngnn 糖霜

quatro car gas o picos de açucar piedra

tāuc

頭

sì tāuc bỳ

四頭米 4 medias de arroz de a[xxx] gantas 四頭 quatro me dias de ari na 四頭麥 quatro me dias de tri go

sì tāuc mˇy sì tāuc bè' tě

p.a medidas como anegas medias cele mines

para pie dras tablas eta.

sì tě chìo' sì tě páng

四 石 quatro piedras 四 枋 quatro tablas

tēng

重

p.a doble ces

51

52

53

TRANSCRIPT AND TRANSLATION

329

TRANSLATION [1–4]

[6–18]

si2 tann7 thng5-sng1 四担糖霜 ‘four loads or picos of sugar’ (51) thau5 頭: for measures such as fanegas, medias, celemines si2 thau5 bi2 四頭米 ‘four medias of rice’ si2 thau5 mi3 四頭 ‘four medias of flour’ si2 thau5 beh8 四頭麥 ‘four medias of wheat’

[20–27]

(52) te3

: for stones, planks, etc.

si2 te3 cioh8 四 石 ‘four stones’ si2 te3 pang1 四 枋 ‘four planks’

[29–30]

(53) ting5 重: for folds

Notes f. 27v, left column 5 phit8 ≈ phit4 ‘cl.: horses’ 21, etc. co3 ≈ co7 ‘cl.: house’ f. 28v, left column 6 ua3 ≈ ua7 ‘paint, draw’ 12 =薬~藥 26 mia2 ≈ mia5 ‘cl.: person’

336

ARTE DE LA LENGUA CHIO CHIU

f. 28v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

kéng

間

p.a aposen tos

sì k{í}ng chùc

四間厝

quatro a posentos

kîp'

級

p.a escalo nes

chìt' kîp'

一級

nò kîp'

二級

un esca lon dos esca lones

tèng

頂

p.a sonbre ros suios

chèg' tèng kín chìt' tèng bô

一頂巾

un sonbre ro suio un sonbrero

一頂帽

62

63

64

TRANSCRIPT AND TRANSLATION

337

TRANSLATION [2–6]

(62) king1 間: for rooms si2 king1 chu2 四間厝 ‘four rooms’

[8–14]

(63) kip4 級: for steps cit8 kip4 一級 ‘one step’ no2 kip4 二級 ‘two steps’

[16–22]

(64) ting2 頂: for their hats cik8 ting2 kin1 一頂巾 ‘one of their hats’ cit8 ting2 bo7 一頂帽 ‘a hat’

Note 16–17

The translation ‘their hats’ presumably refers to the hats of the Sangleys, which differed in shape from the hats worn by the Spaniards. The literal meaning of kin1 巾 is ‘napkin, towel‘.

338

ARTE DE LA LENGUA CHIO CHIU

f. 29r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

pôu

部

chìt' [pôu chê'c]

一部冊

[xxx] [tāy]

pa obras 65 de {li}bros como las de santo tomas aunque se an munchos uer pos una obra como los del co chu

[chìt'] tāy […]

一 戯

para tabla dos un tabla do de come dias

[sê]ng

乘

p.a sillas de acarreto

sỳ sêng kiô

四乘簥

4 sillas de acarreto co mo de mug.rs

tēy

第

pa ordenes contando primero se gundo et.a

tēy ît' pùn

第一本

tēy ît' lāng

第一人

el primer libro por orden el primer honbre

66

67

68

TRANSCRIPT AND TRANSLATION

339

TRANSLATION [1–9]

(65) pou7 部: for written works, books, like those of Santo Tomas although ?. cit8 pou7 cheh4 一部冊 ‘single works, as those of co chu’

[11–15]

(66) tai5 : for stages cit8 tai5 hi3 一 戯 ‘a comedy’

[17–22]

(67) sing7 乘: for sedan chairs si2 sing7 kio7 四乘簥 ‘four sedan chairs, as for women’

[24–33]

