Language Diversity in the Sinophone World: Historical Trajectories, Language Planning, and Multilingual Practices 0367504510, 9780367504519

Table of contents :
Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
List of Figures
List of Tables
List of Contributors
Acknowledgments
Introduction: Language Diversity in the Sinophone World
Part I: Historical Trajectories
1 What Was Standard Chinese in the Nineteenth Century? Divergent Views in the Times of Transition
2 Manchu, Mandarin, and the Politicization of Spoken Language in Qing China
3 Romanizing Southern Mǐn: Missionaries and the Promotion of Written Chinese Vernaculars
4 Interactions across Englishes in Mainland China, Hong Kong, Macao, and Singapore
Part II: Language Planning
5 One Legacy, Two Legislations: Language Policies on the Two Sides of the Taiwan Strait
6 Language Policy and Practice in Taiwan in the Early Twenty-First Century
7 A Tale of Two Special Administrative Regions: The State of Multilingualism in Hong Kong and Macao
8 One People, One Nation, One Singapore: Language Policy and Shifting Identities among Chinese Singaporeans
Part III: Multilingual Practices
9 Speakers of “Mother Tongues” in Multilingual China: Complex Linguistic Repertoires and Identity Construction
10 Multilingualism and Language Policy in Singapore
11 The Discourses of lào yīngwén: Resistance to and Subversion of the Normative Status of English in Taiwan
12 Conventionalized Code-Switching in Taiwan: English Insertions in Taiwan Mandarin
13 Ubiquitous but Unplanned: The Utterance-Final Particle ê in Taiwan Mandarin
14 Diverse Language, Diverse Grammars: On Quirky Phenomena in Mandarin
Index

Citation preview

Language Diversity in the Sinophone World

Language Diversity in the Sinophone World offers interdisciplinary insights into social, cultural, and linguistic aspects of multilingualism in the sinophone world, highlighting language diversity and opening up the burgeoning field of Sinophone studies to new perspectives from sociolinguistics. The book begins by charting historical trajectories in Sinophone multilingualism, beginning with late imperial China through to the emergence of English in the mid-nineteenth century. The volume uses this foundation as a jumping off point from which to provide an in-depth comparison of modern language planning and policies throughout the Sinophone world, with the final section examining multilingual practices not readily captured by planning frameworks and the ideologies, identities, repertoires, and competences intertwined within these different multilingual configurations. Taken together, the collection makes a unique sociolinguistic-focused intervention into emerging research in Sinophone studies and will be of interest to students and scholars within the discipline. Henning Klöter is Full Professor of Modern Chinese Languages and Literatures in the Institute for Asian and African Studies at Humboldt University of Berlin, Germany. Mårten Söderblom Saarela is an Assistant Research Fellow at the Institute of Modern History, Academia Sinica, Taiwan.

Routledge Studies in Sociolinguistics

How we really say ‘no’ Linguistic mitigation in English and Spanish Nydia Flores-Ferrán Linguistic Variation and Social Practices of Normative Masculinity Authority and Multifunctional Humour in a Dublin Sports Club Fergus O’Dwyer A Sociolinguistic View of a Japanese Ethnic Church Community Tyler Barrett Socio-grammatical Variation and Change In Honour of Jenny Cheshire Edited by Karen V. Beaman, Isabelle Buchstaller, Sue Fox and James A. Walker Social Theory and Language The Construction of Meaning Glyn Williams Scripting Japan Orthography, Variation, and the Creation of Meaning in Written Japanese Wesley C. Robertson Language Diversity in the Sinophone World Historical Trajectories, Language Planning, and Multilingual Practices Edited by Henning Klöter and Mårten Söderblom Saarela

For more information about this series, please visit https://www.routledge. com/Routledge-Studies-in-Sociolinguistics/book-series/RSSL

Language Diversity in the Sinophone World Historical Trajectories, Language Planning, and Multilingual Practices Edited by Henning Klöter and Mårten Söderblom Saarela

First published 2021 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, Henning Klöter and Mårten Söderblom Saarela; individual chapters, the contributors The right of Henning Klöter and Mårten Söderblom Saarela to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Klöter, Henning, 1969– editor. | Söderblom Saarela, Mårten, editor. Title: Language diversity in the Sinophone world : historical trajectories, language planning, and multilingual practices / edited by Henning Klöter and Mårten Söderblom Saarela. Description: New York, NY : Routledge, 2021. | Series: Routledge studies in sociolinguistics | Includes bibliographical references and index. | Summary: “Language Diversity in the Sinophone World offers interdisciplinary insights into social, cultural, and linguistic aspects of multilingualism in the Sinophone world, highlighting language diversity and opening up the burgeoning field of Sinophone studies to new perspectives from sociolinguistics. The book begins by charting historical trajectories in Sinophone multilingualism, beginning with late imperial China through to the emergence of English in the mid-19th century. The volume uses this foundation as a jumping off point from which to provide an in-depth comparison of modern language planning and policies throughout the Sinophone world, with the final section examining multilingual practices not readily captured by planning frameworks and the ideologies, identities, repertoires, and competences intertwined within these different multilingual configurations. Taken together, the collection makes a unique sociolinguistic-focused intervention into emerging research in Sinophone studies and will be of interest to students and scholars within the discipline”—Provided by publisher. Identifiers: LCCN 2020019821 (print) | LCCN 2020019822 (ebook) | ISBN 9780367504519 (hardback) | ISBN 9781003049890 (ebook) Subjects: LCSH: Language policy—China. | Multilingualism— China. | Chinese language—Variation. Classification: LCC P119.32.C6 .L35 2020 (print) | LCC P119.32.C6 (ebook) | DDC 306.44/951—dc23 LC record available at https://lccn.loc.gov/2020019821 LC ebook record available at https://lccn.loc.gov/2020019822 ISBN: 978-0-367-50451-9 (hbk) ISBN: 978-1-003-04989-0 (ebk) Typeset in Times New Roman by codeMantra

Contents

List of Figures List of Tables List of Contributors Acknowledgments Introduction: Language Diversity in the Sinophone World

vii ix xi xv 1

H E N N I NG K L Ö T E R A N D M Å RT E N S ÖDE R BL OM SA A R E L A

PART I

Historical Trajectories

11

1 What Was Standard Chinese in the Nineteenth Century? Divergent Views in the Times of Transition

13

R IC H A R D VA N N E S S SI M MON S

2 Manchu, Mandarin, and the Politicization of Spoken Language in Qing China

39

M Å RT E N S ÖDE R BL OM SA A R E L A

3 Romanizing Southern Mǐn: Missionaries and the Promotion of Written Chinese Vernaculars

60

D ON S NOW

4 Interactions across Englishes in Mainland China, Hong Kong, Macao, and Singapore

80

C H R I S T I A N E M E I E R KOR D

PART II

Language Planning 5 One Legacy, Two Legislations: Language Policies on the Two Sides of the Taiwan Strait H E N N I NG K L Ö T E R

99 101

vi Contents

Figures



Tables



Contributors

Su-Chiao Chen 陳淑嬌 is professor at the Department of Applied Foreign Languages and director of the Language Center at National Taiwan University of Science and Technology, Taiwan. She is past president of Taiwan ESP Association. Her research focuses on issues of multilingualism, acculturation, and language policy and planning in education, specifically in relation to minority or immigrant groups in multilingual societies. She has published several books and numerous articles in refereed journals. Yeow Wah Fong 冯耀华 is a PhD candidate at the National Institute of Education, Nanyang Technological University, with research interests in language use and identity with specific reference to the Chinese community in Singapore, language policy and planning, and design and development of Chinese language instructional materials. Yeng Seng Goh 吴英成 is Professor of Linguistics and Language Education in the Office of Education Research at the National Institute of Education, Nanyang Technological University. He received his BA in Chinese language and literature from National Taiwan University and a PhD in linguistics from SOAS, University of London. He is a bilingual and bicultural scholar whose key research interests include the teaching of Chinese as a(n) second/foreign/international language, Chinese-English language policies and language education, contrastive linguistics and translation, heritage language education, ICT in teaching and learning; universityindustry collaboration. Henning Klöter  (PhD, Leiden University, 2003) is Professor of Modern Chinese Languages and Literatures at the Institute of Asian and African Studies (IAAW), Humboldt University of Berlin. He has previously held positions at the universities of Göttingen, Mainz, Bochum, and National Taiwan Normal University. His major publications are Written Taiwanese (2005) and The Language of the Sangleys: A Chinese Vernacular in Missionary Sources of the Seventeenth Century (2011).

xii Contributors David C.S. Li 李楚成 is a professor and head of the Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University. He studied in Hong Kong and France and obtained his PhD in Germany. Trained in general linguistics, he became interested in social aspects of language learning and use in multilingual settings. He has published papers on World Englishes, Hong kong English, code-switching, and multilingualism in Greater China. He is currently researching Sinitic brushtalk in early modern Sinographic East Asia. Lijun Li 李丽君 is a PhD student, research assistant, and lecturer at the Department of English and American studies, University of Hamburg. Her research interests include contact linguistics, Chinese substrates, World Englishes, translation studies, multilingualism, and second and foreign language acquisition. She currently focuses on language contact and diachronic development of Singapore English in relation to the substrate languages spoken in the multilingual context of Singapore. Sihua Liang 梁斯华 is the author of Language Attitudes and Identities in Multilingual China: A Linguistic Ethnography. She received her PhD from the University of Cambridge and is currently working as a research associate at Sun Yat-sen University in Guangzhou, China. Her training is in sociolinguistics and linguistic anthropology, and she is especially interested in language policy, (trans)languaging practice, and identity construction of Chinese dialect speakers in linguistically diverse regions. Chin-hui Lin 林欽惠 is a lecturer in Mandarin at the Institute of Asian and African Studies (IAAW), Humboldt University of Berlin. She worked at the Department of Chinese Studies, Leiden University from 2001 to 2012 and at the Department of East Asian Studies, Göttingen University from 2013 to 2015. She holds an MA (2003) in teaching Chinese as a foreign language from National Taiwan Normal University and received her PhD in Chinese linguistics from Leiden University in 2014. Her research interests include teaching Chinese as a foreign language, conversation analysis, and language contact. Christiane Meierkord is Professor of English Linguistics at Ruhr University Bochum. She is author of Interactions across Englishes: Linguistic Choices in Local and International Contact Situations (2012), co-editor of Ugandan English: Its Sociolinguistics, Structure and Uses in a Globalising PostProtectorate (2016), and has published extensively on the use of English as a lingua franca. Her current research focuses on English in postprotectorates and the spread of English at the grassroots of societies. Peter Siemund is Professor of English Linguistics at the University of Hamburg. He pursues a cross-linguistic typological approach in his work on reflexivity and self-intensifiers, pronominal gender, interrogative constructions, speech acts and clause types, argument structure, tense and

Contributors  xiii aspect, varieties of English, language contact, and multilingual development. His publications include Pronominal Gender in English: A Study of English Varieties from a Cross-Linguistic Perspective (Routledge 2008), Varieties of English: A Typological Approach (CUP 2013), and Speech Acts and Clause Types: English in a Cross-Linguistic Context (OUP 2018). Richard VanNess Simmons is Professor in the School of Chinese at The University of Hong Kong and Professor of Chinese in the Department of Asian Languages and Cultures at Rutgers University. He holds a PhD in Sinology and Chinese dialectology from the University of Washington, Seattle. Simmons’ research areas include dialectology, dialect geography, historical linguistics, history of Chinese, history of Mandarin, and Chinese sociolinguistics. He has done extensive fieldwork investigating and mapping China’s Mandarin and Wu dialects. Don Snow (PhD in East Asian Languages and Cultures, Indiana University) is director of the Language and Culture Center at Duke Kunshan University. His research focuses on written forms of Chinese vernaculars, and he is the author of Cantonese as Written Language: The Growth of a Written Chinese Vernacular (Hong Kong University Press, 2004). His recent articles examine the histories of written Suzhounese, Shanghainese, and Chaoshanese. Mårten Söderblom Saarela is a historian of late imperial China with an interest in the cultural and intellectual history of language. He holds a PhD from Princeton University and is currently an assistant research fellow at the Institute of Modern History, Academia Sinica. He was previously a postdoc at the Max Planck Institute for the History of Science. Söderblom Saarela is the author of  The Early Modern Travels of Manchu: A Script and Its Study in East Asia and Europe (University of Pennsylvania Press, 2020). Hsi-Yao Su 蘇席瑤  is associate professor in the Department of English at National Taiwan Normal University. Her research focuses on the connection between language, identity, and language ideologies in the context of Taiwan and transnational Chinese communities. She has published in Journal of Sociolinguistics, Journal of Pragmatics, Journal of ComputerMediated Communication, Gender and Language, Discourse, Context & Media, Journal of Asian Pacific Communication, Language and Linguistics, Concentric: Studies in Linguistics, and Monumenta Taiwanica. Choi-Lan Tong 湯翠蘭 is an associate professor of the School of Languages and Translation, Macao Polytechnic Institute. She obtained her BA and MA in Taiwan and her PhD in Guangzhou. Trained in general linguistics and historical Chinese phonology, she has published extensively on the linguistic characteristics of Yuè dialects in the Pearl River Delta region, especially Macao. Through the lens of multilingualism and multidialectal

xiv Contributors interaction, she has a keen interest in researching the intricate relationships between language/dialect and culture. Julia Wasserfall is a research associate and lecturer at the Institute for Asian and African Studies (IAAW), Humboldt University of Berlin. She has a master’s degree in Chinese language and literature from Ruhr University Bochum (2013) and recently finished her PhD dissertation project on English language ideologies in Taiwan and Singapore at Humboldt University of Berlin (2019). Jeroen Wiedenhof  was trained as a linguist and sinologist in Leiden and Beijing. He has joint appointments as assistant professor at Leiden University’s Institutes for Area Studies and for Linguistics (LIAS, LUCL). He is a researcher in Chinese and general linguistics, and a lecturer in linguistics, Mandarin and Classical Chinese. His interests include syntax and semantics, synchronic phonology, and evolutionary linguistics. The international edition of his A Grammar of Mandarin was published by John Benjamins in 2015.

Acknowledgments

First drafts of the chapters in this volume were presented at a conference at the University of Göttingen in 2015 which was generously sponsored by the German Research Foundation (DFG), the Chiang Ching-kuo Foundation for International Scholarly Exchange, and the Department of International and Cross-strait Education of the ROC Ministry of Education (Taiwan). We thank all institutions for their support. We also express our sincere gratitude to Karen Finney-Kellerhoff who has meticulously proofread all chapters. Aurel Eschmann, Valeriya Morozhnikova, and Eleanor Schaumann are to be thanked for their excellent editorial assistance. Our thanks also go to the two anonymous reviewers of our book proposal. Finally, we would like to thank all contributors for their constructive participation in the publication process.

Introduction Language Diversity in the Sinophone World Henning Klöter and Mårten Söderblom Saarela

Abstract The introduction places this collection into the context of sinophone studies, a rapidly growing field that critically engages with issues in Chineselanguage literature, film, and popular culture from countries with ethnic Chinese populations. Addressing the neglect of language in sinophone studies, the purpose of the volume is introduced as opening the field to language and its diversity and vice versa. Without claiming definiteness, the regional framework of the volume is then defined in terms of places that pursue official language policies on the use of one or more Sinitic languages, notably the People’s Republic of China (PRC), Hong Kong, Macao, Singapore, and Taiwan, with consideration of the historical background from which these polities and their linguistic communities emerged. The introduction is rounded off with brief chapter summaries.

Introduction This volume is a collection of essays on the social, political, and cultural aspects of multilingualism in China and its East and Southeast Asian periphery from the eighteenth century to the present day. It is an intervention in the developing field of sinophone studies, which until now has developed primarily in reference to modern literature written in Chinese but originating outside China. By looking at the historical trajectories of Chinese multilingualism, its shaping through language planning, and its present-day realization through actual, everyday practice, the chapters of the book aim at foregrounding language in sinophone studies. They reveal the diversity in the ideas about—and use of—language within Chinese-speaking communities, as well as the tensions that condition their development.

Sinophone and Language The young field of sinophone studies is marked by a paradox. On the one hand, the very concept of sinophone obviously relates to language. On the other hand, this notion rarely occurs in scholarship on Sinitic languages, but it does occur in relation to literary and other artistic works articulated in Chinese in the twentieth and twenty-first centuries. In response to this

2  Henning Klöter and Mårten Söderblom Saarela paradox, we define the general purpose of this volume as opening the field of sinophone studies to language and its diversity, and vice versa. For the past 15 years, the concept of sinophone has been a controversial issue in debates on literature, film, and popular culture from countries with ethnic Chinese populations. One of the general merits of the term is that it helps to disambiguate various meanings of the word Chinese. As McDonald writes, the English word Chinese is multiply ambiguous, corresponding to Zhōngguó (de) ‘related to China’, or Huárén (de) ‘related to Chinese ethnicity’ or Zhōngwén (de) ‘related to Chinese language’. It was no doubt for this reason that in recent years within Chinese studies the third of these meanings has received a new English coinage, sinophone. (2011: 5) McDonald’s disambiguation occurs in the context of teaching Chinese as a foreign language; a similar approach can be found in the field of Chinese literature. Significantly, however, by setting up a dichotomy between “inside China” and “outside China,” Shih’s oft-quoted definition of sinophone takes a new twist. Shih, who is often credited with inaugurating the field of sinophone studies, argues that the category of the sinophone dispels the confusion surrounding “Chinese literature” as either “literature from China” (中國 文學) or “literature in Chinese” (華文文學) (Shih 2013a: 34). Elsewhere, she explicitly takes up the inside/outside dichotomy and defines sinophone literature largely in terms of the latter, writing that “Sinophone studies takes as its objects of study the Sinitic-language communities and cultures outside China as well as ethnic minority communities and cultures within China where Mandarin is adopted or imposed” (Shih 2013b: 11, emphasis in the original). The repeated use of plurals in this quotation implies that sinophone literature comprises many languages. As Shih puts it, “Sinophone is … polyphonic and multilingual” (2013b: 10). Yet polyphonicity remains restricted to the Sinitic languages belonging to the Sino-Tibetan language family, notably topolects of Southeastern China (Cantonese, Southern Mǐn, Hakka, Chaozhou) which are also the mother tongues of ethnic Chinese communities in Southeast Asia. Writers of non-Chinese ethnicity (e.g. Tibetan) using Chinese as a literary medium are understood to fall within the purview of sinophone studies, but sinophone rarely applies to linguistic or artistic practices within these communities that are not articulated in Chinese.

Borders and Boundaries By positing a sinophone world in the title of this volume, we do not mean to imply that this “world” has clear boundaries. As a matter of fact, the sinophone world cannot be captured uncontroversially in terms of political

Introduction  3 borders. If we adopt a loose definition, the sinophone world would include any place inside or outside China where a Sinitic language is currently spoken or has been spoken in the past, be it Beijing, Berlin, or Bangalore. Indeed, as soon as the term is no longer primarily used in reference to modern literature, the concept of sinophone approaches other terms that have been used to refer to related linguistic practices across different polities in East Asia. Historians have talked about a “sinosphere” in English, and in related terms in the languages of the region (e.g. Japanese Kanji bunkaken 漢字文化圏), to capture the shared use of the Chinese script and of the Classical Chinese language. More recently, scholars have begun to pry open the deceptive uniformity of the logographic script and its arrangement on the page (which famously fooled some of the first European sojourners in East Asia) while investigating the complexity of the interaction between spoken and written languages in the region. King proposes the term “sinographic cosmopolis” for the area where Classical Chinese was used (King 2015). The diversity of language and language-contact situations in East Asia is currently being investigated from various angles, sinophone studies being one of them. Adopting too loose a definition of sinophone opens a Pandora’s box of possibilities, which would inevitably lead to arbitrariness and a low degree of comparability of case studies. While we hope that this volume will contribute to the wider discussion of linguistic pluralism in China and beyond, we have decided to restrict “our” sinophone world to polities that pursue official language policies on the use of one or more Sinitic languages, notably the PRC, Hong Kong, Macao, Taiwan, and Singapore, with consideration of the historical background from which these polities and their linguistic communities emerged. To be sure, the presence of Sinitic language policies is a defining criterion for our heuristic configuration of the sinophone world, but it is by no means the only one. Another feature that links the areas under consideration is the ethnolinguistic make-up of the respective language communities. All five places are inhabited mainly by ethnic Chinese people who speak at least one Sinitic language and whose traditions, customs, and values are to a considerable extent rooted in Chinese culture. Yet sinophone studies are based on a notion of polyphonicity or language diversity. Language diversity is not just about the existence of different languages within different communities; it is also about the diversity of attitudes, values, and ideological beliefs which speakers attach to different languages. Such beliefs, in turn, are to a great extent shaped by official language planning, for example, which defines a standard pronunciation or stipulates a language as the medium of instruction in schools. But language planning cannot explain everything; and there is never an exact match between the goals of language planning and actual language use. In other words, if we want to understand how people speak and how speakers perceive language use within their communities, we have to go beyond official language planning and ask to what extent multilingual practices are shaped by historical, cultural, or other factors.

4  Henning Klöter and Mårten Söderblom Saarela This volume therefore also seeks to explore areas outside language planning that manifest the ways speakers make use of the various linguistic resources of their environments and how they perceive language diversity within their communities. Against this backdrop, we conceptualize the sinophone as a lens through which to look at language diversity and the various factors that shape its specific manifestations.

Historical Trajectories Within sinophone studies, notions of multilingualism usually come together with a critique of the dominance of Mandarin vis-à-vis other Sinitic varieties, for example, when Shih writes of a “willingly acquired or forcefully imposed language” in her discussion of the use of Mandarin by China’s ethnic peoples or, in a similar vein, when Chow critiques the “hegemony of Mandarin” (2013: 47). To be sure, as our brief overview here shows, throughout most of the twentieth century until today, the spread of Mandarin has been at the core of language planning in most parts of the sinophone world. This spread has been accompanied by a pervasive and vaguely defined notion that Mandarin represents a pure, homogeneous, and—by sheer virtue of its existence—prestigious linguistic norm. Such a notion results from several factors such as wishful thinking, historical simplification, and uncritical mingling of linguistic description and prescription. It needs to be emphasized, however, that this volume does not seek to critique either the dominance of Mandarin or the ideologies underlying this dominance. Instead, some chapters explore the position of Mandarin from various interrelated perspectives. It opens and closes with a seemingly simple and yet central question: What actually is Mandarin? The very fact that the English word “Mandarin” has different Chinese equivalents, each with different sociolinguistic connotations, testifies to the complexity of the answer. Up until the early twentieth century, Mandarin was widely known as Guānhuà (lit. ‘language of officials’), a term which literally referred to the lingua franca of court officials in imperial China but was probably both much more widely used and less prestigious than the name suggests (Karasawa 1996). Thus, generally speaking, notions of linguistic dominance clearly have a historical dimension that cannot be separated from present-day configurations of polyphonicity. Approaching the sinophone world historically, however, poses a similar problem to our restriction in terms of space: Where to draw the line? Theoretically speaking, the ideas underlying sinophone studies can be placed into any historical context, regardless of its distance from or proximity to the present. For the sake of focus, however, we have limited our historical perspective to the late imperial and Republican periods, as sociolinguistic changes that occurred after the eighteenth century have strongly conditioned present-day configurations of language diversity. As shown in several chapters, none of the recognized historical milestones represents a clear-cut boundary in terms

Introduction  5 of sociolinguistic developments, be it the foundation of the Republic in 1912 following the fall of the Qing dynasty a few weeks earlier or the foundation of the PRC in 1949. During the nineteenth century and beyond, as pointed out in Simmons’ opening chapter, there were concepts of “standardness,” but no standard Mandarin in a narrow prescriptive sense. Instead, perceptions of correctness were highly selective. The nineteenth century witnessed a gradual shift from norm based on the pronunciation of the city of Nanjing toward a northern Beijing-based pronunciation scheme. Since the Manchu court and the Hàn Chinese literati elite had conflicting ideas as to what kind of Mandarin should be spoken, this shift remained highly controversial until the early twentieth century. Simmons effectively counters the still widespread view that the standard followed the location of the capital. In fact, several communities and groups held opposing views of the linguistic norm, and the southern elite resisted the centripetal force of the capital for a long time. Like Simmons, Söderblom Saarela emphasizes how much attention representatives of the imperial court and elite individuals paid to language issues in general and to correct pronunciation in particular. He also points out that in late imperial China the language situation within the social elite was characterized by a high degree of multilingualism, with Manchu occupying an important position alongside Classical Chinese and Mandarin. Indeed, the coexistence of several languages within the government and political elite contributed to a heightened sensitivity toward what counted as proper or correct language. The opening of China’s major coastal cities following a series of opium wars in the middle of the nineteenth century had a deep and lasting sociolinguistic impact. The presence of Western diplomats, traders, travelers, and missionaries ushered in a new period of Chinese-Western contacts and, more important for our investigation, laid the foundation for the spread of English within the sinophone world. Interestingly, however, Western influence was not restricted to English as an international language, but also affected the situation of regional Sinitic varieties. The late imperial period did not witness any official effort to promote the use or raise the status of topolects. Instead, as shown in Snow’s chapter, language planning for the latter resulted mainly from the activities of Western missionaries. These activities were closely bound to the translation of the Bible and other texts into regional Sinitic varieties. Snow argues that the promotion of vernacular literacies by missionaries on the Southeastern coast was a strategy that radically broke with the dominant conventions. In the city of Xiamen, the missionaries’ language ideology was translated into an ambitious and successful project of teaching an alphabetic orthography of the regional vernacular to local illiterates. Meierkord’s chapter, accordingly, foregrounds the spread of international languages in general and of English in particular. Although English is by no means the only international language, it is arguably the most important

6  Henning Klöter and Mårten Söderblom Saarela non-Sinitic language that has contributed to multilingualism in the sinophone world. The status and the spread of English are different in each sinophone community. In the former British colonies of Hong Kong and Singapore, English has maintained its official status and its strong position in higher education. In the PRC, by contrast, it remains a foreign language. Throughout the sinophone world the ubiquitous presence of English has made it a main contact variety and contributor to emerging creole languages.

Language Planning Entering the area of language planning and policy, the second part of this volume looks at official attempts to shape language use of the speakers in different areas of the sinophone world. Linking a historical approach with modern language planning, Klöter examines (dis)continuities in language policies from Republican China to modern language planning in the PRC and Taiwan after the political split of the two polities in 1949. He demonstrates that the 1920s marked a transitional period during which an earlier approach to language standardization inclusive of different regional features gave way to an exclusive approach that defined correctness in terms of the pronunciation of the Beijing area. Interestingly, this shift has since then been upheld on both sides of the Taiwan Strait, despite the fundamental ideological differences between the two polities. In both cases, the spread of standard Mandarin has been very successful, not only in terms of actual language use but also in terms of attitudes toward the standard. In Taiwan, however, the dominance of Mandarin has not remained uncontested. The 1990s witnessed a new awareness of local language cultivation amid a steady upsurge of English as the major international language. As discussed by Chen, the government has been promoting an English language policy for communicative proficiency and a local language policy for fostering a Taiwan identity, following decades of a “Mandarin-only” policy. The simultaneous promotion of English, local languages, and Mandarin is aimed at putting Taiwan on the map as a multilingual place in a globalized world. Yet her study suggests that government commitment to executing the local language policy has been rather weak, in contrast to the implementation of the English-language policy. The frequency of English use nevertheless remains very low. In contrast to mainland China and Taiwan, the governments of Hong Kong and Macao have historically not placed any emphasis on the spread of Mandarin, partly due to their long-term colonization by the United Kingdom and Portugal. Both territories are now Special Administrative Regions (SAR) of the PRC. Li and Tong’s overview shows that since the handovers in 1997 and 1999, both SAR administrations have preserved their autonomy in the field of language planning. Hong Kong and Macao can rightly be considered the only places in the sinophone world where a Sinitic language other than Mandarin dominates public domains. Both governments

Introduction  7 continue the colonial-era practice of assigning official status to Cantonese, the major regional language, alongside English and Portuguese. Cantonese remains as the medium of instruction in most schools and, in Hong Kong, at some universities. In addition, in the two SARs, Cantonese elements are visible both in the media. However, due to the strong ties with mainland China, Mandarin is growing in importance. Singapore was once a hotspot of sinophone polyphonicity, but language planning in the multiethnic island state has since independence in 1965 gone in the opposite direction with regard to regional Sinitic varieties. As shown by Goh and Fong, the official language policy has been shaped by the visions of the first Prime Minister Lee Kuan Yew. Following Lee’s guidelines, great emphasis was placed on ethnic integration, standardization, and internationalization. As in Taiwan before 1949, Mandarin was historically hardly in use among the Chinese population. It was only after 1965 that it became an official language—alongside English, Malay, and Tamil—and gradually gained dominance over the other Sinitic varieties through education and intensive campaigning. Despite official promotion of Mandarin, this chapter suggests that Chinese Singaporeans increasingly use English as a home language.

Multilingual Practices Plans are one thing, but the outcomes are often different. To be sure, in many parts of the sinophone world, language planning has been thoroughly successful in terms of its original goals. The fact that Mandarin is widely and efficiently used as a medium of communication is the result of successful language planning. The same is true for modern written Chinese based on Mandarin, which forms the basis of high-level literacy and literary production throughout the sinophone world. There are, however, many examples that testify to the failure of language planning to attain its goals; the low frequency of English use in Taiwan discussed in Chen’s chapter being but one example. In addition, there are facets of language use that represent neither unattained goals nor unintended consequences of official planning such as the use of written Cantonese in Hong Kong media. Many language-related phenomena elude planning, depending instead on social, cultural, or ethnic factors such as migration, loyalty to ancestral traditions, or the influencing force of popular culture. It is these phenomena beyond the scope of planning that are analyzed in the third part of the volume. Liang’s chapter considers language attitudes in Guangzhou, which was historically a Cantonese-speaking city. The city’s recent economic success has made it a popular destination for businessmen and migrant workers from all over the country, leading to a high degree of sociolinguistic complexity. Reflecting on notions of “mother tongue” and hometown language in the narratives of second-generation migrant students in Guangzhou elicited through ethnographic fieldwork, Liang identifies mismatches between

8  Henning Klöter and Mårten Söderblom Saarela the participants’ regional identities, shaped by a sense of belonging to an ancestral place, and “mother-tongue” competence. In Singapore, actual language use is likewise not identical to the official goals of language planning. Tying in with Goh and Fong’s chapter, Siemund and Li argue that the official language policy has had a deep impact on language use. A diverse multilingual society has effectively turned into one that is characterized by individual bilingualism. Siemund and Li suggest significant differences in terms of use of the declared mother tongue and great variation in the observable language repertoires, often tied to factors such as ethnicity and social background. An unplanned—and, from an official perspective, undesirable—effect of multilingualism in Singapore is the emergence of Colloquial Singapore English, or Singlish, a variety of English strongly characterized by influences from Chinese dialects and Malay. Taiwan has not witnessed the emergence of an Anglo-Sinitic contact variety comparable to Singlish in Singapore. Yet English makes up an essential part of language practices there. Su’s chapter shows that it is impossible to identify a clear-cut pattern in attitudes toward English. She shows that the strong support of strengthened English education is sometimes met with negative sentiments toward the use of English in daily life. Her analysis is pegged on the slang term lào yīngwén, roughly meaning ‘to speak English fluently (in a showy manner)’. Through a study of well-known Taiwanese bloggers, Su discusses the various ways in which English is discursively constructed and ideologically represented in various local contexts. The standardization of Mandarin on both sides of the Taiwan Strait has led to a widespread notion of linguistic normativity. As a consequence, there is a strong tendency in the scholarship to ignore non-normative forms or to label them as regional or wrong. Against this backdrop, our volume closes with the proposal that the sinophone should be conceptualized as a bidirectional lens. On the one hand, it opens our perspective to language variation beyond our familiar and normative categories. On the other hand, the observation of linguistic variation that a sinophone perspective facilitates ought to also make us critically reflect on those same categories. How does our understanding of Mandarin change once we discard the parameters stipulated and defended by official language planning? What kind of transformations does Mandarin undergo due to its use by multilingual speakers? Indeed, how do we account for multilingualism within Mandarin? These questions have a deep historical dimension, as Simmons shows in his analysis of shifting norms of Mandarin in late imperial China. But they also have far-reaching consequences with regard to the analysis of contemporary Mandarin. This argument implies that we should consider linguistic analysis of language variation as a distinct form of multilingual practice. A case in point is the insertion of English elements into spoken Mandarin. Whereas Su’s chapter identifies different attitudes toward this ubiquitous phenomenon, Wasserfall’s chapter looks at the concrete manifestations of English items once they become a part of spoken Taiwan Mandarin. She

Introduction  9 points out that speakers of different backgrounds use and recognize a number of English items, while submitting them to different degrees of morphological and phonological integration. Wasserfall’s chapter complements Su’s analysis by showing that English words function as markers of belonging and form an in-group code. In contrast to Wasserfall’s analysis of two distant and unrelated languages, Lin’s chapter looks at internal Sinitic language contact. She analyzes the effect of language contact on the emergence of Taiwan Mandarin utterance-final particles (UFP), or particles without referential meaning that fulfill several pragmatic functions. Her analysis of the contact-induced emergence of UFP in Taiwan Mandarin not only relies on language data but also reaches out to the social history of the speakers of different Sinitic varieties and social mechanisms of language contact. Our volume is rounded off with Wiedenhof’s chapter on “quirky” phenomena in Mandarin, by which he means instances of rare or exceptional usage which tends to escape linguistic attention. There is much variation within Mandarin which has so far remained undocumented due to the social, cultural, and educational principles that inform conventional grammar writing, including Mandarin locative markers, nominal predicates, endearment tones, and the /r/ phoneme. The existence of these quirky phenomena, Wiedenhof argues, suggests that new conceptualizations are needed in the description of Mandarin. From the perspective of language standardization, the inherently dynamic qualities of language call for a critical reflection on the temporary limitations of a linguistic standard. As a whole, this volume demonstrates that if brought together, sociolinguistics and history have a lot to contribute to our understanding of the communities, polities, and beliefs that form the “sinophone world.” Like much of sinophone studies, the book focuses on the littoral and overseas periphery of the Chinese-speaking world. It mainly treats areas that migration, colonialism, and the divisions of the Cold War have placed beyond the scope of the simple term “Chinese.” Within this purview, the chapters of the book lay bare the sometimes gaping discrepancies between, for instance, language policy and actual language use or between conflicting but contemporaneous understandings of the linguistic norm. It is clear, however, that such contradictions are not unique to the polities and communities studied in these pages. The various imperial formatio ns from which the sinophone world emerged—the Qing empire among them—were similarly characterized by an unstable and evolving multilingualism. Future studies will explore the dynamics of linguistic hierarchies, government plans, and the language use of various communities that are located in the very “center” from which “sinophone studies” initially aspired to distance itself. Linguistic diversity in the PRC—within and beyond what scholars would label “Sinitic” and speakers “Chinese”—deserves more detailed study in historical depth. The same holds for late imperial China and its hinterland.

10  Henning Klöter and Mårten Söderblom Saarela

References Chow, Rey. 2013. “On Chineseness as a Theoretical Problem.” In Sinophone Studies: A Critical Reader, edited by Shu-mei Shih, Chien-hsin Tsai, and Brian Bernards, 43–56. New York: Columbia University Press. Karasawa Yasuhiko 唐澤靖彦. 1996. “Teisei goki Chūgoku ni okeru hanashikotoba no kōyō (ichi): Kanwa no shakaiteki yakuwari” 帝政後期中国における話しことば の効用（１）―官話の社会的役割 [The use of spoken languages in late imperial China (one): The social role of Guānhuà]. Chūgoku tetsugaku kenkyū 10: 105–47. King, Ross. 2015. “Ditching ‘Diglossia’: Describing Ecologies of the Spoken and Inscribed in Pre-modern Korea.” Sungkyun Journal of East Asian Studies 15, no. 1: 1–19. McDonald, Edward. 2011. Learning Chinese, Turning Chinese: Challenges to Becoming Sinophone in a Globalised World. London: Routledge. Shih, Shu-mei. 2013a. “Against Diaspora: The Sinophone as Places of Cultural Production.” In Sinophone Studies: A Critical Reader, edited by Shu-mei Shih, Chienhsin Tsai, and Brian Bernards, 25–42. New York: Columbia University Press. ———. 2013b. “Introduction: What Is Sinophone Studies?” In Sinophone Studies: A Critical Reader, edited by Shu-mei Shih, Chien-hsin Tsai, and Brian Bernards, 1–16. New York: Columbia University Press.

Part I

Historical Trajectories

1

What Was Standard Chinese in the Nineteenth Century? Divergent Views in the Times of Transition Richard VanNess Simmons Abstract In late imperial China, the most widely spoken form of the Mandarin lingua franca called Guānhuà was the Nanjing-type southern Mandarin language with five tones. Prior to the nineteenth century, Western students of spoken Chinese generally studied this five-tone Mandarin, a practice which had started with the efforts of the Jesuit missionaries to learn Chinese. But in the mid-nineteenth century, many Westerners began to advocate that the language of Beijing should be the focus of their study. In the same period, however, the dialect of Beijing carried little prestige among the general Chinese population. By the end of the nineteenth century there was still strong resistance to a Beijing pronunciation standard among Chinese. By analyzing a complex interplay of ideological and social factors, this chapter counters the widespread view that the standard followed the location of the capital.

Introduction The modern conceptualization of China’s so-called standard language in the Qing (1644–1911), and even in the Ming (1368–1644) dynasty, rests on at least three faulty assumptions that prevent an accurate view of the actual historical and linguistic situation. The first is that the national capital is a priori the locus of the spoken standard, which simply follows wherever the capital is established. The second, related, assumption is that the historical koiné known as Guānhuà (官話, ‘the language of the officials’), which came to be known as Mandarin in the West, was thus based on the Beijing dialect as soon as the capital moved to Beijing. A third assumption is that the common spoken standard was identical to the reading pronunciation standard for Classical Chinese. These assumptions have deeply affected the received view of sinophone history and are reflected, for example, in Benjamin A. Elman’s landmark study of the Ming and Qing Chinese examination system, in which he remarks: Beginning in the early Ming, the dominant values, ideas, questions, and debates that prevailed in court and among officials were translated

14  Richard VanNess Simmons into a classical language whose pronunciation was based on the standard Mandarin dialect of the court (kuan-hua 官話) in the capital region in north China (after the Mongol invasion and after the Ming transfer of the primary capital from Nanking in 1415–21) and not on the dialects of the more populous and prosperous south, although a form of “southern” kuan-hua remained in use during the Ming in the parallel ministries that were maintained in Nanking as the southern capital…. Without a competing capital such as Nanking, Peking alone during the Ch’ing dynasty provided the standard language for officials. (Elman 2000: 373–74; emphasis added) The question of the vernacular phonetic basis for the recitation of classical language texts is a complex one. It is by no means simply the case that the pronunciation of the classical language was “based on the standard Mandarin dialect of the court.” A wide variety of factors came into play in determining how to pronounce a classical text, including local reading traditions, traditional rime books and rime tables, as well as each individual speaker’s dialect background.1 Regional conventions and many types of local reading practice allowed for local pronunciation and recitation of texts in dialect. Many people must have also learned some form of Guānhuà pronunciation that could be used in recitation as well and that was also useful to know for mastery of the Mandarin-based lingua franca which was necessary for communication beyond one’s local community. This lingua franca is the focus of the present discussion, which seeks a more nuanced picture of the linguistic and geographical basis of the “standard Mandarin dialect” in Qing times, the “standard language for officials” in the period, and Guānhuà in both its northern and southern forms.2 This chapter examines these linguistic standards and varieties from the perspective of the mid to late nineteenth and early twentieth centuries, and uncovers a set of assumptions, practices, and attitudes that clarify the nature and shape of the “standard Mandarin dialect” of late imperial China. We use the term “standard” in this discussion loosely to encompass a range of linguistic norms and attitudes that held sway, and slowly evolved, in China and the sinophone world from premodern times through the late nineteenth century and into the early twentieth century. Prior to the nineteenth century, the standard was a vague set of conventions for speaking and writing that the educated literati generally held to be most preferred and most useful for oral and written communication throughout China’s vast territory and across the many dialects and varieties of spoken Chinese. The various premodern conventional spoken norms were never precisely or rigorously codified by the government prior to the twentieth century. Toward the end of the nineteenth century, as the Chinese learned of national language standards that were being developed in Europe, Japan, and elsewhere, they began increasingly to call on the government to develop a

Standard Chinese in the Nineteenth Century  15 national language standard for China too. This would ultimately come to be a governmentally codified standard. To better understand the evolution of the traditional Guānhuà norms, and attitudes toward them, in the late nineteenth and early twentieth centuries, we will look at the linguistic specifics that entered the debate regarding exactly what the modern official standard should be.3 The results of our examination bring us to the following conclusions: 1 2 3 4 5

The conventional pronunciation norm most widely promulgated in the Qing is a five-tone Mandarin that is best characterized as the southern type. Various permutations of this norm were prevalent throughout the Qing and even persisted into the Republican period. While the Manchu court may have preferred Beijing pronunciation,4 the city’s dialect was otherwise widely dismissed among the Han Chinese, including the literati elite. Westerners clearly turned to Beijing as the standard only after the Taiping rebellion (1851–64). Chinese acceptance of a Beijing standard did not come until decades later.