(68) tei5 第: for counted sequences, first, second, etc. tei5 it4 pun2 第一本 ‘the first book of a series’ tei5 it4 lang5 第一人 ‘the first man’

Notes 1–6

9 11–13

Douglas (1873:379): poo7 ‘a set, as of books’. Santo Tomas refers to the University of Santo Tomas in Manila, established in 1611 by the Dominican friar Miguel de Benavides (1550–1605) (for details, see Villarroel, 2005). The additional explanation after is unclear; the word in line 6– 7 may be read as ‘words’ or ‘verses’, neither of which makes sense. The meaning of is unclear, one possible reading is koo2 cu1 ‘ancient books’. =臺 Notes continued on p. 341

340

ARTE DE LA LENGUA CHIO CHIU

f. 29r, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

âg'

握

para mana das atadas

69

chìt' âg' chángc

一握葱

un atado de cebollas

būe

枚

p.a cossas re 70 dondas {no}

chèg' būe 一枚 chò

4 una ciru ela pasa

tôa

para ra çones

chèg' tôa 一 tò lỳ 道理

una raçon

toán

para nego 72 çios o cossas

端

71

TRANSCRIPT AND TRANSLATION

341

TRANSLATION [2–6]

(69) ak4 握: for handfuls, bundles cit8 ak4 chang1 一握葱 ‘a handful of onions’

[8–13]

(70) bue5 枚: for round things cik8 bue5 co2 一枚 ‘a dried plum’

[15–19]

(71) tua7

: for reasons

cik8 tua7 to2-li2 一 道理 ‘one reason’

[21–22]

(72) tuan1 端: for affairs or things

Notes f. 29r, left column 14 The EMH transcription is illegible due to paper damage. The transcription hi3 in the translation is the modern Hokkien reading. 24 tei5 ≈ te7 prefix for ordinal numbers f. 29r, right column =棗 12 15, 18 =段

342

ARTE DE LA LENGUA CHIO CHIU

f. 29v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

chèg' toán sû

一端事

nìon

una cossa tael

sì nìon guīn sì nìon ſsôan

四 艮

chīn

銭

maizes

sỳ chˉyn guīn sỳ chˉyn xiōng

四銭艮

4 mases de plata 4 mases de seda

hún

分

四 線

四銭絨

73

4 taes de plata 4 taes de hilo

conde rin

73

74

TRANSCRIPT AND TRANSLATION

343

TRANSLATION [2–3]

[5–10]

cik8 tuan1 su7 一端事 ‘a thing’ (73) nio2

: tael

si2 nio2 gin5 四 艮 ‘four taels of silver’ si2 nio2 suann7 四 線 ‘four taels of thread’

[12–17]

(73) cinn5 銭: maces si2 cinn5 gin5 四銭艮 ‘four maces of silver’ si2 cinn5 jiong5 四銭絨 ‘four maces of silk’

[18–20]

(74) hun1 分: candareen

Notes 5 5, etc.

10 12

=兩 Sixteen nio2 (Mnd. liǎng) ‘tael’ equals one kin1 (Mnd. jīn) (Morse, 1908: 171). Morse explains that in ‘weighing the precious metals […] the tael is the heaviest unit, and it has decimal subdivisions down to the one thousandmillion-millionth […] part of a tael, those in daily use being the following: 10 li (cash) = 1 fên (candarin); 10 fên = 10 tsien [qián, EMH: cinn5] (mace); 10 tsien = 1 liang [EMH niu2] (tael)’ (1908:150). English candereen ~ candarin corresponds to , and maces to ~ . Clarke lists eight regional equivalents of tael, ranging from 28.4 g to 60.49 g, and five regional equivalents of mace, ranging from 1.85 g to 3.92 g (1891:110, 114–115). suann7 ≈ suann3 ‘thread’ The erroneous repetition of number 73 is according to the original.