The Qing Standard and Its Characteristic Phonology From the perspective of dialect evolution, Qing-period Mandarin was not a single language. Since well before the Ming, the Mandarin dialects already evidenced great diversity.5 The oldest and deepest split within the Mandarin dialects is that between northern and southern types. Though similar in many ways, there are clear and significant phonological differences between these two types, which will become clear as our discussion unfolds. In historical perspective, the southern type is older and more conservative, while the northern type is an innovative variety that made its first encroachment deep into Chinese territory only following the collapse of the Northern Song (960–1127) and the Jin (1115–1234) takeover of north China. Prior to that, the southern type was dominant in the central plains, including in Kaifeng, the Northern Song capital, until its speakers fled south with the dynasty. Later, the descendant of this southern type was widely spoken in the regions of modern Anhui and Jiangsu, the territories from which Zhū Yuánzhāng 朱元璋 (1328–98) marched forth to expel the Mongol Yuan (1271–1368) dynasty and establish the Ming dynasty. The language of the capital of Zhū Yuánzhāng’s new Ming dynasty, Nanjing, was a variety of this southern type, which thus can also be called the Jiāngnán 江南 (‘south of the Yangtze’) type.6 This Jiāngnán Mandarin carried forward the prestige of the southern Mandarin that had been established in the Song and came to serve as the

16  Richard VanNess Simmons exemplar for the Mandarin-based lingua franca spoken by officials in the Ming—Guānhuà.7 The prestige of the language is demonstrated by the preference for Guānhuà pronunciation and usage, particularly that of southern Mandarin, among the educated literati and others of elevated social status and mobility in their conversation and reading practices, as well as in their interactions across dialect and other linguistic boundaries. Our discussion below provides a number of examples that illustrate this preference and the various forms it takes. The wide geographical distribution and deep historical roots of Ming-period Guānhuà, dating back to its northern Song origins, added to the prestige that Mandarin gained through association with the Ming dynasty’s founding emperor. This prestige and the accompanying influence held fast even following the move of the Ming capital to Beijing in 1421 and through the end of the Qing dynasty. Though it had taken deep root in the Ming bureaucracy right from the start, this preferred language of the officials is not explicitly identified as “Guānhuà” in written records until decades after the move to Beijing, in the late fifteenth or early sixteenth century. One of the earliest mentions is in the Sìyǒuzhāi cóngshuō (四友齋叢說; Collected discussions from the Four Friends Studio) by Hé Liángjùn 何良俊 (1506–73), who says that “Yǎyí [i.e. Wáng Chǒng 王寵 (1471–1533) of Suzhou] did not like to employ his hometown language; when speaking he always used Guānhuà” (Hé 1995–2002: 619).8 This concept of Guānhuà had coalesced around the language that had been spread throughout the empire by the Ming bureaucracy and military and that had come to serve as the lingua franca of government officials, merchants, soldiers and their officers, and others who traveled beyond their own local communities. While there is no evidence for a single, discrete term referring to this lingua franca prior to the late Ming, the idea of a correct and elegant reading pronunciation did have earlier characterizations. For instance, the imperially commissioned early Ming southern Mandarin rime book Hóngwǔ zhèngyùn (洪武正韻; Rectified rimes of the Hóngwǔ reign [1328–98]) makes note of “correct pronunciation” (正音) and “elegant vocalization of the central plains” (中原雅聲) in its introductory discussion (Yuè and Sòng 1983: 4, 6).9 In the mid-seventeenth century a non-Chinese ethnic group, the Manchus, swept in from the northeast to push out the Ming rulers and establish the Qing dynasty. By this time, southern Mandarin-based Guānhuà had become deeply and firmly entrenched as the prestigious lingua franca, or common koiné—the common supra-regional spoken language—of the native Chinese Han literati. So much so that the new non-Chinese rulers yielded to its prestige and essentially codified it in the Imperial Dictionary that was compiled under the fourth Qing emperor, Kāngxī 康熙 (r. 1661–1722): the Kāngxī zìdiǎn (康熙字典) completed in 1716 under the direction of Zhāng Yùshū 張玉書 (1642–1711; Zhāng Yùshū 1985; hereafter: Kāngxī imperial dictionary). It is no surprise, then, that the editor of the writings of the French Jesuit mission to China, Jean-Baptiste du Halde (1674–1743), published a

Standard Chinese in the Nineteenth Century  17 description of Jiāngnán (“Kiang nan”) Mandarin as “convenient for the government” in 1735: The Mandarin-Language is properly that which was formerly spoken at court in the province of Kiang nan, and spread into the other Provinces among the polite People; and hence it is that this Language is better spoken in the Provinces adjoining to Kiang nan then in the others, but by slow degrees it was introduced in all Parts of the Empire, which is very convenient for the government. (Du Halde 1741: 389–90; emphasis added) This language is reflected in the Kāngxī imperial dictionary in the first of two rime tables which the Qing compilers included in the introductory material to the dictionary.10 (The second rime table in the dictionary is a traditional Qièyùn 切韻 rime table conventionally considered to reflect Middle Chinese, seventh century, reading pronunciation.) It is unnecessary, and perhaps impossible, to identify a specific dialect that underlies the Mandarin rime table in the dictionary. It is even entirely possible that the table was not based on any single dialect. But it reflects features that are characteristic of the Jiāngnán prestige language and that were also taken to be indispensable elements of the commonly accepted colloquial standard. The principal southern Mandarin features that can be discerned in the rime table of the Kāngxī imperial dictionary are: •

The shift of Common Chinese voiced obstruent initials to voiceless aspirated initials • In a two register píng 平 tone.

•

The loss of Common Chinese final consonants -p, -t, and -k • With rù 入 tone preserved as a category, having a weak final glottal stop -ʔ.11

•

A resulting system of five tones: • Upper (shàng 上, qīng 清, or yīn 陰) píng 平, lower (xià 下, zhuó 濁, or yáng 陽) píng 平, shǎng 上, qù 去, and rù 入.

The status of the rù tone constitutes the primary phonological distinction between northern and southern types of Mandarin. Southern Mandarin had the rù tone and northern Mandarin did not.12 Chinese scholars were well aware of this difference very early on. In a late Ming discussion of the regional characteristics of Guānhuà, Zhāng Wèi 張位 (1538–1605) in his Wèn qí jí (問奇集; Collected notes on the unusual, preface 1590) addresses the issue in a section titled Gèdì xīangyīn (各地鄉音; Regional Pronunciations of Various Places): In general, the rù tone is pronounced as a píng tone north of the Yangtze; and there are also many pronunciations that have no corresponding

18  Richard VanNess Simmons Chinese character which cannot be fully listed here. South of the Yangtze the dental sounds frequently suffer from muddled pronunciation; but this is merely a regional accent in Guānhuà. If the various local dialects are used, it is even more difficult to communicate. (Zhāng Wèi 1995–2002: 215) The early Qing preference for the southern Guānhuà koiné is also seen in the Manchus’ adoption of a southern Mandarin velar pronunciation of certain syllables that were palatalized when they first began learning Chinese from speakers in northeast China. The fact that the Manchus adopted this feature in their Chinese transcriptions after their conquest of the Ming is strong corroboration of the prestige of the southern Mandarin that is reflected in the Kāngxī imperial dictionary, the authoritative status of which the Qing court went to severe lengths to maintain (Kaske 2008: 43). China’s ruling elite clearly continued to prefer the Jiāngnán Guānhuà that had dominated in the Ming; and the Manchus were persuaded that it was the preferred prestige form.13 Hence the closest thing there is to a recorded “standard” Guānhuà promulgated widely in the Qing is a five-tone southern type Mandarin. One hundred years after the Kāngxī imperial dictionary was completed, the southern type of Mandarin continued to prevail as the preferred colloquial speech, though the northern type was also heard in both north and south China. Coexisting with northern Mandarin, the Jiāngnán type of Guānhuà had wide currency as the dominant lingua franca throughout the empire. A contemporary description of the situation is provided by Gāo Jìngtíng 高靜亭 (fl. nineteenth c.), the compiler of the first indigenous textbook of Guānhuà, the Zhèngyīn cuōyào (正音撮要; Synopsis of standard pronunciation).14 Gāo’s view of language as reflected in this textbook reveals that Guānhuà speakers in his time did not abandon their own native vernaculars or dialects, but rather used Guānhuà as a Mandarin koiné for communication across China’s vast territories while continuing to use their own local languages at home and with people from the same province.15 Yet while Gāo tells us this Guānhuà koiné encompassed both northern and southern vernaculars (běihuà 北話 and nánhuà 南話), we find that the Mandarin pronunciation represented in his Zhèngyīn cuōyào is firmly of the southern type.16 This is seen in the pronunciation glosses in his textbook, which reveal that his preferred Guānhuà has the typical five tones we saw earlier reflected in the Kāngxī imperial dictionary rime table, and illustrated in Table 1.1.17 Clearly this five-tone type Mandarin had a powerful Table 1.1 The five tones of Qing Guānhuà

yīn 陰 yáng 陽

píng 平

shǎng 上

1 2

3

qù 去

rù 入

4

5

Standard Chinese in the Nineteenth Century  19 normative influence, one that would in fact ensure that the five-tone southern Mandarin would maintain its dominant prestige in the Chinese linguistic consciousness through the end of the Qing and into the twentieth century. The southern Mandarin koiné was thus well established and already at least 500 years old by the start of the nineteenth century, when considered in light of its Song origins noted earlier. Clearly it had not been diminished by the move of the capital in the Ming; nor had it been superseded by the northern type after close to two centuries of Manchu rule in the Qing. Below we will examine Western descriptions and views of the standard in the nineteenth century.

Western Ideas of the Qing Guānhuà Standard The Mandarin Studied by Western Scholars in the Early Nineteenth Century The Qing Guānhuà phonology outlined earlier obviously contrasted with the dialect of Beijing in many ways. Both Chinese and Western scholars of the Chinese language were well aware of this in the nineteenth century. Joseph Edkins (1823–1905), one of the leading scholars of the Chinese languages at the time, noted the wide regional currency of Guānhuà in his 1864 grammar of Mandarin: The native name of the pronunciation used at court, and in public offices is KWAN HWA, or mandarin* dialect. This dialect is in its essential features, the common language of the people in the provinces north of the Yang-tsï-kiang, in Sï-c‘hwen, Yün-nan, Kwei-cheu, and in parts of Hu-nan and Kwang-si. At least, there is sufficient similarity in the sounds employed through this wide extent of country, embracing twothirds of China, to warrant their being called by a common name. *The word mandarin, though Portuguese in origin, is too convenient to be resigned as an English equivalent for Kwan. The term ‘court dialect’ is not extensive enough to represent the public colloquial medium of conversation in all public offices through the country, and among the people throughout two thirds of it. (Edkins 1864: 7) In a slightly later passage Edkins details the differences between Guānhuà and the dialect of the capital: In the Peking dialect, 京話 ching hwa, the words of [the] fourth toneclass are all distributed among the other four classes, with no rule but custom to determine into which they have wandered… The finals n, ng, are kept distinct from each other after all vowels, and are the only

20  Richard VanNess Simmons consonants that can terminate a word. The initials h and k, when they stand before i or ü, change to s and ts (or ch).* *Scholars who are natives of Peking distinguish the metropolitan dialect from the Kwan-hwa. Sounds used in reading, and words found in printed mandarin books, form the Kwan-hwa. Sounds not used in reading and words not found in books are referred to [sic] the local dialect. (Ibid., 8) The “fourth tone-class” refers to the rù tone, which is lost in Beijing and most northern Mandarin dialects. Following this, Edkins provides a broad phonological definition of Mandarin: Mandarin, so far as sound is concerned, may be defined as that system which has either five or four tones, has only n and ng for consonant finals, and is wanting in the letters g, d, b, z, v, among its initials. (Ibid., 9) Of this definition, the only feature that distinguishes southern Mandarin from northern Mandarin is the number of tones. Northern and southern Mandarin types both have lost the earlier final consonants *-p, *-t, *-k, and *-m, leaving only -n and -ng; and both varieties of Mandarin have no voiced initials, which is what Edkins means when he says that Mandarin “is wanting in the letters g, d, b, z, v, among its initials.” But with regard to tone, those Mandarin varieties with four tones (upper píng, lower píng, shǎng, and qù), and lacking the rù tone, would belong to the northern type; those with five tones, including rù, would belong to the southern type. Though it had existed alongside the southern type for a long time as a less prestigious form of the Mandarin lingua franca, Western scholars of Chinese only began to take serious note of the northern type with four tones as a possible preferred alternative to the southern Jiāngnán type in the mid-nineteenth century, only a decade or two before Edkins penned the earlier passages. The earliest version of Mandarin documented by Western scholars is that in a lost Portuguese-Chinese dictionary that was compiled by Matteo Ricci (1552–1610) together with Lazzaro Cattaneo (1560–1640) in 1598–99; it had the five tones of southern Mandarin (Yang 1989, 208–10).18 Soon thereafter Nicolas Trigault (1577–1628) adopted a revised version of the Ricci- Cattaneo five-tone Mandarin Romanization in his Xīrú ěrmù zī (西儒耳目資; An aid to the eyes and ears of the Western scholar). The Ricci-Cattaneo five-tone system of diacritics for Mandarin was then widely adopted by later scholars and missionaries, including Martino Martini (1614–61), Francisco Varo (1627–87), Prospero Intorcetta (1626–96), and Joseph-Henri de Prémare (1666–1736).19 In the nineteenth century Robert Morrison (1782–1834) made note of the traditional Mandarin system of five tones in a short grammar of Chinese that he compiled (Meadows 1847: 61; Morrison 1815: 20) and his dictionary included the rù tone together with píng, shǎng, and qù (Morrison

Standard Chinese in the Nineteenth Century  21 20

1815–23: I/I, xvii). Séraphin Couvreur (1835–1919) and Georg von der Gabelentz (1840–93) also used the Ricci-Cattaneo five-tone system in their Romanizations (Luó 1934, table 3). A preference for the four-tone northern Mandarin does not appear in the Western scholarly community until the late nineteenth century. Among Western grammars and dictionaries that can serve as reliable sources for either southern or northern types of Mandarin in the Qing, all of the sources for the northern four-tone type date from 1864 and later (Coblin 2003: 199), i.e. Edkins (1864), Wade (1867), Parker (ca. 1870, in Giles 1892), Stent (1871). A couple of decades earlier, in 1844, Samuel Wells Williams had issued his English and Chinese Vocabulary in the Court Dialect. But Williams’ volume actually presented only a variety of southern Mandarin with a clearly separate and distinct rù tone. Subsequently, in 1846, Robert Thom (1807– 1846) published his Extracts from Works Written in the Mandarin Language, as Spoken at Peking, which transcribed and translated lessons from Gāo Jìngtíng’s Zhèngyīn cuōyào and other colloquial texts (Thom 1846). However, though he is said to have had a good spoken command of the Beijing dialect (Kaske 2008: 68), Thom did not give any details about pronunciation or tones in his text. The earliest to publish an accurate guide for the study of the Mandarin of Beijing appears to have been Thomas Taylor Meadows (d. 1868). Meadows advocated learning the “court pronunciation”—that of “Pekin”—and the capital’s four-tone system with a new Romanization he developed for that purpose in an eclectic volume about China which he penned in 1847, Desultory Notes on the Government and People of China.21 In support of his advice to learn the dialect of the capital, Meadows claimed that one-third of the government officials stationed in Guangdong 廣東 in 1844, 74 out of 231, were Beijing natives. He further advised against attempting to learn both the “Nankin and Pekin colloquial,” reasoning as follows: As it is quite impossible to attain a practical proficiency in any two variations of the colloquial, it will certainly be found most advisable for the greater part of those who intend studying the language with the view to putting their knowledge to actual use in business, to apply themselves to the Pekin alone, as a spoken medium. (Meadows 1847: 45–46) Evaluating Meadows’ advice, the editors of the Chinese Repository conceded that the capital city’s dialect was “the most useful for those who expect to come in contact solely or chiefly with official people.” But they doubted that it was widely useful for spoken communication in China in general (Bridgeman and Williams 1848: 93). Introducing the tones of the Beijing colloquial, Meadows noted that “in the court pronunciation, only four shěng are heard; in the Nankin, five.”22 Meadows’ Romanization marked the four tones of Beijing with the numbers

22  Richard VanNess Simmons 1 to 4. A decade later, in developing his earlier materials for the study of the Beijing dialect, Thomas Francis Wade (1818–95) formally adopted Meadows’ method of marking the tones with four tone numbers for his own Romanization. Wade attributed his inspiration to Meadows: “It is simplest, as Mr. Meadows suggests, to distinguish the four tones of Pekingese by numbers. I write the shêng [‘tones’] of the syllables pa, accordingly as follows: pa1, pa 2, pa3, pa4 ” (Wade 1859: 85). The Impact of the Taiping Rebellion on the Western Standard Precisely in the decade and a half from 1851 to 1864, Edkins, Meadows, and Wade were among the cultural and linguistic principals in the drama playing out in China around the Western community’s response to the Taiping rebellion. The leader of the uprising, Hóng Xiùquán 洪秀全 (1814–64), announced the founding of Tàipíng Tiānguó 太平天國 in 1851. Hóng led his armies to conquer Nanjing in 1853, where he massacred the city’s Manchu population and set up the capital of his Heavenly Kingdom. After a decade of civil war, the Qing finally defeated the Taiping when Zēng Guóquán 曾國荃 (1824–90) and Zēng Guófān 曾国藩 (1811–72) recaptured Nanjing in 1864, the year Hóng died, fatally weakened by disease in the final weeks of his failed empire. As the war between the Qing and the Taiping wreaked disastrous turmoil, destruction, and famine in China’s interior, the foreign community in China debated the merits of Hóng’s Heavenly Kingdom and whether or not to take sides in the civil war: to support the rebels and let the Manchu dynasty fall or to back the imperial armies and prop up the Qing empire. The foreign community was eventually drawn into the fray following the Arrow War (or Second Opium War) in 1856, but they continued to be ambivalent as to the preferred outcome. Wade, from a military background, served as interpreter for James Bruce, the 8th Earl of Elgin, who was British High Commissioner in China at the time. Despite the British hostilities toward the Qing, Wade had great respect for the Chinese elite and, especially, the Manchu elite and harbored deep contempt for the rebels, who were from poorer classes in southern China (Platt 2012: 41). Edkins and Meadows on the other hand were more sympathetic to the rebels. Edkins was a missionary affiliated with the London Missionary Society. He and his friend James Legge (1815–97) had become acquainted with Hóng Xiùquán’s cousin Hóng Réngān 洪仁玕 (1822–64) during Edkins’ first posting in Hong Kong in the late 1840s. Edkins used this connection to make contact with the Taiping leaders in Suzhou and Nanjing beginning in 1860, following which he provided the Western community with glowingly positive reports about the rebels (Platt 2012: 78–82, 221–22). Meadows was a British consul with extensive experience in China who believed that the Taiping would ultimately be successful in overthrowing the Qing and advocated letting the rebellion take its course without foreign intervention (Platt 2012: 90–91).

Standard Chinese in the Nineteenth Century  23 In the final years of the civil war, a group of Western allies led by the British joined forces with the imperial troops in a series of battles that prevented the Taiping from controlling Shanghai, finally tipping the balance in favor of the Qing. The ultimate outcome was the utter defeat of the Taiping and the destruction of their capital in Nanjing in 1864. In the decade and a half of civil war, the rebellion had also laid waste to vast areas in central China, including much of the Jiāngnán region, southern Mandarin’s native territory.23 Wade, Edkins, and Meadows were thus deeply involved in contemporary events in China in the mid-nineteenth century and the events of the civil war that so severely threatened the Manchu control of the dynasty. Wade’s linguistic leaning, and his clear preference for the Beijing dialect and the language of the capital’s ruling class, may have been influenced by his political views. Edkins and Meadows, whose sympathies lay more with the rebels and their leaders, all from southern China, must have focused on the capital’s language for more pragmatic reasons. Following the destruction of Nanjing, the three of them and many others would have had to determine which spoken language would be best to put their effort into learning and mastering. Though Nanjing had long been the designated exemplar, and historical home, of the southern type Mandarin, with the virtual decimation of the southern capital’s population and the reestablishment of central imperial authority in Beijing, the northern capital increasingly seemed the obvious choice. Just three years after the Manchu victory in the civil war, in 1867, Thomas Francis Wade published his landmark Mandarin textbook Yü-yen tzŭ-erh chi (語言自邇集; Collected lessons in language beginning with the close at hand), to which he gave a long English title: A Progressive Course Designed to Assist the Student of Colloquial Chinese as Spoken in the Capital and the Metropolitan Department. Wade noted in his preface that “Pekinese is the dialect an official interpreter ought to learn. Since the establishment of foreign legations with their corps of students at Peking, it has become next to impossible that any other should take precedence” (1867: vi). Wade’s textbook spearheaded a shift to the study of the Beijing dialect as the preferred spoken standard for non-Chinese to learn. The Japanese followed this trend and Chinese language instruction in Japan turned from Nanjing dialect based to Beijing dialect based in 1876. Though this shift “was not too late in light of the contemporary world situation,” as concluded by Rokkaku Tsunehiro, the change was so abrupt and unexpected that no Beijing language textbooks were available in Japan. The result was that it took five years to implement the change. Wade’s Beijing-based textbook was at first widely adopted in Japan during the initial stages of the transition to the study of the capital city’s dialect (Andō 1991: 27–29; Rokkaku 1961: 47, 1992: 92). Written half a century earlier, Robert Morrison’s observation regarding the ascendancy of the speech of the Qing dynasty’s capital seems to have

24  Richard VanNess Simmons been prescient, though he had adopted the dialect of Nanjing as the standard for his 1815 dictionary: What is called the Mandarin Dialect, 官話 or Kwan hwa, is spoken generally in 江南 Keang-nan, and 河南 Ho-nan, Provinces, in both of which, the Court once resided; hence the Dialects of those places gained the ascendancy over the other Provincial Dialects, on the common principle of the Court Dialect becoming, amongst People of education, the standard Dialect. A Tartar-Chinese Dialect is now gradually gaining ground, and if the Dynasty continues long, will finally prevail. (Morrison 1815–23: I/I, x) At the beginning of the following century, K. Hemeling (1878–1925) appraised the outcome of the subsequent evolution of Mandarin and the effects of the Taiping rebellion in the preface to his 1907 study of the Nanjing dialect Die Nanking Kuanhua: The ascendancy doubtless gained by the Nanking Kuanhua over other ‘Mandarin’ dialects during the epochs when Nanking was the Capital of the Chinese Empire and probably retained for some time after the removal of the central government to Peking in 1421 A. D. by Emperor Yunglo, had in the course of centuries, since gone by, gradually lost ground. It received its death-blow through the massacres, nearly extirpating the inhabitants of the unfortunate Nanking, perpetrated after its capture, by the T’aip’ing rebels whose Capital it subsequently was from 1853–1864, and again in a worse degree by the Imperial forces when recapturing it. The few handfuls of old resident families that escaped destruction being insufficient to absorb the rush of immigrants from other parts of China after the suppression of the rebellion, the standard pronunciation cultivated by them could not penetrate throughout this city, and even to this day the number of its inhabitants speaking a pure and unadulterated Nanking Kuanhua is comparatively small. The days of splendour are gone never to return. The Kuanhua of Nanking is no longer a serious rival for supremacy of that of Peking: the latter, purified of its localisms, is at present the only standard spoken language recognised throughout the whole Empire, a fact not to be wondered at, if one takes into consideration that Peking has been uninterruptedly the Capital of the Chinese Empire since 1421 A. D. (Hemeling 1907: 2; emphasis added) Western Ambivalence about the Beijing Standard Joseph Edkins, who as a missionary quite likely interacted more with all levels of Chinese society than Wade or Meadows, recommended that “the Peking dialect must be studied by those who would speak the language of

Standard Chinese in the Nineteenth Century  25 the imperial court, and what is, when purified of its localisms, the accredited kwan hwa of the empire.” However, he also noted that “the Peking dialect is more fashionable, but that of Nanking is more widely understood,” because the vernacular of the northern capital was “far removed from the analogies of the dialects in the southern half of the country” (1864: 10). For that reason, Edkins was unwilling to completely abandon the Nanjing dialect in his Mandarin grammar and made only supplementary, though frequent, reference to Beijing pronunciation and usage. Similarly, Georg von der Gabelentz (1840–93) in his 1881 grammar of Chinese was loath to accept the Beijing dialect as standard for serious study of Chinese and recommended “the adoption for scientific purposes of the ‘Mandarin’ dialect reproduced in the writings of the Jesuits of the 17th and 18th centuries, which he considers to have been that of Nanking” (Hemeling 1907: 2). The most vehement vocal disapproval of the Beijing standard may be that of the Dutch Sinologist Gustave Schlegel (1840–1903) in his stridently written introduction to a reprint of an article that W. F. Mayers (1839–78), a British sinologist, had penned in 1867. Mayers had strongly objected to Thomas Francis Wade’s adoption of a Beijing-based Romanization in his Hsin Ching Lu (Xún jīn lù 尋津錄; Record of a search for a ford) of 1859 (given the English title Book of Experiments), and argued in great detail in favor of the superiority of a Nanjing-based system. Schlegel wrote: The late lamented W. F. Mayers, one of the best sinologues England can boast of, broke a lance against the abuse of the Peking system of orthography which then already (in 1867) became threatening. Nearly 30 years have since elapsed and, we are sorry to say, this abuse has thriven like an obnoxious weed, and menaces to overwhelm and smother entirely the standard pronunciation (正音) expressed in the Imperial Dictionary of K‘ang-hi. The younger generation of sinologues does not seem to study the work of their predecessors, and gets as infatuated, (since Peking has been opened) with the bad, half Manchurian, pronunciation, as the Peking Chinese themselves, … the elder Sinologues, Morrison, Medhurst, Bridgman, … Rémusat, … all of them have used the standard pronunciation in their transcriptions, and we do not see why the present generation should adopt the local brogue of Peking as the standard pronunciation for the whole Chinese language. (Schlegel 1895; 499–500; emphasis as in the original) Schlegel then reproduced Mayers’ arguments in their entirety. Mayers had concluded, Enthusiastic admirers of the Peking colloquial may perhaps object to a proposition tending to suggest a limit of the influence of that local

26  Richard VanNess Simmons dialect over the general language of China; but … [this] writer must remain steadfast in his belief in the older consonantal system for general use. (Schlegel 1895: 507) By “the older consonantal system” Mayers is referring to the southern Mandarin retention of velars before high front vowels that we noted the Manchus had also adopted (see earlier section). We find the issue raised again in 1888 in a widely discussed exchange between the Reverends Absalom Sydenstricker (1852–1931) and A. Purist in the pages of the Chinese Recorder under the title “Romanizing the Official Dialect”: Rev. A. Sydenstricker: “Now it is very well known that there is no form of pronunciation of Mandarin that is universal — not even of the “court dialect.” Rev. A. Purist: “The Nankingese official dialect is far preferable to any colloquial… For all practical purposes the Mandarin of the Kiangsu province can be easily understood by most people in a district as large as eight of the American States.” Rev. A. Sydenstricker: “Again, Pekingese commonly goes by the name of the ‘court dialect.’ Such being the case, how can Nankingese be termed the ‘official dialect?’ As to the wide use of Pekingese, so good an authority as Wade brings witness to show that it is ‘perfectly well understood in Hankow,’ and missionaries living in Nanking itself say that the natives are aping Pekingese even there. My own limited observation confirms to me the wide and increasing prevalence of the Pekingese form of Mandarin.” (Sydenstricker and Purist 1888: paragraphs from pages 37, 135, 301) Sydenstricker’s last word in this exchange underscores the inevitable trend toward the dialect of Beijing as the preferred variety of Mandarin for non-Chinese to learn.24

Chinese Conceptions of the Guānhuà Standard in the Late Qing The linguistic concerns of the Chinese themselves after the Taiping rebellion did not follow the trends of the Westerners, and most people continued with their traditional and regionally based views of what Guānhuà should be. A widespread disregard of the Beijing dialect persisted unabated among speakers of southern Mandarin dialects after the civil war. The continuing preference for the old Guānhuà standard can be discerned in the various orthographies invented by the Chinese during the script reform movement (known as the Qièyīnzì yùndòng 切音字運動) of the late nineteenth and early twentieth centuries.

Standard Chinese in the Nineteenth Century  27 With the heightened urgency for reform following the Sino-Japanese war in 1894–95, and calls for improving education and increasing literacy throughout China, many Chinese began enthusiastically to explore writing reform and ways to overcome the difficulties of learning the Chinese characters of the traditional writing system. At the same time contact in the nineteenth century with various national language standards in Europe and elsewhere gave rise to calls for the development of a national language standard in which speech and writing were more closely aligned. Having seen the viability of various Western missionary Romanizations of Chinese dialects and the success of alphabetic writing in achieving literacy among the non-elite members of society, many Chinese seized upon the idea of developing phonetic orthographies for Chinese as a possible path to writing reform.25 In the years between 1896 and 1911 a flurry of inventiveness ensued. Elisabeth Kaske calculated that upward of 29 different schemes were developed by 24 Chinese scholars (2008: 90–159). Some were alphabetic; some were based on the structure of Chinese characters (usually inspired by Japanese kana); some were in imitation of Western forms of stenography or shorthand; still others were purely new inventions—some almost whimsical in form. As they were intended to be phonetic scripts, a specific spoken language had to be chosen whose pronunciation the script would represent. Thus, many of the schemes served as orthographies for various dialects; and some were even intended to be pan-dialectal by making allowances for dialect variation. But orthographies designed to represent varieties of Mandarin were the most strongly represented. Our examination of the qièyīnzì orthographies produced in the late Qing period found 11 that were clearly intended to write a form of Mandarin, all listed in Table 1.2. Heretofore little careful attention has been given to what specific type of Mandarin these orthographies represented. The designers’ identification of the type of Mandarin each is intended to write can be somewhat ambiguous. But examined closely, we find that the number of tones each scheme represents in their system allows us to classify them as southern Mandarin, with five tones, or northern Mandarin, with four tones (and no rù tone). Of the 11 schemes examined, three are firmly northern Mandarin with four tones, one—Liú Mèngyáng’s—is intended to write Beijing pronunciation but provides indication for five tones, and the remaining seven orthographic schemes—a clear majority—are unambiguously the southern type Mandarin with five tones, including a rù tone. Cài Xīyǒng’s 蔡錫勇 (1850–96) stenographic scheme was probably the earliest indigenous Mandarin phonetic orthography. His system indicates the five tones by the position of his shorthand letters in relation to a vertical line. Wáng Bǐngyào’s 王炳耀 (fl. 1896) system was a stenographic type system designed to write forms of Cantonese, Mǐn dialects, and Hakka as well as “Běiyīn 北音” (‘Northern’) vernaculars. It included diacritics to mark upper shǎng and qù, and lower píng, shǎng, qù, and rù, with upper píng and

28  Richard VanNess Simmons Table 1.2 Mandarin phonetic schemes in the late Qing

1 2 3 4 5 6 7

8 9 10 11

Author

Year Title

Type

Self ID

No. of tones

Cài Xīyǒng 蔡錫 勇 (1850–96) Wáng Bǐngyào 王炳耀 Wáng Zhào 王 照 (1859–1933) Lì Jiésān 力捷三

1896 Chuán yīn kuài zì 傳 音快字 1896 Pīnyīn zìpǔ 拼音字譜

Stenographic

Guānyīn 官音 5

Stenographic

Běiyīn 北音

4

1900 Guānhuà héshēng Chinese stroke Jīnghuà 京話 4 zìmǔ 官話合聲字母 1902 Wú shī zìtōng qièyīn Stenographic Guānhuà 官 5 Guānhuà zìshū 無師 话 自通切音官話字書 1906 Zēngdìng héshēng jiǎn Chinese stroke Nányīn 南音 5 zìpǔ 增訂合聲簡字譜

Láo Nǎixuān 勞乃宣 (1843–1921) Tián Tíngjùn 田 1906 Shùmù dàizì jué 数目 廷俊 代字訣 Lú 1906 Zhōngguó zìmǔ Zhuàngzhāng Beijing qièyīn hé 盧戇章 dìng 中國字母北京切 (1854–1928) 音合訂, and Beijing qièyīn jiàokēshū 北京切音教科書 Liú Mèngyáng 1908 Zhōngguó yīnbiāo 劉孟揚 zìshū 中國音標字書 (1877–1943) Jiāng Kànghǔ 1908 Tōng zì 通字 江亢虎 (1883–1954) Liú Shì’ēn 刘 1909 Yīnyùn jìhào 音韻 世恩 記號 Huáng Xūbái 黄 1909 Lādīngwén yìjiě 拉丁 虚白 文臆解

Húběiyīn 湖 北音 Chinese stroke Beijing 北京

5

Latin alphabet Jīngyīn 京音

4 or 5

Latin alphabet Unknown

5

Self designed ? symbols Latin alphabet Guānhuà 官 话

5

4

5

rù unmarked and rù further distinguished by a final consonant. But Wáng Bǐngyào explicitly indicates that Běiyīn has only four tones and thus appears to have been the earliest indigenous system equipped to write a fourtone northern Mandarin. Wáng Zhào’s 王照 (1859–1933) phonetic alphabet was inspired by Japanese kana (仮名) during the time he spent in Japan and had symbols for initials and finals of syllables based on the traditional fǎnqiè (反切) analysis. His system marked the four tones of jīnghuà (京話, ‘Capital language’) with a modified version of a traditional four-corner system that eliminated the rù tone. Wáng Zhào’s was probably the earliest indigenous orthography that unequivocally took Beijing as its standard. Láo Nǎixuān’s 勞乃宣 (1843–1921) scheme was designed to supplement Wáng Zhào’s system and enable it to write southern Mandarin (nányīn 南音),

Standard Chinese in the Nineteenth Century  29 and five tones: Láo added a rù tone mark to Wáng’s modified four-corner system: a leftward falling stroke (piě 撇) in the lower right. Lú Zhuàngzhāng (also rendered Gàngzhāng) 盧戇章 (1854–1928) had designed a variety of schemes for Mǐn dialects in the early 1890s using symbols derived from the Roman alphabet. Then, following a three-year stay in Taiwan at the turn of the century where he was inspired by Japanese kana, Lú invented a system of symbols that could be used to write Mǐn and Cantonese dialects as well as the language of Beijing. For the Beijing orthography, he too used a modified four-corner system to indicate the four tones with small circles. In 1905, Lú Zhuàngzhāng submitted his Beijing orthography to the Qing Ministry of Education (學部). But after review it was rejected for having the following defects: (1) it was awkward to write, (2) its set of initials was incomplete, and (3) it did not indicate the rù tone (Lín Tāo 2010: 427; Lú Zhuàngzhāng 2015; Ministry of Education 1958: 70–71). The systems designed by Liú Mèngyáng 劉孟揚 (1877–1943) and Huáng Xūbái 黄虚白 were both based on the Roman alphabet, with no modifications. Liú designed his system to write the Beijing dialect and used diacritics to mark tones—in fact his diacritics were eventually adopted in Mandarin Phonetic Symbols (注音符號) and later in Hànyǔ Pīnyīn. But Liú included a diacritic for the rù tone as well, with an explanation that “it is preferable to have the tone, just in case, rather than leaving it out with no mention” (甯可 備而不用，不能缺而不講; Liú Mèngyáng 1957: 6). Huáng Xūbái’s was a fivetone Guānhuà system (Luó 1934, table 3). He marked tone with additional letters placed before or after the syllable. As our earlier summary of the various systems illustrates, the majority of the late nineteenth-century orthographies for Chinese were based on five-tone Mandarin. This situation reveals a clear collective predilection for southern Guānhuà among the educated elite in the late Qing. The preference for Jiāngnán Mandarin reflected in these orthographies is also echoed in sociolinguistic attitudes that were broadly current at the time among Han Chinese. These attitudes were informed by a simmering resentment of the Manchu ruling class and the Manchu Banner population in general. This resentment fostered a disparagement of their northern Beijing Mandarin as coarse and uncultured, as noted by a Manchu gentleman of the times in 1903: At present the use of the alphabet to spell the Beijing dialect to assist with what the Chinese characters do not capture, is opposed most strongly by high officials and well-known scholars. It is not due to a lack of humaneness in their hearts, but because there is much that they do not understand when they see such writing. Yet to the northerner, following over two centuries of frequent but subtle fear that we will be denounced for our lowly ignorance, we take the cultured approach in the discussion, not emboldening ourselves to promote the Beijing dialect and engender the ridicule of southerners. (Chángbái Lǎomín 1958: 34)

30  Richard VanNess Simmons Outside the imperial court, the Bannermen and their families were the primary speakers of Beijing dialect. Thus, as the preference for the Beijing dialect began to prevail among Westerners, they had to rely primarily on Bannermen for instruction in the language. But the majority of the Han literati class was quite hostile to the Bannermen and had little use for their dialect. They felt no compunction to learn or promote it. There is no doubt that these sociolinguistic attitudes informed and reinforced the partiality for five-tone southern Mandarin that prevailed in the late Qing. Hence we see that the Chinese idea of elegant Mandarin did not match the Westerners’ penchant for Beijing as a standard at the turn of the twentieth century but rather continued to favor southern Guānhuà.26

Conclusion The preference for the five-tone system of southern Mandarin seen in the majority of the Mandarin Guānhuà orthographies that were proposed in the last years of the Qing indicates a high level of ambivalence in accepting Beijing as the pronunciation standard across a broad spectrum of those concerned with the issue. The rejection of Lú Zhuàngzhāng’s system by the Qing Ministry of Education tells us that even the imperial authorities did not welcome a standard based solely on the dialect of the capital. The Qing authorities in the eighteenth and early nineteenth centuries required those who sat for the Shùntiān 順天, or Metropolitan, prefectural exam in Beijing and other prefectural examinations to pass a pronunciation inspection (審音) and to speak with an acceptable northern accent (北音) (Hirata 2000). But this was primarily to prevent southerners who sat for the exams from passing as northerners and it was often circumvented.27 Imperial partiality for a pure northern accent did not filter down very far even among the elite literati. There is no doubt that the imperial coterie in closest proximity to the emperor preferred the spoken dialect of Beijing, which would have been their native language by the end of the Qing. Thus, the turn to the Beijing standard in the West and Japan came when the capital was opened up and diplomatic exigencies made it prudent to learn the idiom preferred by the imperial court. Yet the imperial connection did not have strong purchase with the general population of the dynasty and broad Chinese acceptance of the capital’s vernacular as the standard form of speech for the whole of China was still a long way off. We close this discussion with a brief anecdote to illustrate the sociolinguistics at play between the Chinese language spoken by the Manchus and the regional colloquial dialects spoken by the Han Chinese just after the end of the Qing28: During the Kāngxī reign, the Qing government set up an imperial garrison at Jīngzhōu 荊州 (in the vicinity of modern Jiānglíng 江陵 in Hubei). The imperial forces built a wall to separate Jīngzhōu City into western and eastern sectors. Han Chinese were moved to the western

Standard Chinese in the Nineteenth Century  31 sector and the eastern sector was given over to the Manchu Banner troops and their families and staff. There was little interaction between the two sectors in the ensuing two centuries; and the eastern sector developed into a Manchu Banner dialect island that spoke the Northern Zhílì 直隸 northern type Mandarin, very close to the Beijing dialect. Throughout the Qing, the Banner community was coddled by the imperial government, fostering a lasting enmity toward them among the Han Chinese of the city. After the collapse of the Qing in the Xinhai revolution of 1911, the Banner community in the eastern sector lost the support of the central government and with it their economic foothold. As a result, most of the Banner descendants and their families quickly became destitute. The wall between the western and eastern sectors was torn down and many of the Manchus were expelled from the city. Those who remained behind eked out a living by tearing down their old residences and selling off the remains, or tried to get by as petty merchants, rickshaw pullers, or through other menial occupations. Due to the lingering Han resentment of the Manchu Bannermen, a checkpoint was once established at the south gate of Jīngzhōu City to identify Manchus on the basis of their accent. Those who entered or exited by the gate were told to read the number 666. Those who pronounced it in the Northern Zhílì Mandarin, as [liou51-liou51-liou51] (liù-liù-liù), instead of the local Jīngzhōu Han dialect pronunciation, [lu51-lu51-lu51] (lù-lù-lù), were arrested and beheaded. Though the claim of beheadings may be apocryphal, still, in order to survive in their drastically degraded circumstances the Manchus of Jīngzhōu quickly applied themselves to learning the local dialect. They could not learn it perfectly, however. So their language evolved into a patois that came to be called ‘eastside accent’ (東邊腔); and they still remained identifiable on the basis of their speech. In an early twentieth-century study, Yū Shǔluán 於曙巒 characterized the situation as follows: The language of the Banner peoples who stayed in Jīngzhōu is of the northern Mandarin type with a southern accent mixed in. It is a particular kind of colloquial pronunciation that has evolved from the mixing of the languages of Beijing and Jīngzhōu. Whether male or female, young or old, it is the same with all, hearing their accent one immediately knows that they are descendants of the Bannermen. (Yū 1926: 48) It was not until six years after Yū wrote this passage, in 1932, that the Ministry of Education of the Republic of China issued the first dictionary officially sanctioned by the Chinese government to set Beijing as the standard for pronunciation, the Gwoin Charngyonq Tzyhhuey (國音常用字彙; Lexicon of frequently used characters in National Pronunciation) (Jiàoyùbù 1932).29 Yet the rù tone was still a featured element in the dictionary. The introduction tells us that “the rù-tone reading should still be kept in addition

32  Richard VanNess Simmons (to the non-rù reading of Beijing)” (入聲的讀法，還應該兼存) (1932, VI). For example, on page 85 of the dictionary the word for “six,” 六 is glossed as liù (spelled “liow” in the dictionary), with an alternate reading also indicated. We find that alternate reading on page 89 and identified as a rù tone reading: lù (“luh”). We can see therein that the linguistic tension represented by the two pronunciations of “six” in the Jīngzhōu anecdote recounted earlier had not yet fully abated. There were still those who preferred the southern lù over the northern liù in certain contexts. Clearly, the lingering prestige of southern Mandarin with its rù tone had a continuing and concrete influence on the standard Chinese language even into the first few decades of the twentieth century. In fact, similar tensions continue today in the sinophone world between those who speak northern and southern accented Mandarin, even though the rù tone is no longer found in either.