344

ARTE DE LA LENGUA CHIO CHIU

f. 29v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

sì hún guīn

四分艮

sỳ hún xiōng

四分絨

lí

尼

[…]

sì lí kím sỳ lý guīn

四尼金

4 cas[…] de [oro] 4 c[…] de [plata]

四尼艮

tùngcn

4 conde rinas de plata 4 [co]nderi [nas] de seda p[…]xa

otra m[edi] da de diez ter çias

sì tùngcn pǒu

四 布

sì tùngcn têy

四 地

chîo'c

尺

75

76

4 medidas destas de man tas 4 medidas de tierra una ter çia

77

TRANSCRIPT AND TRANSLATION

345

TRANSLATION [1–6]

si2 hun1 gin5 四分艮 ‘four candareen of silver’ si2 hun1 jiong5 四分絨 ‘four candareen of silk velvet’

[8–13]

(75) li1 尼 […] si2 li1 kim1 四尼金 ‘four li1 of gold’ si2 li1 gin5 四尼艮 ‘four li1 of silver’

[15–23]

(76) thung2

: another measure of ten terçias

si2 thung2 pou3 四 布 four pieces of cloth in this measure si2 thung2 tei7 四 地 four measures of earth

[25–26]

(77) chioh4 尺: a foot

Notes 4–5 8–13 8, etc. 15, etc.

Douglas (1873:184): jiong5 ‘silk velvet, floss silk’ Some words are illegible due to paper damage. The conjectured readings and are based on the EMH examples 金 kim1 ‘gold’ and 艮 gin5 ‘silver’. li1 ≈ li5 ‘a small weight, one-tenth of a hun’; cf. Douglas (1873:302). = 丈; thung2 ≈ tng7 ‘a chang, or measure of 10 Chinese feet’; cf. Douglas (1873:512). Note that the use of superscript is in conflict with my analysis in Chapters 4 and 5, according to which should be attached to syllables with an aspirated initial and an open nasal final. As this Notes continued on p. 347

346

ARTE DE LA LENGUA CHIO CHIU

f. 30r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

sỳ chîo'c pǒu sỳ chîo'c toân

四尺布 四尺

4 terçias de mantas quatro de damasco

chǔnc [x] 寸

para pul gadas

sì chǔnc pǒu sì chǔnc toân

四寸布

4 pulgadas de mantas 4 pulgadas de damasco

tàu

斗

una me 79 dida de diez gantas

sì tàu bì

四斗米

4 medidas destas que son 40 gan tas.

四寸

78

TRANSCRIPT AND TRANSLATION

347

TRANSLATION [1–4]

si2 chioh4 pou3 四尺布 ‘four feet of cloth’ si2 chioh4 tuan7 四尺 ‘four feet of damask’

[6–12]

(78) chun3 寸: for inches si2 chun3 pou3 四寸布 ‘four inches of cloth’ si2 chun3 tuan7 四寸 ‘four inches of damask’

[14–21]

(79) tau2 斗: a measure of ten gantas si2 tau2 bi2 四斗米 four measures of this, 40 gantas

Notes f. 29v, right column 15, etc. is the only example of occurring with a closed nasal final, I assume that the handwritten transcription is erroneous. 16, etc. According to Chardon, the tercia ‘Castilian foot’ is one third of a vara (1980:301, fn. 20). Clarke lists eleven regional equivalents of vara, ranging from 0.83 m to 1.1 m (1891:90). 25 A chioh4 (Mnd. chí) is commonly translated as ‘foot’, its actual length is subject to regional differences (Morse, 1908:173). According to Douglas, it is ‘a very little shorter than the English foot’ (1873:82). f. 30r, left column 6–12 Douglas: chun3 ‘the Chinese inch, one tenth of a Chinese foot, therefore larger than the British inch’ (1873:98). Carrington (1864) indicates three different English equivalents for Spanish all of which are slightly less than an inch. Differences are due to chronological and regional differences, viz. 0.843 inch (Aragon, p. 5), 0.926 inch (standard according to the ‘Old System,’ p. 88f.), 0.992 inch (Valencia, p. 98). 14–21 Douglas: tau2 ‘a dry measure somewhat like a peck’ (1873:478). The Spanish word ganta is Philippine Spanish, equaling three liters (DLE, 2001).