Acknowledgments The original draft of this paper was written while the author was the Starr Foundation East Asian Studies Endowment Fund Member at the Institute for Advanced Study (IAS) in Princeton in Spring 2015. It was subsequently presented at The International Symposium on Language Diversity in the Sinophone World held at Göttingen University, June 11–13, 2015. The author is deeply grateful for the support provided by the IAS.

Notes 1 See Hirata (2000, 2002) and Kaske (2008: 40–55) for useful discussions of the nuances of the issue. 2 The use of Mandarin as a common language in the Chinese diaspora that is noted by Goh and Fong in this volume is a circumstance that most likely has its earliest roots in the informal Guānhuà standard of the Qing. The role of lingua franca in the rise of national languages, or the competition between them, in other regions of Asia is discussed by Meierkord and by Siemund and Li, both in this volume. 3 For an in-depth treatment of the political and historical background to the development of a national language standard in China, see Kaske (2008: 1–76). 4 See Söderblom Saarela’s chapter in this volume for a discussion of the Qing ruler’s language attitudes and the role of Manchu in the language politics of the time. 5 This summary overview of the history of Mandarin is based in part on Coblin (2000) and Simmons (2017). 6 The “creole Mandarin” of speakers from Guilin in Guangxi mentioned by Söderblom Saarela in this volume would in fact be a version of the Mandarin that arose in China’s southwest following large-scale migrations from Henan and Anhui during the Ming, which thus had close ties to Jiāngnán Mandarin. 7 Coblin (2007: 7) notes that during most of the Míng/Qīng period the preferred way of pronouncing Guānhuà (hereafter GH) had originated in the Yangtze Watershed. In

Standard Chinese in the Nineteenth Century  33 addition to this, there was also a less prestigious northern type of GH pronunciation. The Chinese of traditional times were well aware of this difference. They referred to the southern type pronunciation as nányīn 南音 and to the northern as běiyīn 北音 (Coblin 1997; Lǐ 1980a; Lǔ 1994; Yang 1989). Also see Lǐ Xīnkuí (1997). 8 See further Tsu-lin Mei’s footnote in Norman (2004: 297) and Lǔ Guóyáo (1994: 295) for additional discussion of the origins of the term Guānhuà. According to Yè Bǎokuí (2001: 4–5) the term Guānhuà appears as early as 1483 in the Korean court chronicle Yijo sillok 李朝實錄 [Veritable records of the Yi dynasty]. 9 Hóngwǔ zhèngyùn was compiled by imperial order during Zhū Yuánzhāng’s reign under the supervision of Yuè Sháofèng 樂韶鳳 (d. 1380) and Sòng Lián 宋 濂 (1310–81) and was completed in 1375. Their reference to the central plains follows the model of the Yuan period pronunciation guide for northern Mandarin dramatic performance titled Zhōngyuán yīnyùn 中原音韻 [Rimes of the central plains] (1341) by Zhōu Déqīng 周德清 (1277–1365), who explains that “the language [of operatic drama] must favor the pronunciation of the central plains” (Zhōu 1983: 661); also see Yè Bǎokuí (2001: 28) and Söderblom Saarela in this volume for further discussion of zhèngyīn and the currency of the concept in the Qing. 10 See Lǐ Xīnkuí (1980b) for a detailed introduction to both of the Kāngxī imperial dictionary rime tables. 11 The glottal stop ending is found on most rù syllables in modern Mandarin dialects that preserve the rù tone in the Jiāng-Huái region and elsewhere. The Kāngxī imperial dictionary also clearly states that the “rù tone is short with an abrupt ending” (入聲短促急收藏) and “makes use of a shortened syllable” (借 短言) where it is found. 12 One can compare the fourteenth-century Zhōngyuán yīnyùn, which has merged the rù syllables into the other tones and represents a Mandarin clearly of the northern type. 13 See Simmons (2019: 25–28) for further details on the southern nature of the Mandarin in the Kāngxī imperial dictionary and the Manchus’ preference for certain southern features of pronunciation. Coblin (2003) provides an extensive discussion of the influence of southern Guānhuà on the Mandarin learned by the Manchus as reflected in the Qīngwén qǐméng (清文啟蒙; A Manchu primer), a Manchu textbook for Chinese readers first published in 1730. Elements of the issue are also discussed in Yamazaki Masato (1990). 14 Identified as “the earliest standard pronunciation textbook of the Qing dynasty” by Wáng Wèimín (2006: 53). 15 See Simmons (2019: 28–29) for further details of Gāo’s description of the linguistic milieu in early nineteenth-century China. As Snow in this volume notes, the widespread use of dialect at the local level was the primary impetus for the Protestant missionaries to learn local dialects and to develop Romanizations for them such as for the Mǐn dialects in Fujian. 16 The diversity in early Mandarin, from the basic division into northern and southern types to the diversity of sub-varieties within these two types, is one of the primary factors behind the development of the kinds of “quirky” Mandarin phenomenon described by Wiedenhof in this volume. 17 See Simmons (2019: 29–31) for details on the phonology of the Guānhuà described in Gāo’s textbook. 18 A different, and surviving, version of a dictionary by Matteo Ricci, the Dicionário Português-Chinês, compiled in the 1580s with Michele Ruggieri (1543– 1607), did not mark tones. A full color photographic facsimile of Ricci and Ruggieri’s dictionary is found in Witek (2001).

34  Richard VanNess Simmons















References Andō Hikotarō 安藤彥太郎. 1991. Zhōngguóyǔ yǔ jìndài Rìběn 中国语与近代日本 [The Chinese language in recent Japanese history]. Trans. Biàn Lìqiáng. Beijing: Běijīng Dàxué chūbǎnshè. Bridgeman, Coleman, and Samuel Wells Williams. 1848. Review of Meadows’ Desultory Notes on China. Chinese Repository 17, no. 2: 90–96. Chángbái Lǎomín 長白老民. 1958. “Tuīguǎng Jīnghuà zhìwéi gōngyì lùn” 推广 京話至为公义論 [The argument for the popularization of the dialect of Beijing as a highly important matter of public justice]. Originally published in 1903. In Qīngmò wénzì gǎigé wénjí, edited by Wénzì gǎigé chūbǎnshè, 34–35. Beijing: Wénzì gǎigé chūbǎnshè. Coblin, W. South. 1997. “Notes on the Sound System of Late Ming Guanhua.” Monumenta Serica 45: 261–307. doi:10.1080/02549948.1997.11731303

Standard Chinese in the Nineteenth Century  35 ———. 2000. “A Brief History of Mandarin.” Journal of the American Oriental Society 120, no. 4: 537–52. doi:10.2307/606615 ———. 2003. “A Sample of Eighteenth Century Spoken Mandarin from North China.” Cahiers de linguistique - Asie orientale 32, no. 2: 195–244. doi:10.3406/ clao.2003.1632 ———. 2006. Francisco Varo’s Glossary of the Mandarin Language, Volume 1: An English and Chinese Annotation of the “Vocabulario de la Lengua Mandarina.” Sankt Augustin: Monumenta Serica Institute. ———. 2007. Modern Chinese Phonology: From Guānhuà to Mandarin. Paris: École des Hautes Études en Sciences Sociales, Centre de Recherches Linguistiques sur l’Asie Orientale. DeFrancis, John. 1948. “A Missionary Contribution to Chinese Nationalism.” Journal of the North China Branch of the Royal Asiatic Society 73: 1–34. Du Halde, Jean-Baptiste. 1741. The General History of China. Trans. Richard Brookes, vol. 2. London: J. Watts. Edkins, Joseph. 1864. A Grammar of the Chinese Colloquial Language Commonly Called the Mandarin Dialect. 2nd ed. Shanghai: Presbyterian Mission Press. Elman, Benjamin A. 1984. From Philosophy to Philology: Intellectual and Social Aspects of Change in Late Imperial China. Cambridge, MA: Council on East Asian Studies, Harvard University. doi:10.2307/j.ctt1tg5kdt ———. 2000. A Cultural History of Civil Examinations in Late Imperial China. Berkeley: University of California Press. Féng Shòuzhōng 馮壽忠. 2010. “Wú Zhìhuī yǔ Qīng mò qièyīn yùndòng” 吳稚暉與 清末切音運動 [Wú Zhìhuī and the phoneticization movement in the late Qing]. Zhōngwén jìkān 28, no. 8.1: 13–30. Giles, Herbert A. 1892. A Chinese-English Dictionary. London: Bernard Quaritch. Hé Liángjùn 何良俊. 1995–2002. Sìyǒuzhāi cóngshuō 四友齋叢說 [Collected discussions from the four friends studio]. Preface dated 1569. In Xùxiū Sìkù quánshū, vol. 1125. Shanghai: Shànghǎi gǔjí chūbǎnshè. Hemeling, Karl Ernst Georg. 1907. “Die Nanking Kuanhua.” PhD diss., Universität Leipzig. Hirata Shōji 平田昌司. 2000. “Qīng dài Hónglúsì zhèngyīn kǎo” 清代鸿胪寺正音考 [Examination of the Qing dynasty pronunciation standard of the Court of State Ceremonial]. Zhōngguó yǔwén 6: 537–44. ———. 2002. “Review of A Cultural History of Civil Examinations in Late Imperial China by Benjamin A. Elman.” Tōyōshi kenkyū 61, no. 2: 329–37. Huáng Xiǎolěi黄晓蕾. 2012. “Lùn Qīng mò qièyīnzì yùndòng jíqí xiàndài yìyì” 论清末切音字运动及其现代意义 [The phonetic orthography movement of the late Qing and its contemporary significance]. Lánzhōu Jiāotōng Dàxué xuébào 31, no. 5: 97–100. Jiàoyùbù Guóyǔ Tǒngyī Chóubèi Wěiyuánhuì 教育部國語統一籌備委員會, ed. 1932. Guóyīn chángyòng zìhuì 國音常用字彙 [Lexicon of common usage in national pronunciation]. Romanized on the cover as Gwoin charngyonq tzyhhuey. Shanghai: Shāngwù yìnshūguǎn. Kaske, Elisabeth. 2008. The Politics of Language in Chinese Education, 1895–1919. Leiden: Brill. Klöter, Henning. 2011. The Language of the Sangleys: A Chinese Vernacular in Missionary Sources of the Seventeenth Century. Leiden: Brill.

36  Richard VanNess Simmons Lǐ Xīnkuí 李新魁. 1980a. “Lùn jìndài Hànyǔ gòngtóngyǔ de biāozhǔnyīn” 论近 代汉语共同语的标准音 [On the standard pronunciation of early Mandarin]. Yǔwén yánjiū 1: 44–52. ———. 1980b. “Kāngxī zìdiǎn de liǎngzhǒng yùntú” 《康熙字典》的两种韵图 [The two rime tables in Kāngxī zìdiǎn]. Císhū yánjiū 1: 174–82. ———. 1997. “Jìndài Hànyǔ nán-běi yīn zhī dà jiè” 近代汉语南北音之大界 [The major boundary between northern and southern pronunciations in early Mandarin]. In Lǐ Xīnkuí yīnyùnxué lùnjí, 228–66. Shantou: Shàntóu Dàxué chūbǎnshè. Lǐ Yǔmíng 李宇明. 2002. “Qīngmò wénzì gǎigéjiā de fāngyánguān” 清末文字改 革家的方言观 [Late Qing attitudes about dialect among language reformers]. Fāngyán 3: 193–200. Lín Tāo 林焘. 2010. Zhōngguó yǔyīnxué shǐ 中国语音学史 [History of Chinese phonology]. Beijing: Yǔwén chūbǎnshè. Liú Mèngyáng 劉孟揚. 1957. Zhōngguó yīnbiāo zìshū 中國音標字書 [Handbook of phonetic writing for China]. Originally published in 1906. In Pīnyīn wénzì shǐliào cóngshū. Beijing: Wénzì gǎigé chūbǎnshè. Lǔ Guóyáo 鲁国尧. 1988. “Tàizhōu fāngyīn shǐ yǔ Tōng-Tài fāngyán shǐ yánjiù” 泰 州方音史與通泰方言史研究 [Investigation of the history of the Tōng-Tài dialects and the history of Tàizhōu pronunciation]. Ajia-Afurikago no keisū henkyū 30: 149–219. ———. 1994. “Míng dài Guānhuà jí qí jīchǔ fāngyán wèntí—dú Lì Mǎdòu Zhōngguó zhájì” 明代官话及其基础方言问题——读《利玛窦中国札记》 [Ming period Guānhuà and the question of its underlying dialect: Reading Matteo Ricci’s De Christiana expeditione apud Sinas]. Originally published in 1985. In Lǔ Guóyáo zìxuǎnjí, 292–394. Changsha: Húnán jiàoyù chūbǎnshè. Lú Zhuàngzhāng 盧戆章. 2015. “Zhōngguó zìmǔ Běijīng qièyīn hédìng” 中國字母 北京切音合訂 [Combined edition of Chinese phonetic symbols with spellers for Beijing]. Originally published in 1906. In Pīnyīn wénzì shǐliào cóngshū, vol. 5, edited by Wénzì gǎigé chūbǎnshè, 149–86. Beijing: Guójiā túshūguǎn chūbǎnshè. Meadows, Thomas Taylor. 1847. Desultory Notes on the Government and People of China. London: Wm. H. Allen and Co. Ministry of Education. 1958. “Xuébù zī Wàiwùbù wén” 學部咨外務部文 [Memo of advice from the Ministry of Education to the Ministry of Foreign Affairs]. Originally published in 1906. In Qīngmò wénzì gǎigé wénjí, edited by Wénzì gǎigé chūbǎnshè, 68–71. Beijing: Wénzì gǎigé chūbǎnshè. Morrison, Robert. 1815. A Grammar of the Chinese Language. Serampore: Mission Press. ———. 1815–23. Dictionary of the Chinese Language, 6 vols. Macao: East India Company’s Press. Mote, Frederick W. 1977. “The Transformation of Nanking, 1350–1400.” In The City in Late Imperial China, edited by G. William Skinner, 101–54. Stanford, CA: Stanford University Press. Ní Hǎishǔ 倪海曙. 1959. Qīng mò Hànyǔ Pīnyīn yùndòng biānniánshǐ 清末汉语拼音 运动编年史 [Chronological history of the Chinese phoneticization movement in the late Qing]. Shanghai: Shànghǎi rénmín chūbǎnshè. Norman, Jerry (Luó Jiéruì 罗杰瑞). 2004. “Guānyú Guānhuà fāngyán zǎoqī fāzhǎn de yìxiē xiǎngfǎ” 关于官话方言早期发展的一些想法 [Some thoughts on the early development of Mandarin]. Trans. Tsu-lin Mei. Fāngyán 4: 295–300.

Standard Chinese in the Nineteenth Century  37 Platt, Stephen R. 2012. Autumn in the Heavenly Kingdom: China, the West, and the Epic Story of the Taiping Civil War. New York: Alfred E. Knopf. Rhodes, Edward J. M. 2000. Manchus and Han: Ethnic Relations and Political Power in Late Qing and Early Republican China, 1861–1928. Seattle: University of Washington Press. Rokkaku Tsunehiro 六角恒広. 1961. Kindai Nihon no Chūgokugo kyōiku 近代日本 の中国語教育 [Chinese language education in recent Japanese history]. Tōkyō: Awaji shobō. ———. 1992. Rìběn Zhōngguóyǔ jiàoyù shǐ yánjiū 日本中国教育史研究 [Study of the history of Chinese language education in Japan]. Trans. Wáng Shùnhóng. Japanese original published in 1988. Beijing: Běijīng Yǔyán Xuéyuàn chūbǎnshè. Schlegel, Gustave. 1895. “On the Extended Use of ‘The Peking System of Orthography’ for the Chinese Language.” T’oung Pao (first series) 6, no. 5: 499–508. Simmons, Richard VanNess. 2014. “A Comparative Look at Common Southern Jiāng-Huái and the Southern Mandarin Influences in Hé Xuān’s Yùnshǐ.” In Studies in Chinese and Sino-Tibetan Linguistics: Dialect, Phonology, Transcription and Text / Hànyǔ yǔ Hàn-Zàngyǔ yánjiū: Fāngyán, yīnyùn yǔ wénxiàn, edited by Richard VanNess Simmons and Newell Ann Van Auken, 133–53. Language and Linguistics Monograph Series 53. Taipei: Institute of Linguistics, Academia Sinica. ———. 2017. “Varieties of Mandarin.” In The Encyclopedia of Chinese Language and Linguistics, vol. 2, edited by Rint Sybesma et al., 672–82. Leiden: Brill. ———. 2019. “Northern and Southern Variations on a Theme: Notes on the Mandarin koiné of Qīng China.” Buckeye East Asian Linguistics 4, 23–34. Accessed February 5, 2020 at https://kb.osu.edu/handle/1811/87678. Skinner, G. William. 1977. “Regional Urbanization in Nineteenth Century China.” In The City in Late Imperial China, edited by G. William Skinner, 211–49. Stanford, CA: Stanford University Press. Stent, George С. 1871. A Chinese and English Vocabulary of the Pekinese Dialect. Shanghai: Customs Press. Sydenstricker, Absalom and A. Purist. 1888. “Romanizing the Official Dialect.” Chinese Recorder and Missionary Journal 19, 36–38, 133–35, 300–1. Thom, Robert. 1846. Extracts from Works Written in the Mandarin Language, as Spoken at Peking. Part 1. Ningbo: Presbyterian Mission Press. Wade, Thomas F. 1859. The Hsin Ching Lu 尋津錄, or, Book of Experiments; Being the First of a Series of Contributions to the Study of Chinese. Hong Kong: Printed at the Office of the China Mail. ———. 1867. Yü-yen tzŭ-erh chi 語言自邇集, A Progressive Course Designed to Assist the Student of Colloquial Chinese as Spoken in the Capital and the Metropolitan Department. London: Trübner & Co. Wáng Dōngjié 王东杰. 2010. “‘Shēng rù xīn tōng’: Qīng mò qièyīnzì yùndòng hé ‘Guóyǔ tǒngyī’ sīcháo de jiūjié” ‘声入心通’: 清末切音字运动和 ‘国语统一’ 思潮 的纠结 [‘The sound enters and the mind comprehends’: The intertwined ideologies of ‘National Language Unification’ and the phonetic spelling movement of the late Qing]. Jìndàishǐ yánjiū 5: 82–106. Wáng Qúnshēng 王群生. 1992. “Jīngzhōuchéng de ‘dōngbiānqiāng’” 荆州城的‘东边 腔’ [The ‘eastside accent’ of Jīngzhōu City]. Yǔwén yánjiū 2, no. 23: 78–82. Wáng Wèimín王为民. 2006. “Zhèngyīn cuōyào zuòzhě lǐjí yǔ bǎnběn kǎolùn” 正音撮要作者里籍与版本考论 [A study of the author’s origin and editions of

38  Richard VanNess Simmons Collected essentials of correct pronunciation]. Gǔjí zhěnglǐ yánjiū xuékān / Journal of Ancient Books Collation and Studies 11, no. 6: 53–56. Wang, Yeh-chien. 1964. “The Impact of the Taiping Rebellion on Population in Southern Kiangsu.” Papers on China 19: 120–58. Wells Williams, Samuel. 1844. English and Chinese Vocabulary in the Court Dialect. Macao: Office of the Chinese Repository. Witek, John W., S.J., ed. 2001. Dicionário Português-Chinês (Pú-Hàn cídiǎn 葡漢辭 典 [Portuguese-Chinese dictionary]). Lisbon: Biblioteca Nacional Portugal. Yamazaki Masato 山崎雅人. 1990. “‘Manbun’ Taishin Taiso Bukoutei jitsuroku no shakuyōgo hyōgi kara mita Kango no gaon, kōon no zetsumenonka ni tsuite” 『[ 満文] 大清太祖武皇帝実録』の借用語表記から見た漢語の牙音・喉音の舌面 音化について [On the palatalization of velars in modern Chinese: The Chinese words in the Manchu version of the Veritable records of the Grand Founder, Emperor Wu of the Great Qing]. Gengo kenkyū 98: 66–85. Yang, Paul Fu-mien, S.J. 1989. “The Portuguese-Chinese Dictionary of Matteo Ricci: A Historical and Linguistic Introduction.” In Proceedings of the Second International Conference on Sinology, Section on Linguistics and Paleography, 191–242. Taipei: Academia Sinica. Yè Bǎokuí叶宝奎. 2001. Míng Qīng Guānhuà yīnxì 明清官话音系 [The phonology of Mandarin in the Ming and Qing]. Xiamen: Xiàmén Dàxué chūbǎnshè. Yū Shǔluán 於曙巒. 1926. “Shā shì” 沙市 [Shā city]. Dōngfāng zázhì 23, no. 7: 45–61. Yuè Sháofèng 樂韶鳳 and Sòng Lián 宋濂. 1983. Hóngwǔ zhèngyùn 洪武正韻 [Rectified rimes of the Hóngwǔ reign]. Originally published in 1375. Reprint of a manuscript edition from 1782. In Yǐngyìn “Wényuān Gé Sìkù quánshū,” vol. 239. Taipei: Táiwān shāngwù yìnshūguǎn. Zhāng Wèi 張位. 1995–2002. Wèn qí jí 問奇集 [Collected notes on the unusual]. Preface dated 1590. In Xùxiū Sìkù quánshū, vol. 238. Shanghai: Shànghǎi gǔjí chūbǎnshè. Zhāng Yùshū 張玉書 et al., eds. 1985. Kāngxī zìdiǎn 康熙字典 [Kāngxī imperial dictionary]. Originally published in 1716. Shanghai: Shànghǎi shūdiàn. Zhōu Déqīng 周德清. 1983. Zhōngyuán yīnyùn 中原音韻 [Rimes of the central plains]. Originally published in 1341. Reprint of a manuscript edition from 1782. In Yǐngyìn “Wényuān Gé Sìkù quánshū,” vol. 1496. Taipei: Táiwān shāngwù yìnshūguǎn.

2

Manchu, Mandarin, and the Politicization of Spoken Language in Qing China Mårten Söderblom Saarela

Abstract This chapter discusses the politics of plurilingualism during the last imperial dynasty, notably the coexistence of Literary Chinese, written vernacular Chinese, Mandarin, and Manchu, and the way in which the Qing administration responded to plurilingualism within the government administration. Certain members of the ruling elite spoke Manchu, a language that is genetically unrelated to and typologically distant from Sinitic. Since the Qing rulers tried to integrate Manchu language maintenance with an appropriation of Chinese political institutions and cultural traditions, the court had to negotiate the position of its own language in an environment that used and identified with Literary Chinese, written vernacular Chinese, and spoken Mandarin. It is shown that this policy led to a high awareness of societal plurilingualism and the importance of language planning. Commercially published books of Manchu-Chinese language pedagogy evidence a growing concern for vernacular languages beyond the government in the period.

Introduction In the early to mid-Qing period, roughly corresponding to the rule of the four emperors reigning under the titles Shùnzhì (r. 1644–61), Kāngxī (r. 1662–1722), Yōngzhèng (r. 1723–35), and Qiánlóng (r. 1736–95), spoken language received great attention from the imperial court and emerged as a matter of elite concern. The consolidation of a Manchu polity, the invasion of China and the subsequent coexistence of Manchu and Chinese in the administration, and finally an activist imperial policy at the local level all contributed to the politicization of the speech of the hereditary Banner elite and Chinese civil officials. The publication and circulation of books for learning both Manchu and the prestigious form of spoken Chinese, Mandarin, show that Bannermen and civil examination candidates made efforts to maintain or gain competency in certain spoken languages. This chapter will discuss sources from the period in which spoken language emerges as a problem or a matter of debate. I refer to such attention to language by government representatives or elite individuals as the politicization of spoken language.

40  Mårten Söderblom Saarela This chapter seeks to demonstrate, by example, what the study of language in Qing history can bring to sinophone studies and Chinese historical sociolinguistics. First, by focusing on the relationship between Manchu and Chinese, this chapter contributes to a relatively uncultivated area of sinophone studies: linguistic diversity in the Chinese heartland and in the capital under the rule of an empire quite different from those later ascendant on the southeastern Chinese coast and in Southeast Asia. Second, through using rare books, archival documents, and recently published primary sources, the chapter demonstrates the value of philological research for historical sociolinguistics. An almost unfathomable amount of written sources remains from the Qing period. Who knows just how much they can tell us about the period’s languages and their role in society? This chapter will begin with an overview of the languages relevant for the Manchu Qing state when it invaded China in 1644. In the first section, I will describe the coexistence of Literary Chinese, written vernacular Chinese, Mandarin, and Manchu, and the way in which the Qing administration handled plurilingualism in writing. Second, I will discuss the politicization of spoken languages under Shùnzhì, Kāngxī, Yōngzhèng, and Qiánlóng. Finally, I will look at how spoken language received increased attention also outside the bureaucracy, through translations and language pedagogy.

The Languages of Late Imperial China and the Qing Administration Language was a matter of concern to the Manchus as they built their plurilingual state and fought the Chinese Ming empire in the seventeenth century. At this time, languages of political importance in northern China and southern Manchuria included Classical or Literary Chinese, written vernacular Chinese, and spoken Mandarin. Literary Chinese was the dominant written language of late imperial China. It was based in terms of syntax and core vocabulary on the largely Confucian literature of the late first millennium BCE. The Confucian classics were widely studied, but their language was very far removed from that of their readers’ daily existence. It was read aloud and, to some extent, its vocabulary and syntax were used in formal or affected speech. Yet its pronunciation was underdetermined. Some genres required that certain phonological distinctions were represented, somehow. Yet many texts in Literary Chinese could be read aloud in a variety of ways, depending on the preferences of the reader or listener. The system of literary examinations leading to official appointments, one of late imperial China’s most important institutions, sanctioned the underdetermination of pronunciation through its focus on writing (Elman 2000: appendix 4). By the eighteenth century, an alternative written language with roots in the medieval period was firmly established alongside Literary Chinese. This other language has since the early twentieth century been called “plain speech writing” (白話文). In the eyes of late imperial literati, it

Politicization of Language in Qing China  41 represented the “vernacular” (俗) pole of written expression. That terminology notwithstanding, it should not be equated with the everyday language of the common people. The written vernacular could serve as another cosmopolitan literary language appropriate for certain genres, role types, social contexts or statuses in both fictional and other narrative texts (Wei 2014: 258 and 276) or didactic pieces. The spoken varieties of Chinese did not have equal currency. The cosmopolitan spoken language of late imperial China was Mandarin, as sixteenth-century European visitors called it (Coblin 2000: 539). The equivalent Chinese term was “official speech” or Guānhuà (官話), which in at least one government document was translated into Manchu as “correct pronunciation” (tob mudan) (Hǎo 1735). This term is a direct parallel to Chinese zhèngyīn 正音, which was also used in Qing-period sources to refer to the pronunciation of Mandarin (Ishizaki 2014: 2–5). The exact dialect basis of Mandarin changed over time, from south to north, but it was always based on the demographically and geographically dominant northern dialect group, spoken in an area where both the Ming and Qing capitals (Nanjing and Beijing) were located (Coblin 2000). Qing-period normative Mandarin, and the changes it underwent, was not simply the result of imperial policy, but to a great extent reflected social realities beyond government control. Simmons’s chapter in this volume discusses how the Mandarin norm changed during the course of the Qing period. The civil service examinations and the system of official appointments had the “unintended consequence” (Elman 2014) of promulgating Mandarin through increasing mobility. Travel outside of candidates’ native localities was a necessity for most of those who proceeded to the provincial and national exams. Moreover, features of the system meant that for some Southerners, registering for the lower-level exams in the capital region improved their chances of success. For those who passed, appointments could be anywhere but the home region, and they rotated. The geographic mobility of the examination elite encouraged the use of Mandarin. Still, familiarity with the language was rare in many areas and sub-elite social groups. Manchu, which is Tungusic, was the ancestral language of the Qing imperial clan and other aristocratic lineages. Today, Manchu is rarely spoken as an everyday language, but efforts to revive its use have been undertaken in communities in the Chinese Northeast. However, the Sibe, an ethnic group in Xinjiang, write and speak a language that is close to Manchu. The situation was very different in the seventeenth and eighteenth centuries, when Manchu was used in the Qing capital of Beijing. Written Manchu had a short history and was not as varied as written Chinese, nor was it as far removed from the spoken language. This characteristic of the Manchu language was very consequential for its relationship to Chinese. The Manchu language was influenced by Mongolian, from which the Manchus borrowed their script in the early seventeenth century. Furthermore, from the early days of its existence as a written language, Manchu

42  Mårten Söderblom Saarela coexisted with northern vernacular Chinese. Nurhaci (1559–1626), founder of the Manchu state, almost certainly knew how to both speak and read Chinese (Stary 2013: 182–83). Before the Manchus invaded China proper, they incorporated the Liáodōng 遼東 area and its ethnically diverse population in their polity. Individuals from Liáodōng moved with the Manchus to Beijing and on to the provincial garrisons after 1644. Already by the mid-seventeenth century, many speakers of Manchu were bilingual. The pre-conquest Manchus instituted the so-called Eight Banners (Ma. jakûn gûsa; Ch. 八旗) to group their closest subjects, which were from the very beginning ethnically and linguistically diverse. The Banners were military units on the battlefield but also key institutions in peacetime, allocating stipends and housing to soldiers and their dependents. There were Manchu Banners, Mongol Banners, and the “heavy troops” (Ma. ujen cooha) known as the Chinese army (漢軍), or the Chinese Banners. Differences in status, professional duties, and language existed in the Banners throughout the Qing period, and they changed over time. Yet there was a strong association between the Manchu language and the Banners, and not just the Manchu Banners (Porter 2019). The plurilingual Qing state was thoroughly bureaucratized and language appeared important for statecraft primarily in the context of writing. Notably, Nurhaci intervened, or so the received account tells us, in linguistic usage at his court by instituting a written Manchu language. After the conquest of China, the sudden coexistence of two languages in the bureaucracy caused difficulties. The government’s evolving record-keeping practices appeared open to debate. Should “Manchu writing [be] promulgate[d] throughout the empire, so that it can be practiced and translated by all” (Liú Yánchén 2010: 41; Shìzǔ Zhāng huángdì shílù 1986: 196)? Yáng Sìzhòng 楊四重 (n.d.) voiced this opinion to the court in early 1646. The problem was not solved by 1703, when Wáng Yán 王掞 (presented scholar in 1670), then President of the Ministry of Punishments, complained that prisoners’ oral testimony was only recorded in Manchu (清書), which the Ministry’s Chinese officials could not read (Lǐ Huán 2009: vol. 4, 189–92). Conversely, there were reports of Manchus who were illiterate in Chinese being duped by their Chinese subordinates when they went to serve as local officials (Hirata 2001: 33). New developments in the system of official communications led to adjustments regarding which language should be used when writing. The coexistence of written Manchu and written Chinese within the bureaucracy was regulated, but the contingencies of day-to-day administration led to evolving practices and the reiteration of norms.

The Politicization of Spoken Language The spoken language had already been politicized before the conquest. Similarity in spoken language among the Jurchen tribes justified political unity

Politicization of Language in Qing China  43 in Nurhaci’s eyes (Yeh 2002: 42), but the difficulties with spoken Manchu among Chinese officials posed a problem to his successor (Tàizōng Wén huángdì shílù 1986: 148–49; Yáng Yàgēng 2005: 17). The record of Nurhaci’s alleged invention of the Manchu written language also took spoken language into account. The record described the creation of a Manchu written language as an attempt to replicate the situation in the Ming empire, where “when [one] reads the writing of the Chinese state aloud, people knowing Chinese writing and people not knowing it all understand” (Dà Qīng Mǎnzhōu shílù 1969: 108). This statement, even in its underestimation of the degree of Chinese diglossia,1 shows clearly the early Manchu leadership’s ideal of a language shared by all, even those who could not read or write and would only ever communicate orally with government representatives. The Shùnzhì and Kāngxī Periods The Shùnzhì and Kāngxī courts intervened in the spoken language of their subjects, albeit on a smaller scale than their immediate successors. The focus was entirely on civil officialdom and the Banner elite. However, at least in the case of compiling and publishing normative lexicographical works, these efforts should be treated as instances of language planning. The spoken language remained an issue. Yáng Sìzhòng suggested that the young Shùnzhì emperor learn “Chinese pronunciation” (漢音) (Liú Yánchén 2010: 41; Shìzǔ Zhāng huángdì shílù 1986: 196). A special group of officials for a time served as interpreters between the Manchu and Chinese elements of the bureaucracy (Miyazaki 1991–94: 332–33). Manchu’s position was strongest in the agencies staffed exclusively by Bannermen. In 1690, the Kāngxī emperor suggested that Jesuit missionaries who were studying Manchu frequent these agencies to practice Manchu conversation (Du Halde 1736: vol. 4, 266–67). The emperors’ interaction with officials during court audiences contributed to the politicization of the spoken language. An early post-conquest example is found in the institution of language training in the Hanlin Academy (翰林院). Highly ranked recent examination graduates of both civilian and Bannerman backgrounds were assigned to study Manchu and Chinese. The Shùnzhì emperor personally interviewed them to ascertain their Manchu skills by oral examination. It was probably the expectation that this group of officials would talk directly to the emperor that motivated the stipulation that those selected to study Manchu should have “clear and distinct” (明爽) or “suitable” (合式) pronunciation (聲音) (Liú Yánchén 2010: 42; Shìzǔ Zhāng huángdì shílù 1986: 518–19). I think we should understand this instruction to mean that only those with passable Mandarin should be chosen to interact orally with the emperor. The Kāngxī emperor witnessed language attrition during audiences with younger Manchu officials, and he aspired to reverse it by means of a linguistic reference work (Han-i araha manju gisun-i buleku bithe 1708: preface, 3a).

44  Mårten Söderblom Saarela His court also intervened in the realm of Chinese speech, if only its pronunciation. Yùzhì Kāngxī zìdiǎn (御製康熙字典, lit. ‘Imperially commissioned character standard of the [reign of] Secure Peace’, or the ‘Imperial dictionary’) from 1716 arranged characters based on their graphic form, but it was greatly concerned with pronunciation. The book was presented as a comprehensive guide to character readings. It was, indeed, treated as a guide to proper pronunciation shortly after its publication (Yuán 1729: Lí Fènghuì’s 黎鳳翽 [?] preface). Yet what the Kāngxī imperial dictionary promoted was a learned reading pronunciation, not Mandarin as a complete language.2 Kāngxī’s rime dictionary, published after his death, served the same purpose (Lǐ and Wáng 1881). This book sought to make sound glosses intuitive and pronounceable. In most cases, the characters used in the glosses were thus meant to be read in contemporary Mandarin pronunciation (Luó 2008: 463). Other measures introduced by the court provided an incentive for Southerners to learn Mandarin, though the measures themselves did not have that intention. Rather than representing instances of language planning, these measures constituted the use of an individual’s language as a marker of their ancestry and of their belonging in a geographical sense. In certain areas of the south, it had become common to register for the low-level civil examinations in the capital area in order to boost one’s chance of success. In an effort to counter this trend, the pronunciation of examination candidates in the capital was checked for southern accents (Hirata 2001: 50–53). If fraudulently registered candidates advanced as far as the final palace examination, the Kāngxī emperor was confident that he could personally weed them out. In 1712, Kāngxī claimed to “thoroughly know the accents (語音) of the thirteen provinces.” On the day of the palace examination, he warned, the candidates “might have to report to me in person” (Shèngzǔ Rén huángdì shílù 1986: vol. 6, 470). If the accent did not fit that of their registered home region, Kāngxī would know. The problem, for Kāngxī, was not that the candidates spoke with an accent; what was unacceptable was to dissimulate one’s origin. The Yōngzhèng Period Under Kāngxī’s successor, Yōngzhèng, Mandarin Chinese was periodically promoted among wider layers of the population in certain areas. In the largely non-Chinese Southwest, provincial officials opened schools for teaching the language, which contributed to an increase in L1/Mandarin bilingualism in the region (Rowe 1994: 439). Mandarin was also promoted among the local Chinese elite in the Southeast. In 1728, Yōngzhèng issued an edict, widely studied since, that ordered officials in the provinces of Guangdong and Fujian to “ensure that the language [of prospective officials of these provinces] is clear” and not in “local pronunciation” (鄉音) (Shìzōng Xiàn huángdì shílù 1986: vol. 7, 1074). Yōngzhèng’s edict was, in part, another

Politicization of Language in Qing China  45 example of the court audiences—as the emperors’ chance to speak directly with outside officials—contributing to a politicization of the spoken language. In this case, Yōngzhèng was unable to understand Cantonese and Fujianese officials when they appeared before him to present their credentials for appointment. Yōngzhèng was a very hands-on emperor down to the local level, and he sensed danger. The realization that officials from Fujian and Guangdong, who because of the law of avoidance would be serving outside their home area, would not be able to communicate the emperor’s edifying instructions to the bas peuple contributed to him urging them to learn Mandarin (Yeung 2000: 120). More generally, Yōngzhèng shared the widespread belief that local officials should not rely on mistrusted clerks and runners to mediate their communication with the people. In response to the edict, the authorities said that they would verify the Mandarin ability of lower-level exam candidates in the provinces concerned and founded specialized schools there. Initially, however, results were unsatisfactory, and in 1734, an official on the scene memorialized with a request to strengthen controls both in the schools and on the examination grounds (Yáng Bǐng 1734). The campaign to promote Mandarin extended to the Green Standard troops stationed in Fujian. In 1735, governor-general Hǎo Yùlín 郝玉麟 (d. 1745) requested that knowledge of Mandarin be a criterion for promotion in the military. If the officers knew Mandarin, and use of the local language (Ch. 鄉談; Ma. tesu ba-i mudan) was prohibited in interaction with the soldiers, the use of Mandarin would spread among the troops (Hǎo 1735). The campaign to promote Mandarin among examination candidates in the southeastern provinces is considered to have failed, as governmentsponsored Mandarin teaching had by and large ceased in these areas by the second half of the eighteenth century. Students were soon allowed to sit for the examinations regardless of their Mandarin ability. Several reasons have been inferred for this alleged failure. We should not, however, take the reversal of Yōngzhèng’s policy to mean that southeastern literati who aspired to government office no longer learned Mandarin; they just did not go to a specialized, official institution to do so (Hirata 2001: 46; Murata 2003: 65–66). Mandarin’s Manchu Connection The Mandarin campaign in the Southeast had a Manchu connection. First, it has been suggested that Shěn Qǐyuán 沈起元 (1685–1763), a Fujian county magistrate and a driving force behind the schools for “correct pronunciation” in the province, took spoken language seriously in part because he had been assigned to the Manchu class when he was in the Hanlin Academy (Katayama 1978: 55). The experience of learning a new language in adulthood had taught him that foreign languages were not to be taken lightly; he once complained that the rules governing Manchu grammatical particles were so intricate that “one would have to be a sage to compose” in the

46  Mårten Söderblom Saarela language (Shěn and Shěn 1999: 557). Through firsthand experience with the Manchu-Chinese linguistic diarchy, so the argument goes, Shěn would have become alert to language as a political problem. Second, the promotion of the foreign language of Mandarin on the far-flung southeastern coast came to involve Manchu in response to the campaign’s human resources problem. To endow a Mandarin school was easy enough if there was money to go around, but whence the Mandarin teachers? Individuals from Guilin in Guangxi, who were speakers of a dialect of the northern group, commonly tutored Cantonese students in at least some kind of “creole Mandarin” (a term I borrow from Murata 1995: 11), which was a solution of sorts (Yeung 2000: 134). Bringing in teachers from other provinces was not always a good option, however. The government-run schools in Fujian enlisted teachers from the neighboring provinces of Zhejiang and Jiangxi, but since they were unable to communicate with the monolingual, Mǐn-speaking students, they were sent back and replaced with local instructors already trained in Mandarin (Katayama 1978: 58). These local instructors, in turn, might have had some kind of personal or professional connection to Manchus or other Bannermen. Indeed, the Bannermen in the garrisons in southeastern and central China represented an important reserve of Mandarin talent. In the second half of the nineteenth century, Manchus could promote themselves as Mandarin teachers in print, but I have not seen any evidence of this in the eighteenth century. Southern civilians, however, could gain linguistic capital by associating with Bannermen. Such was the case of Yuán Yīzhōu 袁一州 (n.d.) of Guangdong, who published a Mandarin textbook in 1729 on the basis of the notes he had been using for more than 40 years of teaching (Takata 1997). Yuán had worked as a tutor in the home of a Banner official, during which period he had also “practiced Manchu writing” (習清書). Yuán was an ideal Mandarin teacher, as he was skilled both in the local vernacular and the spoken prestige language (cf. Cài Kǎi’s 蔡鎧 and Lí Fènghuì’s prefaces in Yuán 1729). A fascinating aspect of Yuán’s book is that it to a certain extent combined teaching in both of the spoken languages promoted by the Qing court. The express purpose was to teach Cantonese speakers Mandarin, but one page in the textbook was dedicated to Manchu terms, in Manchu script with Chinese transcription and translation, that a future official might find useful (ibid., 19b). Another circumstance similarly suggests that Manchu played a role in the teaching of Mandarin in the Southeast following the 1728 edict. The source is Higashikawa Tokuji 東川徳治 (1870–1938), who in a 1919 book based on fieldwork in Taiwan wrote that in northern Fujian, there had in 1729 been “two Academies for correct pronunciation, which concentrate on the teaching of Manchu (清文)” (Higashikawa 1919: 124). Did the schools actually teach Manchu, or might they have taught northern Chinese with a smattering of bureaucratic Manchu vocabulary, as in Yuán’s book? Unfortunately,