348

ARTE DE LA LENGUA CHIO CHIU

f. 30r, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

=======\\=======\\======= ansi cuentan los dias de la sema na acomodandose a nosotros que ellos no tienen semanas =======\\=======\\======= lèi pǎi 礼拜 semana i domingo lèi pǎi ŷt' 礼拜一 lunes lèi pǎi xī 礼拜二 martes lèi pǎi sán 礼拜三 miercoles lèi pǎi sì 礼拜四 jueves lèi pǎi gòu 礼拜五 biernes lèi pǎi làc' 礼拜六 sabado =======\\=======\\======= cuenta de los meses del año chìt' nˉy 一年十 el ano ti chàp' ene doçe xī gùe' xìt' 二月日 meses chián gùe' 正月 primer mes ques para ellos febrero

TRANSCRIPT AND TRANSLATION

349

TRANSLATION [1–5]

[6–15]

[16] [17–22]

Days of the week are counted like this, adapting our [system], as they do not have weeks. lei2-pai3 lei2-pai3 it4 lei2-pai3 ji5 lei2-pai3 sann1 lei2-pai3 si2 lei2-pai3 gou2 lei2-pai3 lak8

礼拜 礼拜一 礼拜二 礼拜三 礼拜四 礼拜五 礼拜六

week or Sunday Monday Tuesday Wednesday Thursday Friday Saturday

Counting the months of the year: cit8 ni5 cap8 ji5 geh8 jit8 一年十二月日 ‘the year has twelve months’ ciann1 geh8 正月 ‘the first month’ (which is February for them)

Notes 2–4 21–22

In premodern China, the tradition of counting seven days as one week had not been adopted. Instead, ten days were known as a unit of one xún 旬. The first day of the year according to the Chinese lunar calendar falls in the period between mid-January and mid-February according to the Gregorian calendar.

350

ARTE DE LA LENGUA CHIO CHIU

f. 30v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

xī gùe' sán gùe' sỳ gùe' gòu gùe' làc' gùe' chìt'c gùe' {p}ê' gùe' càu gùe' chàp' gùe'

2. março 3. abril 4. mayo 5. junio 6. julio 7. agosto 8. septiem bre 九月 9. otubre 十月 10. nobi embre 十一月 11. dicien bre 十二月 12. enero 二月 三月 四月 五月 六月 七月 八月

chàp' ît' gùe' chàp' xī gùe' _______\\________\\________ quenta del mes cuentan por lunas. no por sol. ________\\________\\________ chéc ît' 一 primero de luna o del mes c ché xī 二 2 chéc sán 三 3

TRANSCRIPT AND TRANSLATION

351

TRANSLATION [1–16]

[17–19] [20–24]

ji5 geh8 sann1 geh8 si2 geh8 gou2 geh8 lak8 geh8 chit4 geh8 peh4 geh8 kau2 geh8 cap8 geh8 cap8 it4 geh8 cap8 ji5 geh8

(2) March (3) April (4) May (5) June (6) July (7) August (8) September (9) October (10) November (11) December (12) January

Months are counted according to the moon and not the sun. che1 it4 che1 ji5 che1 sann1

Notes 20, etc. 24, etc.

二月 三月 四月 五月 六月 七月 八月 九月 十月 十一月 十二月

=初 =初

一 二 三

the first of the moon or of the month 2 3

352

ARTE DE LA LENGUA CHIO CHIU

f. 30v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

chéc sỳ chéc gòu chéoc làc' chéc chît'c chéc pê' chéc càu chéc chàp' chàp' ît' chàp' xī chàp' sán chàp' sì chàp' gòu chàp' làc' chàp' chît'c chàp' pûe'

四 五 六 七 八 九 十 十一 十二 十三 十四 十五 十六 十七 十八

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

TRANSCRIPT AND TRANSLATION

TRANSLATION [1–15]

che1 si2 che1 gou2 che1 lak8 che1 chit4 che1 peh4 che1 kau2 che1 cap8 cap8 it4 cap8 ji5 cap8 sann1 cap8 si2 cap8 gou2 cap8 lak8 cap8 chit4 cap8 pueh4