Politicization of Language in Qing China  47 in this instance Higashikawa did not cite his source nor did he elaborate on the role that Manchu would have played in the two schools. The Qiánlóng Period The Qiánlóng emperor paid a lot of attention both to written and spoken Manchu. He noted with displeasure that Bannermen on guard duty spoke Chinese among themselves (Yáng Yàgēng 2005: 25). Furthermore, following a similar line of reasoning to that of Hǎo Yùlín, Qiánlóng argued that “if military officers do not know Manchu themselves, how can they teach it to the soldiers?” (Zhōngguó Dìyī Lìshǐ Dǎng’àn Guǎn 2011: vol. 12, 443, item 2118). As for the regulation of Chinese, Qiánlóng did not pursue his father’s campaign to promote Mandarin in the Southeast through language tests in the civil examinations. At least one pronouncement which Qiánlóng made late in life (1795) suggests that he, in his imperial persona of Confucian primus inter pares, left literary reading pronunciation to the discretion of the literati (Yeung 2000: 126). A person’s speech, however, was an important part of his identity, and Qiánlóng paid it due attention. Qiánlóng checked Bannermen’s Manchu-speaking skills through court audiences. Manchus were expected to recite their CV in Manchu, and be prepared to answer questions relating to it in that language when they appeared before the throne. In an indication of the importance of the court audience for maintaining a degree of Manchu-speaking ability, it was reported in 1744 that reciting their personal history was the only thing that some Bannermen in the garrisons could still do in Manchu (Yáng Yàgēng 2005: 25–26). In 1775, Qiánlóng noted that Golmin, a translation exam graduate from Mukden in Manchuria, “when received in a court audience, can recite his CV but is unable to utter a single word in Manchu when questioned” (Zhōngguó Dìyī Lìshǐ Dàng’àn Guǎn 2011: vol. 11, 386, item 1877). Speaking Manchu was a marker of being Manchu and hence a good servant of the dynasty. Similarly, the way Chinese officials spoke said something about who they were. In 1777, Qiánlóng proclaimed: Whenever I, the emperor, audit an official, he recites his CV to me. Within a few moments, I am able to identify his accent (語音). When the various servants [of the throne] check the pronunciation and examine [the physical appearance of candidates], they will, if they apply themselves to listen for it, find that it is not difficult to distinguish northern and southern pronunciation. (Gāozōng Chún huángdì shílù 1986: vol. 21, 900) These statements were made in relation to Qiánlóng’s audit of the official Zhāng Tíngtài 張廷泰 (n.d.), then serving in Shàoxīng 紹興, Zhejiang. Listening to Zhāng reciting his CV, the emperor noticed that his “pronunciation approached that of Shàoxīng.” Upon further questioning, it turned

48  Mårten Söderblom Saarela out that Zhāng, registered as from the Beijing area, had lived in Shàoxīng for several years as a child because of his father’s sojourn in the city, long enough, apparently, for him to acquire the local accent. Clearly, Qiánlóng thought, Zhāng should not be serving in what was practically his hometown (Gāozōng Chún huángdì shílù 1986: vol. 21, 899–900). As with Kāngxī’s scrutiny of examination candidates during the palace examination, what mattered for Qiánlóng was not the accent as such, but rather the disconnect between accent and stated geographical origin. This attitude toward the speech of subjects contrasts markedly with that adopted by state-builders in early twentieth-century China, as discussed in Klöter’s chapter in this volume. Qiánlóng associated his scrutiny of Zhāng with measures he instituted as early as 1745. At that time, Qiánlóng expanded the existing system of checking the pronunciation of candidates sitting for the low-level civil examinations in the Beijing region. He introduced designated “pronunciationchecking censors” (審音御史), a post in which both Manchus and Chinese were stipulated to serve (Hirata 2001: 51). Manchus were tasked with evaluating the Mandarin abilities of Chinese civilians, who had to orally communicate their credentials (name, origin, and ancestry). Qiánlóng viewed these measures, at least retrospectively, as an extension of his own scrutiny of officials’ speech during court audiences. He did not intend that more Chinese officials learn Mandarin through these efforts, but the institution of the checks clearly gave officials the incentive to do precisely that. In other contexts, Qiánlóng explicitly affirmed the official character of Mandarin pronunciation. He instituted restrictions on the geographical origin of the officials working at the Court of State Ceremonial (鴻臚寺) to the metropolitan district surrounding Beijing (Hirata 2000: 539). This institution was a location visited by foreign delegations and thus important for the image that the Qing empire presented to the world; presumably, Korean and other visitors were to see that the Qing empire spoke in the two voices of Manchu and Mandarin (Wāng and Yáo 2014: 25–26). It should be noted, however, that the “institutionalization,” in Hirata Shōji’s words, of Mandarin in this context properly speaking refers to reducing the underdetermination of the pronunciation of Literary Chinese, rather than promoting Mandarin as a full-fledged language with its own syntax and lexicon. The official announcements would have been in Literary Chinese, the important thing being that they were pronounced in Mandarin.

Manchu Translations of Confucian Texts as Vernacularization The politicization of speech extended beyond pronunciation checks in court audiences, official examinations, and appointments and promotions. It also happened through books that stressed the spoken and vernacular character of language. In terms of official publications, translations of Chinese Confucian books into Manchu served to standardize Manchu vocabulary,

Politicization of Language in Qing China  49 spelling, and phraseology. These translations suggested to readers that Manchu was a vernacular—and hence speakable—language. Spoken language was also brought to attention through genres with which the government was only minimally involved such as Manchu textbooks. In Manchu textbooks with Chinese translations, both Manchu and Chinese were often treated as spoken—even colloquial—languages. This last section will discuss how Manchu translations and textbooks made readers aware of Manchu and Chinese as vernacular languages. The Qing court sponsored the translation of several Chinese Confucian books that were of interest for students hoping to pass the civil service or the translation examination.3 The nature of these translations helps explain why Manchu was often associated with vernacular or spoken Chinese. In reference to the Confucian classics, Qiánlóng’s court scholars asserted the superiority of Manchu translations over Chinese annotations, which “stick too closely [to the original] and rarely penetrate it.” Without observing the generic conventions of commentaries in the Chinese tradition, the Manchu translation “naturally cuts away the expository subtleties to create something to rely on” (Qīndìng Bāqí tōngzhì 1987: vol. 665, 1026–27; see further Durrant 1979: 657; Yeh 2002: 87). Stressing the self-sufficiency of the Manchu translations paralleled an increased focus on the basic text of the classics also in the Chinese civil service examinations of the period (Yeh 2006: 25–36). According to the court scholars, the Chinese classics had been explained for centuries in Chinese, but remaining within the realm of the literary language, earlier explanations could not compete in clarity with the complete recasting of the text into a different language through translation. Unlike the Chinese commentators, the translators of the text into Manchu could easily do what Martin Luther claimed to have done when translating the Bible into German. Luther “did not have to ask the letters in the Latin language how one should speak German”; he arrived at the appropriate expressions from speaking to people around town (Luther 1938: 15). Similarly, not bound to the form of the Chinese original, the Manchu translators could forge an entirely new text. That is not to say that the Manchu translators were cavalier in their treatment of the Chinese original. On the contrary, from the mid-eighteenth century onward at least, they sought to the greatest extent possible to render one Chinese term by one unique Manchu word (Watanabe 2013: 235). Cognizant, as well, of the uncertain meaning of some passages, the translators often chose to reproduce obscure points rather than to resolve them in favor of a certain interpretation (Abel-Rémusat 1818: 279). When faced with Chinese words that had no ready Manchu equivalents—e.g., the flora and fauna in the Poetry classic (詩經)—translators in 1768 introduced neologisms (Yeh 1992: 222). Even so, it has been argued that their wordier and more explicative translation was of greater value to students than an earlier version (1655) that used more Chinese loans (Yamazaki 2000: 259). Similarly, in the case of

50  Mårten Söderblom Saarela Mencius, the Manchu version (probably dating from 1755 or 1756) at times helped clarify the meaning of the original (Siegmund 2015: 311n7 and 323). The idea that the Manchu translations were easy was persistent. The Bannerman scholar-official Sungyun (1752–1835) expressed the same opinion as Qiánlóng’s court scholars. In a Manchu collection of anecdotes, Sungyun related how Manchu students of Confucian literature—who, we must assume, were fluent in spoken Chinese and to some extent literate in the language— were able to make sense of the original Classical Chinese text only after they had seen the Manchu translation. For these students, the Manchu version took the place of a rendering into spoken Chinese by the teacher; with it, one “understands even without an [oral] explanation” (yala giyangnarakû [< Ch. jiǎng 講] de ulhimbi) (Stary 1983: 215, 437). Much earlier, Shěn Qǐliàng 沈啟 亮 (fl. 1645–93), a Chinese Southerner and Manchu pedagogue, expressed a similar opinion. He wrote that “once translated” into Manchu, “a character that was obscure in the Chinese version is immediately washed completely clear.” Manchu, he asserted, is “simple and gets the meaning across,” and it “allows people to see clearly at first glance” (Shěn 2008: 2). Translation theorizers remarked on the fact that written Manchu did not have the stylistic range of Chinese. Chū-qí-shǔ 初齊曙 (b. c. 1660s–70s) wrote in 1727 that “Chinese language (漢話) has always had many literary and vernacular words that, belonging to the same [semantic] category, can be interchangeably substituted.” Chinese could have eight words for one thing; Manchu might only have one (Chū-qí-shǔ and Chū-jǐng-fū 1727: under the heading “Biān yán” 辨言). The anonymous author of an undated bilingual manuscript for students of Manchu similarly asserted that Manchu texts, unlike Chinese, could “forgo complicated expressions that are extraneous to the explanation.” In Manchu, “the sound of an expression is a complete match,” and there was no need to explain things like the “six scripts”— identified in the structure of Chinese characters since antiquity—or the “four tones” of Middle Chinese that were still important for poetic composition and literary recitation in the Qing period. Without any need for such a hermeneutical apparatus, Manchu could place the reader “as if face to face with thousands of years in their full extent” (Qīngwén hòuxué jīnfá n.d.: vol. 1, preface). It would be a mistake to think of the post-conquest Manchu translations of the Confucian canon as the introduction of the Chinese literary heritage into a foreign culture. From the point of view of early and mid-Qing readers of the Manchu Confucian books, the translations represented, rather, the explanation of universal truths in a language that was closer to that of their everyday life than was the Classical Chinese original. It is true that the language of the Confucian translations was not kitchen Manchu; the translations differed from conversational Manchu in style, in the use of new vocabulary, and in the extension of the semantic range of existing vocabulary to match that of their Literary Chinese interpretanda (Abel-Rémusat 1818: 281–82). Yet a speaker might still employ such Manchu words in

Politicization of Language in Qing China  51 speech without seeming out of place. As Manchu proficiency was increasingly gained only through formal instruction (see, e.g., Crossley 1994; Liú Xiǎoméng 2012), the language was taught in order to be spoken as well as read. Once acquired, Manchu was quotidian, whereas the Chinese language of the classics was not. The Manchu translations of the Confucian classics were, perhaps, not what the farmer might sing at his plough or the weaver hum at his shuttle, as Erasmus hoped for the translated Gospels and Pauline Epistles (Erasmus 1965: 97)—if only because urban Bannermen did not farm and weaving was women’s work. Yet the creation of a Confucian corpus in Manchu has some similarities with the translation projects of the European Reformation. The relative proximity of written Manchu to the spoken language meant that a Manchu translation, to Manchu-Chinese bilinguals, was a form of vernacularization. The Qing experience of Manchu translations of the Confucian classics thus presents, I argue, a precedent for later Chinese vernacularization projects. The late nineteenth-century Protestant missionaries whom Snow studies in his chapter in this volume appear to have been either completely unaware of this Manchu precedent, or too convinced of their own epoch-making importance to pay it any attention. Speech in Manchu-Chinese Language Pedagogy Translations of the Confucian classics involved translating Literary Chinese—not everyday spoken Chinese—into Manchu. Still, as the preceding section suggested, the vernacular character of Manchu contributed to it being paired, in syllabaries and textbooks, with vernacular or at least sayable Chinese. Simple Manchu syllable lists were used in the government schools in the seventeenth century, but the imperial court never intervened in the genre of elementary language pedagogy. The palace print shop put out dictionaries and transcription manuals, but not syllabaries or textbooks. Bilingual Manchu-Chinese pedagogy developed instead through private and commercial initiative. Bilingual language textbooks tended to match Manchu dialogues, perceived as vernacular, with the written Chinese vernacular. At least one pedagogue, Wǔ-gé 舞格 (fl. 1730), initially felt a little awkward at the prospect of publishing a book with such a strong vernacular character. When his Chinese friend asked him to turn his private teaching notes into a textbook for publication, he resisted, saying that “the explanations (Ma. suhengge; Ch. 註) are entirely in trivial, untidy language and, being coarse and vulgar, unpresentable” (Wǔ-gé 1730a: preface, 2b). Yet presented and published they were, with great success. In a later edition, a note was added after the chapter that contained the vernacular dialogues. In Chinese only, the note targeted prospective readers—“highly discerning gentlemen”—who, picking up the book with the intention of studying Manchu, might be offended by the linguistic register. Since the students would be readers of Chinese, the

52  Mårten Söderblom Saarela “oral language” (口頭言語) that might take such prospective readers aback presumably included the Chinese text as well as the Manchu. The author of the note, however, argued that vernacular language had a clear pedagogical value (Wǔ-gé 1730b: 2:57a). Thus when Manchu texts, such as the dialogues of textbooks, were translated into Chinese, the vernacular character of the original was carried over into the Chinese version. Manchu was evidently treated as a vernacular language; relatedly, it was also treated as a spoken language. The Chinese Southerner Liào Lúnjī 廖綸璣 (fl. 1670–80s) was a pioneer in this regard. Liào’s bilingual Manchu syllabary was published in the South. Liào felt that people in his day—Chinese people, we must assume—tried to learn Manchu only as a written language, disregarding the spoken language. Liào disagreed; a person who wants to understand Manchu “must mimic the lips and cheeks”—that is, learn to pronounce the sounds—“before practicing the characters” (Liào 1996: 59). Surprisingly, Shěn Qǐliàng was of a different opinion. He stressed the need for learners to study Manchu primarily on the basis of the script rather than to “first study the tongue’s movements” (先學 滾舌) (Shěn 1686: preface, 2a), which is what would happen if one interacted with the Manchu speech community. Yet Shěn thereby confirmed what was probably the reality in Beijing at this time, where Chinese residents would have had reason to find ways to communicate orally with the city’s Bannermen. Shěn’s books were commercial publications; if he were to make money from them, the Chinese of Beijing would have to start learning Manchu in the study rather than on the street corner. When Chinese characters did not serve to translate the Manchu, but only to transcribe it, they had to be read in northern or Beijing Chinese. The underdetermination of the pronunciation of the Chinese script was thereby reduced. An early Manchu syllabary, for example, specified that the sound glosses should be read in the tradition of “northern rimes” (北韻; Líng and Chén 1699: vol. 1, statement of editorial principles, 1a). Manchu-Chinese dictionaries, including the one that Shěn Qǐliàng published in 1683, often translated Manchu vocabulary into vernacular Chinese (Shěn 2008). In 1701, in a primer ascribed to Shěn, the sound glosses were similarly explained to “follow the rimes of the capital” (從京韻; Jiānzhù shíèr zìtóu 1701: 1a). Another extended syllabary, dating from 1733, contained a Manchu-script transcription of the Mandarin pronunciation of the entire Chinese preface, which was written in simple literary language. A reader who was literate in Manchu and fluent in northern vernacular Chinese could thus read the preface without knowing Chinese characters (as long as he succeeded in imputing the correct tones, which were not marked; Mǎn-Hàn shíèr zìtóu 1861). Some early bilingual pedagogical works allowed the reader to learn either Manchu or Chinese. A collection from 1699, for example, contained a classified vocabulary in which the pronunciation of the Manchu words was transcribed using Chinese characters, as suitable for Chinese literates learning Manchu, and the pronunciation of the Chinese words was transcribed

Politicization of Language in Qing China  53 using the Manchu script, as suitable for Manchu literates learning Mandarin Chinese (for the pronunciation was that of Mandarin; Líng and Chén 1699: section titled “Mǎn-Hàn shìlèi jíyào” 滿漢事類集要). Furthermore, the author of a textbook from 1746, who rearranged material from an imperial Manchu dictionary and then translated it into simple Literary Chinese, claimed that by using his book the reader could learn both languages (Juntu 2001: 2–3). The author specified that the book “not only taught Chinese characters, but also Chinese composition,” which implies that written and not spoken Chinese was the object of study. Yet the structure of the book suggests a link to speech. Instead of the original’s list of lemmata with definitions, the new textbook presented the material in the spoken format of the dialogue. The Chinese language of this textbook was not the written vernacular, but it was intended to be sayable. Even after Manchu ceased to be the default language for everyday oral communication among Bannermen from the second quarter of the eighteenth century, they learned the language aiming for proficiency both in speech and writing. “If you do not practice the Manchu language as a child,” Sungyun wrote, “and wish to learn it in adulthood, you will be something like a man without a tongue; unable to get a single sound into the conversation, you will be left behind, nothing but the butt of their jokes” (Stary 1983: 267, 467). Remember that for Manchu males at the highest echelon, who aspired to high official positions, being summoned to an imperial audience was a real possibility. Such individuals had to worry about their ability in spoken Manchu.

Conclusion This chapter has shown that in the early to mid-Qing period, the government cared about the use of spoken language in a variety of contexts. Eighteenth-century emperors saw a connection between an individual’s language use and his legal status (civilian or Bannerman) and stated regional origin.4 Manchus should speak Manchu and people from Beijing should not sound as though they were from Shàoxīng. If they did, well, then they were in reality from Shàoxīng and should be treated accordingly. Yet some forms of Chinese were clearly favored over others. Under Yōngzhèng, language planning explicitly sought to promulgate Mandarin in Chinese officialdom, but there are signs that ability in Mandarin was desired for communication with the throne also under other emperors (e.g., the “clear and distinct” speech required for speaking to Shùnzhì). The court encouraged career-conscious Bannermen and serving and aspiring Chinese officials to speak in a certain way, and translations of Confucian literature and bilingual language pedagogy drew readers’ attention to both Manchu and Chinese as spoken languages. Manchu and “vernacular” (俗) Chinese—often explicitly defined as having a Mandarin pronunciation—were described as presenting certain advantages over Literary Chinese. Together, these developments within

54  Mårten Söderblom Saarela officialdom, education, and commercial print culture show that spoken language was politicized in the early to mid-Qing period. It was something that elite males could not afford to ignore. Developments in the nineteenth century remain much less known, especially as far as central government language policy and practice are concerned. In particular, the place of Manchu in the official linguistic order of the late Qing period deserves closer attention. Manchu did not disappear, that much is clear. For example, a late nineteenth-century vernacular Chinese novel set in the Beijing Banner milieu displays a character worrying about his son’s ability to recite his personal history when he appears before the emperor (Lǐ Tíng 2005, 176; Táng 2014, 54–55). In this case, the son—a Chinese Bannerman—had just passed the Chinese civil examinations. Even under those circumstances, he was expected to speak some Manchu. It still played a part in the political order. To conclude, let me just cite a piece of anecdotal evidence which suggests that Manchu and Mandarin remained intertwined and that the court audience represented an important institution for policing the speech of actual and would-be officials alike. In Lao She’s 老舍 (1899–1966) novel Beneath the Red Banner (正紅旗下), written in 1961–62 but left unfinished, the impoverished Manchu Bannerman Fú-hǎi 福海 reflects on language use in Manchu Beijing. The novel is set in 1899 and describes the descent of Beijing’s Bannermen into poverty. Fú-hǎi was never destined for an official career and “knew only a smattering of Manchu.” Yet even he evaluated language usage according to what would be appropriate during a court audience. Fú-hǎi, we read, “used two different types of speech.” On the one hand, he spoke “plain everyday speech, full of colloquialisms, aphorisms, painters’ jargon and some Manchu phrases commonly used by Bannermen but easily understood by Han Chinese.” On the other hand, there was “the more formal language of social intercourse” fit for interaction with superiors. Fú-hǎi “never had an audience with the emperor, but he imagined that if he were called to court for a conference on matters of state, he would use this second type of language when presenting his memorial to the throne.” This second type was characterized by “a variety of elegant phrases … as many Manchu words as possible,” which had to be enunciated using “perfect pronunciation.” Fú-hǎi’s language is that of a literate but poor male Beijing Manchu, not a nobleman trained by three private tutors like another of the novel’s characters (Lao She 2017: 31, 68, 117–18; translation from Lao She 1982: 50, 111, 192–93). Thus, one should be careful not to make too much out of Fú-hǎi’s lack of Manchu proficiency. Rather, what is interesting in this account is that Fú-hǎi imagined that a theatrical and formal form of Mandarin, heavily laden with Manchu words, was appropriate in a setting like the court audience. That was during the very last years of the Qing. What had the situation been like 50 or a 100 years earlier? This chapter is unable to answer such questions. I have simply argued that the coexistence of the Manchu and Chinese languages from the time of the Qing conquest in the mid-seventeenth century to about 1800, a period

Politicization of Language in Qing China  55 of activist rulership, contributed to the spoken language of various groups receiving increased attention from the imperial court. The existence of a varied literature of Manchu-Chinese language pedagogy suggests that the concern for vernacular languages extended beyond the government.