四 五 六 七 八 九 十 十一 十二 十三 十四 十五 十六 十七 十八

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

353

354

ARTE DE LA LENGUA CHIO CHIU

f. 31r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

chàp' càu xī chàp' xì' ît' xì' xī xì' sán xì' sì xì' gòu xì' làc' xì' chît'c xì' pûe' xì' càu sán chàp'

19 20 21 22 23 24 25 26 27 28 29 30 fin de la luna ________\\________\\________ las nunas unas tienen a trein ta dias. otros a veinte nuebe ________\\________\\________ gùe' tôa sán 月大三 la luna mayor es de chàp' xìt' 十日 treinta dias 十九 二十 卄一 卄二 卄三 卄四 卄五 卄六 卄七 卄八 卄九 三十

f. 31r, right column 1 2 3 4 5

gùe' sìo xì'

月小卄

càu xìt'

九日

6 7 8 9 10 11

Bǎng lèg' sỳ chàp' pê' nī

萬 曆 四 十 八 年

la luna menor es de vein te y nueue dias

TRANSCRIPT AND TRANSLATION

355

TRANSLATION [1–14]

[15–17] [18–21]

cap8 kau2 ji5 cap8 jih8 it4 jih8 ji5 jih8 sann1 jih8 si2 jih8 gou2 jih8 lak8 jih8 chit4 jih8 pueh4 jih8 kau2 sann1 cap8

十九 二十 卄一 卄二 卄三 卄四 卄五 卄六 卄七 卄八 卄九 三十

19 20 21 22 23 24 25 26 27 28 29 30 (end of the moon)

Some months have thirty days and some months have 29 days. geh8 tua7 sann1 cap8 jit8 月大三十日 ‘The major month has thirty days.’

[f. 31r, right column] [1–5]

geh8 sio2 jih8 kau2 jit8 月小卄九日 ‘The minor month has 29 days.’

[6–11]

Bang3-lik8 si2 cap8 peh4 ni5 萬曆四十八年

Note 6–11

bang3 ≈ ban7 ‘ten thousand’; the year is 1620 according to the Gregorian calender, see p. 8.

356

ARTE DE LA LENGUA CHIO CHIU

f. 31v, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

chù

子

tiùc

丑

ˉyn

寅

baù

卯

sīn chì

巳

gòu

午

bˇy

未

sýn

申

yù

酉

sûr'

戌

hǎi

亥

de media noche asta las dos de las dos a las qua tro de quatro a seis de seis a las ocho desde las o cho a las diez de las diez al las doçe de las doçe del dia asta las dos de dos a qua tro de la tar de de quatro asta las seis de la tarde desde las seis asta las ocho de las ocho asta las diez de las diez as ta media no che

TRANSCRIPT AND TRANSLATION

357

TRANSLATION [2–32]

cu2 tionn2 in5 bau2 sin5 ci2 gou2 bi3 sin5 iu2 sut4 hai3

子 丑 寅 卯 巳 午 未 申 酉 戌 亥

from midnight until two from two to four from four to six from six to eight from eight to ten from ten to twelve from twelve in the day until two from two to four in the afternoon from four to six in the afternoon from six to eight from eight to ten from ten to midnight

Notes 1ff.

12 15 20 30

The twelve characters listed here are known as the ‘twelve branches’ (shíèr zhī 十二支), which, together with the shígān 十干 ‘ten stems’ form the gānzhī system. For a brief introduction and further references to this system and its role in traditional Chinese chronology and time calculation, I refer to Wilkinson (2000:175–219). =辰 ci2 ≈ ci7 bi3 ≈ bi7 hai3 ≈ hai7

358

ARTE DE LA LENGUA CHIO CHIU

f. 31v, right column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

chù sˉy 子時初 a las doçe chéc de las noche chù sī 子時中 a la una tióng chù sˉy 子時尾 a la una i buè media

ît' kén

一更

xī kén sán kén

二更 三更

sỳ kén 四更 gòu kén 五更

primer guar ro de la noche 2 3 media noche 4 4

TRANSCRIPT AND TRANSLATION

359

TRANSLATION cu2 si5 che1 cu2 si5 tiong1 cu2 si5 bue2 4 1 [11–17] it kenn 5 ji kenn1 sann1 kenn1 si2 kenn1 gou2 kenn1 [2–7]

子時初 子時中 子時尾 一更 二更 三更 四更 五更

at midnight at one at half past one first quarter of the night (2) (3) midnight (4) (4)

Notes 11–17 17

The division of daytimes into ten night watches (five during daytime hours, five at night; EMH kenn1, Mnd. gēng) was one traditional Chinese way of dividing the day into smaller units (cf. Wilkinson, 2000:198ff.). (4) = (5)

360

ARTE DE LA LENGUA CHIO CHIU

f. 32r, left column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

bīn hua chây' chà chà lāi bīn hoa chà kìc lāy kín tǒan chà lāi chìa' chà pûin lāi chìa' pûin oǎn lāi xìt' sìo tǎu lāi xìt' tǎu lāy

冥向載 manana 早早來 de mana na be 冥向早 maná enleban tandose 今旦早 oi de mana na 來 bino 食早 despues de almorsar 來 bendras 食 el tienpo de de almorsar 晏來 asta comer 日小 como a las diez 午來 uen 日午來 a medio dia ben 起來

TRANSCRIPT AND TRANSLATION

361

TRANSLATION [3–6]

bin5-hua caih4 ca2 ca2 lai5 冥向載早早來 tomorrow early in the morning come  ‘You come tomorrow early in the morning!’

[7–10]

bin5 hua ca2 ki lai5 冥向早起來 tomorrow morning getting up  ‘getting up in the morning’

[11–13]

kin1 tuann3 ca2 lai5 今旦早來 today in the morning he came  ‘He came this morning.’

[14–16]

ciah8 ca2 puinn7 lai5 食早 來 after breakfast you will come  ‘You will come after breakfast.’

[17–19]

ciah8 puinn7 uann3 lai5 食 breakfast time till eating

[20–22]

jit8 sio2 tau3 lai5 日小午來 about ten come  ‘Come around ten!’

[23–24]

jit8 tau3 lai5 日午來 at noon come  ‘Come at noon!’

晏來

Notes 3ff. 3–4 14–16 17–19

As the boundaries of lines indicating syntactic units are rather unclear on this and the following folio, the syntactic divisions follow my tentative analysis of the EMH sentences.

Posited tone: yīnrù Tone numeral (TLPA): 4 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

Arte ak4 ‘handful’ bah4 ‘meat’ bat4 ‘know’ bin5-hua caih4 ‘tomorrow’ bueh4 ‘wish, soon’ cat4 ‘cl.: reason’ cheh4 ‘book’ chioh4 ‘foot’ chit4 ‘seven’ ciah4 ‘cl.: ships, big animals’ ciah4 si3 ho2 ‘very good’ cik4 ‘one’ cioh4 ‘borrow’ co7 tit4 ‘able to do’ gok4-hu5 ‘crocodile’ iah4 ‘sheet of paper’ it4 ‘one’ (bound) jit8 ssuah4 ‘in the evening’ kap4 ‘with’ khah4 comp. marker khiap4 ‘wicked’ khit4 dative, passive marker kip4 ‘step’ pak4 ‘cl.: pictures, paintings’ peeh4 ‘hundred’ peh4 ‘eight’ po2-pueh4 ‘riches’ put4 ‘not’ sioh4 ‘love’ sut4 11th stem tik4 ‘reed’ toh4 ‘table’ ssoh4 ‘rope’

Douglas -bah4 ‘flesh pork’ bat4 ‘know’ bin5-na2-cai3 ‘tomorrow’ bueh4 ‘wish, will’ cat4 ‘cl.: affairs’ cheh4 ‘book’ chioh4 ‘Chinese foot’ chit4 ‘seven’ ciah4 ‘cl.: ships, etc.’ ciah4 ‘thus, so, then’ -cioh4 ‘borrow’ tit4 aux.: ability gok8-hi5 ‘crocodile’ iah4 ‘leaf of a book’ it4 ‘one’ (bound) suah4 ‘end’ kap4 ‘with’ khah4 comp. marker khiap4 ‘cowardly’ khit4 passive marker kip4 ‘degree’ pak4 ‘cl.: scrolls’ pεh4 (Zh) ‘hundred’ peh4 (Zh) ‘eight’ po2-pue3 ‘precious things’ put4 ‘not’ sioh4 ‘love’ sut4 cyclical character tik4 ‘bamboo’ toh4 ‘table’ soh4 ‘rope’ total

I

II

         

 

                     31

 

      28

I

II

 

 

      

          

Tone diacritic in the Arte: macron Posited tone: yángpíng Tone numeral (TLPA): 5 1. 2. 3.

Arte a5 ‘or’ bak8-cing5 ‘now’ bin5-oa-jit8 ‘tomorrow’

Douglas a2 ‘or’ bak8-cing5 ‘at present’ bin5-na2-jit8 ‘tomorrow’

377

APPENDIX

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.

Arte bo5 ‘not have’ bu5-kik8 ‘infinitely’ bue5 ‘cl.: round things’ can5 ‘layer’ cang5 ‘cl.: trees, bushes’ cha5 ‘firewood’ chai3-thau5 ‘radish’ chan5 ‘field’ chng5 ‘bed’ ciau5 ‘all’ cinn5 ‘money’ co7-pu5 ‘together’ cua5 ‘snake’ cun5 ‘ship’ ei5 ‘shoe’ gau5 ‘wise’ gin5 ‘silver’ gu5 ‘ox’ guan5 ‘fountain’ hee5 ‘shrimp’ hing5 ‘return’ hing5-sionn3 ‘figure’ ho5 a measure hong5-tei3 ‘king’ hu5 ‘fish’ in5 third branch ji5 ‘character’ ji5 ‘two’ (bound) jiong5 ‘silk’ kei1 thi5 ‘cockrow’ khim5 ‘lute’ kun5 ‘cl.: herds, companies’ lai5 ‘come’ lang5 ‘man’ liam5-cui2-lang5 ‘Christian’ lin5 ‘camlet’ ling5 ‘satin’ me5-hui ‘in the evening’ mui5 ‘door’ ni5 ‘year’ ou5 ‘heap’ pang5 ‘room’ penn5-penn5 ‘equally’ phue5 ‘skin’ ping5 ‘side’ pou5-to5 ‘grape’ pou5-to5 ‘grape’ puann5 ‘plate’ si5 ‘time, moment’ siam5 ‘arm’s length’

Douglas bo5 ‘not have’ bu5 ‘not’ -can5 ‘layer’ cang5 ‘cl.: trees, bushes’ cha5 ‘firewood’ thau5 substantive suffix chan5 ‘field’ chng5 ‘bed’ ciau5 ‘complete’ cinn5 ‘money’ cue3-pu5 ‘come together’ cua5 ‘snake’ cun5 ‘ship’ e5 ‘shoe’ gau5 ‘wise’ gin5 (Zh) ‘silver’ gu5 ‘ox’ guan5 ‘fountain’ he5 ‘shrimp’ hing5 ‘return’ hing5 ‘form, shape’ ho5 ‘small weight’ hong5-tei3 ‘emperor’ hu5 (Ch) ‘fish’ in5 cyclical character ji7 ‘character’ ji7 ‘two’ (bound) jiong5 ‘silk velvet, floss silk’ kei1 thi5 ‘the cock crow’ khim5 ‘harp’ kun5 ‘herd, company’ lai5 ‘come’ lang5 ‘man’ -lin5 ‘thin glossy sort of silk’ ling5 ‘thin silk stuff’ mi5 ‘night’ mui5 (Zh) ‘door’ ni5 ‘year’ -pang5 ‘room’ pinn5-pinn5 ‘equally’ phue5 ‘skin’ ping5 ‘side’ pou5-to5 ‘grape’ pou5-to5 ‘grape’ puann5 ‘plate’ si5 ‘time, etc.’ siam5 ‘join the hands’

I  

II  

                              

                      

    

    

        

        

     

378 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

APPENDIX Arte sin5 fifth branch sin5 hun5 ‘soul’ sin5 hun5 ‘soul’ sin5 nineth branch tai5 ‘cl., stage’ tang5 ‘with’ tang5-ching3 ‘metal part’ tei5 ‘cl.: counted sequences’ thou5 ‘earth’ thng5 ‘sugar’ tiau5 ‘cl.: sins, roots’ ting5 ‘cl.: folds’ tio5-si3 ‘God’ tiu5 ‘taffeta’ uan5 ‘cl.: balls’ un1-cionn5 ‘big dresses’

Douglas siam5 cyclical character sin5 hun5 ‘soul’ sin5 hun5 ‘soul’ sin5 cyclical character tai5 ‘small platform’ tang5 ‘with’ tang5 ‘copper, brass’ te7 ‘prefix: ordinal numbers’ thou5 ‘earth’ thng5 ‘sugar’ tiau5 ‘cl.: long things’ ting5 ‘cl.: folds’ -tiu5 ‘silk stuffs’ uan5 ‘cl.: balls’ in1-chionn5 total

I            

II       

   64

   61

I      

II      

 

 

             

 

   

Tone diacritic in the Arte: circumflex Posited tone: yángqù Tone numeral (TLPA): 7 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Arte bei7 ‘sell’ bo7 ‘hat’ bue7 ‘has not’ ca:1-bou7-jit8 ‘yesterday’ cai7 ‘load’ cei7 ‘many’ cionn7 ping5 ‘this side’ co7 ‘make’ cu7-ki2 ‘oneself’ cue7 ‘rows of arranged things’ cun7 ‘when’ kang7 ‘with’ kau7 ‘until’ kiam7 ‘sword’ ku7-ni5 ‘last year’ lau7 ‘pinch’ liam7 king1 ‘pray’ lou7 ‘journey’ na7 ‘if’ phio7 ‘license’ pou7 ‘cl.: books’ puinn7 ‘rice’ sai1-hu7 ‘workman’ si3-cui7 ‘who’

Douglas be7 (Zh) ‘sell’ bo7 ‘hat’ bue7 (Zh) ‘has not’ ca7-boo7-jit8 ‘yesterday’ cai7 ‘cargo’ ce7 (Zh) ‘many’ -co7 ‘make’ cu7-ki2 ‘one’s self’ -si5-cun7 ‘time’ kang7 ‘together’ kau3 ‘till’ kiam3 ‘sword’ ku7-ni5 ‘last year’ lau7 ‘leak out’ liam7 ‘recite’ loo7 ‘journey’ na7 ‘if’ phio3 ‘warrant’ poo7 ‘a set, as of books’ puinn7 (Zh) ‘rice’ sai1-hu7 ‘skilled workman’ cui7 ‘who’

        

379

APPENDIX

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Arte si5-ce7 ‘when’ sing7 ‘cl.: sedan chair’ su7 ‘thing, affair’ tai7-ke1 ‘together’ tann7 ‘load’ tann7 ‘speak’ tei7 ‘earth’ thou7 ‘give’ tiann7-tiann7 ‘only’ tua7 ‘great’ tua7 ‘cl.: reason’ tuan7 ‘damask’ tuinn7 ‘cl.: meal’ suann7 ‘thread’

Douglas --su7 ‘affair’ tai7-ke1 ‘all of you’ tann3 ‘load’ tann3 ‘speak’ te3 ‘earth’ thou3 ‘give’ tiann7-tiann7 ‘only’ tua7 ‘great’ -tuan7 ‘satin’ tuinn3 (Zh) ‘cl.: meal’ suann3 ‘thread’ total

I

II

       

 

   33



I 

II 

    

    

            ?  

   

 

24

Tone diacritics in the Arte: vertical stroke