Acknowledgments I thank Viacheslav Zaytsev and Irina Popova (for facilitating research in Saint Petersburg in August 2016); Masato Hasegawa and the participants of the Chinese Text Reading Group, Max Planck Institute for the History of Science (for reading a source on March 13, 2017); Hasegawa and Kjell Ericsson (for interpreting a Japanese passage); Janet Theiss, Margaret Wan, Wolfgang Behr, Matthew Mosca, Xue Zhang, and Henning Klöter (for comments); Theiss (for a discussion of the paper on April 11, 2017 sponsored by the University of Utah); Ori Sela (for inviting me to present at Tel Aviv University on May 23, 2017); Joachim Kurtz (ditto at Heidelberg University on June 20, 2017); Qiu Yuanyuan (for facilitating research in Beijing in December 2017); and Nathan Vedal (for correcting mistakes). Research in Saint Petersburg and Beijing was funded by the Max Planck Institute for the History of Science.

Notes 1 See Snow’s chapter in this volume for this concept in relation to Chinese. 2 See Simmons’s chapter in this volume for further details on the pronunciation described in the Kāngxī imperial dictionary. 3 Peter Kornicki (2018: 210–12) has also recently discussed the Manchu translations of Confucian texts in the context of vernacularization. 4 This perceived link between origin and language would be an interesting point of comparison for future discussions of the much later Chinese notions of “mother tongues” or “mother dialects,” as discussed by Liang in this volume.

References Abel-Rémusat, Jean-Pierre. 1818. “Notice sur les quatre livres moraux attribués communément à Confucius.” Notices et extraits des manuscrits de la bibliothèque du roi 10: 269–426. Chū-qí-shǔ 初齊曙, and Chū-jǐng-fū 初敬敷. 1727. “Fānyì fāwēi” 翻譯發微 [Explanations of the subtleties of translation]. Held at Dalian Library with the call number M 22–161 B. Coblin, W. South. 2000. “A Brief History of Mandarin.” Journal of the American Oriental Society 120, no. 4: 537–52. doi:10.2307/606615 Crossley, Pamela Kyle. 1994. “Manchu Education.” In Education and Society in Late Imperial China, 1600–1900, edited by Benjamin A. Elman and Alexander Woodside, 340–78. Berkeley: University of California Press. Dà Qīng Mǎnzhōu shílù, Dà Qīng Tàizǔ Gāo huángdì shílù 大清滿洲實錄、大清太祖 高皇帝實錄 [Manchurian veritable records of the Great Qing; veritable records of the grand founder, Emperor Gāo of the Great Qing]. 1969. Taipei: Huáwén shūjú.

56  Mårten Söderblom Saarela Du Halde, Jean-Baptiste, ed. 1736. Description géographique, historique, chronologique, politique, et physique de l’empire de la Chine et de la Tartarie Chinoise. 4 vols. The Hague: Henri Scheurleer. doi:10.1163/9789004192560.ht-509 Durrant, Stephen. 1979. “Sino-Manchu Translations at the Mukden Court.” Journal of the American Oriental Society 99, no. 4: 653–661. doi:10.2307/601450 Elman, Benjamin A. 2000. A Cultural History of Civil Examinations in Late Imperial China. Berkeley: University of California Press. doi:10.1086/ahr/107.1.168 ———. 2014. “Unintended Consequences of Classical Literacies for the Early Modern Chinese Civil Examinations.” In Rethinking East Asian Languages, Vernaculars, and Literacies, 1000–1919, edited by Benjamin A. Elman, 198–219. Leiden: Brill. doi:10.1163/9789004279278_009 Erasmus, Desiderius. 1965. Christian Humanism and the Reformation: Selected Writings. Ed. and tran. John O. Olin. New York: Harper Torchbooks. Gāozōng Chún huángdì shílù 高宗純皇帝實錄 [Veritable records of the Eminent Exemplar, Emperor Chún]. 1986. In Qīng shílù. Beijing: Zhōnghuá shūjú. Han-i araha manju gisun-i buleku bithe [Imperially commissioned mirror of the Manchu language]. 1708. Ed. Maci and Margan. Beijing: Wǔyīng diàn. Hǎo Yùlín 郝玉麟. 1735. Fragmentary memorial. First Historical Archives, Beijing. Call number 02-01-02–2495–007. Dated Yōngzhèng 13/4/14 (May 6). Higashikawa Tokuji 東川徳治. 1919. Nan-Shi ni okeru kyōiku oyobi shūkyō no hensen 南支ニ於ケル教育及ヒ宗教ノ変遷 [Changes in education and religion in southern China]. Taiwan: Rinji Taiwan kyūkan chōsagai. Hirata Shōji 平田昌司. 2000. “Qīngdài Hónglú Sì zhèngyīn kǎo” 清代鸿胪寺正音 考 [Examination of the correct pronunciation used at the Court of State Ceremonial in the Qing period]. Zhōngguó yǔwén, no. 6: 537–44. ———. 2001. “Seidoka sareru Shindai kanwa” 制度化される清代官話 [The institutionalization of Mandarin in the Qing period]. In Min-Shin jidai no oningaku, edited by Takata Tokio, 31–59. Kyoto: Kyōto Daigaku jinbunkagaku kenkyūjo. Ishizaki Hiroshi 石崎博志. 2014. “Seion shiryō no tokushitsu” 正音資料の特質 [Characteristics of the sources on correct pronunciation]. Nihon Tōyō bunka ronshū 20: 1–30. Jiānzhù shíèr zìtóu 箋註十二字頭 | giyan ju ši el dzi teo [The twelve heads, annotated]. 1701. Attributed to Shěn Qǐliàng 沈啟亮. Beijing: Fùkuí zhāi. Incomplete copy held at the Vatican Apostolic Library with the call number BORG. CINESE 351.7. Juntu 屯圖. 2001. Yī xué sān guàn Qīngwén jiàn 一學三貫清文鑑 [Mirror of the Manchu language, which will direct you to three things when you consult only one]. Originally published in 1746. In Gùgōng zhēnběn cóngkān, vol. 723. Haikou: Hǎinán chūbǎnshè. Katayama Hyōe 片山兵衛. 1978. “Shindai Bin, Etsu chihō no seion kyōiku ni tsuite: Yōsei chō wo chūshin ni” 清代閩粤地方の正音教育について—雍正朝を中心に— [On education in correct pronunciation in Fujian and Guangdong in the Qing period, with a focus on the Yōngzhèng reign]. Chūō Daigaku Ajiashi kenkyū 2: 51–73. Kornicki, Peter Francis. Languages, Scripts, and Chinese Texts in East Asia. Oxford: Oxford University Press, 2018. doi:10.1093/oso/9780198797821.001.0001 Lao She. 1982. Beneath the Red Banner. Trans. Don J. Cohn. Beijing: Panda Books. ——— 老舍. 2017. Zhènghóng qí xià; Èr Mǎ 正红旗下 ・ 二马 [Beneath the red banner; Mr. Ma and son]. Zhènghóng qí xià dates from 1961–62. In Lǎo Shě wénjí. Chongqing: Chóngqìng chūbǎnshè.

Politicization of Language in Qing China  57 Lǐ Guāngdì 李光地 and Wáng Lánshēng 王蘭生. 1881. Qīndìng yīnyùn chǎnwēi 欽 定音韻闡微 [Imperially authorized elucidation of the subtleties of phonology]. First published in 1726. Yangzhou: Huáinán shūjú. Lǐ Huán 李桓, ed. 2009. Guócháo qíxiàn lèizhēng, chūbiān 國朝耆獻類徵初編 [Categorized and verified accounts of venerable historical figures of our dynasty, original installment]. Originally published in 1884–90. In Sānshísān zhǒng Qīngdài rénwù zhuànjì zīliào huìbiān. Jinan: Qí-Lǔ shūshè. Lǐ Tíng 李婷. 2005. Jīngzú rénjiā: “Érnǚ yīngxióng zhuàn” yǔ mínsú wénhuà 京旗人 家——《ㄦ女英雄传》与民俗文化 [People of the capital: Tale of filial heroes and popular custom and culture]. Harbin: Hēilóngjiāng rénmín chūbǎnshè. Liào Lúnjī 廖綸璣. 1996. Shíèr zìtóu 十二字頭 | juwan juwe uju [The twelve heads]. In Zhèngzì tōng, edited by Liào Wényīng, compiled by Zhāng Zìliè. Originally published in 1671. Beijing: Zhōngguó gōngrén chūbǎnshè. Líng Shàowén 凌紹雯 and Chén Kěchén 陳可臣, eds. 1699. Xīnkè Qīngshū quánjí 新刻清書全集 | ice foloho manju-i geren bithe [Complete collection of Manchu writing, newly cut]. Nanjing: Tīngsōng lóu. Liú Xiǎoméng 刘小萌. 2012. “Qīngdài Mǎnrén de jiāshú: yǐ Wán-yán Lín-qìng jiā wéi lì” 清代满人的家塾——以完颜麟庆家为例 [The family schools of Qing-period Manchus: The example of Wanggiyan Linkin’s family]. In Mǎnxué lùncóng, edited by Zhào Zhìqiáng, 2: 129–50. Shenyang: Liáoníng mínzú chūbǎnshè. Liú Yánchén 刘彦臣. 2010. “Qīngdài ‘guóyǔ qíshè’ zhèngcè yánjiū” 清代“国语骑 射”政策研究 [Studies on the “dynastic language and mounted archery” policy in the Qing period]. PhD diss., Northeast Normal University (China). Luó Chángpéi 罗常培. 2008. “Wáng Lánshēng yǔ Yīnyùn chǎnwēi” 王兰生与《音 韵 阐微》 [Wáng Lánshēng and the Elucidation of the subtleties of phonology]. Originally published in 1943. In Luó Chángpéi wénjí, 8: 430–63. Jinan: Shāndōng jiàoyù chūbǎnshè. Luther, Martin. 1938. Sendbrief vom Dolmetschen. Originally published in 1530. Ed. Rudolf Lehmann. Modernized orthography. Leipzig: Reclam. Mǎn-Hàn shíèr zìtóu 滿漢十二頭 [The twelve heads in Manchu and Chinese]. 1861. Beijing: Wénxīng táng. Held at Peking University Library with the call number X 419.1 0033. Miyazaki Ichisada 宮﨑市定. 1991–94. “Shinchō ni okeru kokugo mondai no ichimen” 清朝における国語問題の一面 [One aspect of the dynastic language issue in the Qing dynasty]. Originally published in 1947. In Miyazaki Ichisada zenshū, 14: 282–337. Tokyo: Iwanami shoten. Murata Yūjirō 村田雄二郎. 1995. “‘Bun-haku’ no achira ni: Kindai Chūgoku ni okeru kokugo mondai” 「文白」の彼方にー近代中国における国語問題 [Beyond “literary and plain”: The national language question in early modern China]. Shiso 7: 4–34. ———. 2003. “Shindai no seion kyōiku to Yōsei tei” 清代の正音教育と雍正帝 [Correct pronunciation education in the Qing period and the Yōngzhèng emperor]. Odysseus 8: 52–69. Porter, David. 2019. “Bannermen as Translators: Manchu Language Education in the Hanjun Banners.” Late Imperial China 40, no. 2: 1–43. doi10.1353/late.2019.0006 Qīndìng Bāqí tōngzhì 欽定八旗通志 [Imperially authorized comprehensive treatises of the Eight Banners]. 1987. In Wényuān Gé Sìkù quánshū, vol. 665–71. Original dating from 1786. Shanghai: Shànghǎi gǔjí chūbǎnshè. Qīngwén hòuxué jīnfá 清文後學津筏 | manju gisun-i amaga tacire urse be ibebure tasan-i bithe [Manual for the advancement of the beginning student of Manchu]. n.d. Held at Dalian Library with the call number M22–121.

58  Mårten Söderblom Saarela Rowe, William T. 1994. “Education and Empire in Southwest China: Ch’en Hongmou in Yunnan, 1733–38.” In Education and Society in Late Imperial China, 1600– 1900, edited by Benjamin A. Elman and Alexander Woodside, 417–57. Berkeley: University of California Press. Shěn Qǐliàng 沈啟亮. 1686. Shíèr zìtóu jízhù 十二字頭集註 [Collected notes on the twelve heads]. Chōnglǐ táng. Held at the library of the National Museum of China. ———. 2008. Dà Qīng quánshū 大清全書 | daicing gurun-i yooni bithe [Complete book of the Great Qing]. Original published in 1683. Shenyang: Liáoníng mínzú chūbǎnshè. Shěn Qǐyuán 沈起元, and Shěn Zōngyuē 沈宗約. 1999. Jìngtíng Gōng niánpǔ 敬 亭公年譜 [Chronological biography of Mr. Jìngtíng]. Original published in 1847. In Běijīng Túshūguǎn cáng zhēnběn niánpǔ cóngkān, vol. 92. Beijing: Běijīng Túshūguǎn chūbǎnshè. Shèngzǔ Rén huángdì shílù 聖祖仁皇帝實錄 [Veritable records of the Sage founder, Emperor Rén]. 1986. In Qīng shílù. Beijing: Zhōnghuá shūjú. Shìzōng Xiàn huángdì shílù 世宗憲皇帝實錄 [Veritable records of the Epochal Exemplar, Emperor Xiàn]. 1986. In Qīng shílù. Beijing: Zhōnghuá shūjú. Shìzǔ Zhāng huángdì shílù 世祖章皇帝實錄 [Veritable records of the Epochal Founder, Emperor Zhāng]. 1986. In Qīng shílù. Beijing: Zhōnghuá shūjú. Siegmund, Felix. 2015. “Einige Bemerkungen zu mandschurischen Übersetzungen des Mengzi, vor allem zum Mengdzi bithe.” In Auf Augenhöhe. Festschrift zum 65. Geburtstag von Heiner Roetz, edited by Wolfgang Behr et al., Bochumer Jahrbuch zur Ostasienforschung 38: 309–28. Stary, Giovanni, ed. and trans. 1983. Emu tanggû orin sakda-i gisun sarkiyan. Erzählung der 120 Alten: Beiträge zur mandschurischen Kulturgeschichte. Wiesbaden: Harrassowitz. ———. 2013. “Nurhacis Kindheit: Das größte Geheimnis der Ch’ing-Dynastie?” Originally published in 1989. In Selected Manchu Studies: Contributions to History, Literature, and Shamanism of the Manchus, edited by Hartmut Walravens, 180–83. Berlin: Klaus Schwarz. Tàizōng Wén huángdì shílù 太宗文皇帝實錄 [Veritable records of the Grand Exemplar, Emperor Wén]. 1986. In Qīng shílù. Beijing: Zhōnghuá shūjú. Takata Tokio 高田時雄. 1997. “Shindai kanwa no shiryō ni tsuite” 清代官話の資 料について [On the sources for Qing-period Mandarin]. In Tōhō Gakkai sōritsu gojisshūnen kinen tōhōgaku ronshū, 771–84. Tokyo: Tōhō gakkai. Táng Hǎiyàn 唐海艳. 2014. “Kējǔ zhìdù duì Qīngdài Mǎnyǔwén shǐyòng de yǐngxiǎng” 科举制度对清代满语文使用的影响 [The influence of the examination system on the use of Manchu language and writing in the Qing period]. Mǎnyǔ yánjiū, no. 1: 54–57. Wāng Yínfēng 汪银峰 and Yáo Xiǎojuān 姚晓娟. 2014. “Cháoxiǎncháo Yānxíng shǐ bǐxià de Mǎnyǔ” 朝鲜朝燕行使笔下的满语 [The Manchu language in the writings of Chosŏn emissaries to Beijing]. Mǎnyǔ yánjiū, no. 2: 24–28. Watanabe Junsei 渡辺純成. 2013. “Manshūgo shisō, kagaku bunken kara miru kundokuron” 満洲語思想・科学文献からみる訓読論 [“Reading-by-gloss” theory seen from Manchu philosophical and scientific documents]. In Zoku “kundoku” ron Higashi Ajia kanbun sekai no keisei, edited by Nakamura Shunsaku et al., 220–59. Tokyo: Iwata shoin. Wei, Shang. 2014. “Writing and Speech: Rethinking the Issue of Vernaculars in Early Modern China.” In Rethinking East Asian Languages, Vernaculars, and

Politicization of Language in Qing China  59 Literacies, 1000–1919, edited by Benjamin A. Elman, 254–301. Leiden: Brill. doi:10.1163/9789004279278_011 Wǔ-gé 舞格. 1730a. Qīngwén qǐméng 清文啟蒙 | cing wen ki meng bithe [Manchu language primer]. Beijing: Sānhuái táng. ———. 1730b. Mǎn-, Hànzì Qīngwén qǐméng 滿漢字清文啟蒙 [Manchu language primer in Manchu and Chinese script]. Revised edition. Published after 1772. Beijing: Hóngwén gé. Yamazaki Masato 山崎雅人. 2000. “Lùn Mǎnwén Shījīng xīn, jiù fānyì zhī chàyì” 论满文《诗经》新旧翻译之差异 [On the differences between the old and new Manchu translations of the Poetry classic]. In Mǎnxué yánjiū dì liù jí, edited by Yán Chōngnián, 246–61. Beijing: Mínzú chūbǎnshè. Yáng Bǐng 楊炳. 1734. Routine memorial. First Historical Archives, Beijing. Call number 03–0039-002. With an imperial rescript dated Yōngzhèng 12/8/3 (August 31). Yáng Yàgēng 杨亚庚. 2005. “Qīng shílù suǒ jiàn Qīng qiánqī yǔyán wénzì zhèngcè” 《清实录》所见清前期语言文字政策 [Language and script policy of the early Qing period as seen in the Qing veritable records]. Master’s thesis, Jilin University. Yeh Kao-shu 葉高樹. 1992. “Shījīng Mǎnwén yìběn bǐjiào yánjiū: Yǐ ‘Zhōu nán,’ ‘Shào nán’ wéi lì” 《詩經》滿文譯本比較研究—以〈周南〉、 〈召南〉為例 [A comparative study on the Manchu translations of the Poetry classic using the examples of the “Odes of Zhōu and South” and “Odes of Shào and South”]. Guólì Táiwān Shīfàn Dàxué lìshǐ xuébào 20: 219–34. ———. 2002. Qīng qiánqī de wénhuà zhèngcè 清朝前期的文化政策 [Cultural policies of the first half of the Qing period]. Bǎnqiáo: Dàoxiāng chūbǎnshè. ———. 2006. “Mǎn-Hàn hébì Qīndìng fānyì wǔjīng sìshū de wénhuà yìhán: Cóng ‘yīn guóshū yǐ tōng jīngyì’ dào ‘yīn jīngyì yǐ tōng guóshū’” 滿漢合璧《欽定繙 譯五經四書》的文化意涵：從「因國書以通經義」到「因經義以通國書」 [The cultural import of the Manchu-Chinese bilingual Imperially authorized translations of the Five Classics and Four Books: From “Grasping the meaning of the classics through the dynastic language” to “Grasping the dynastic language through the meaning of the classics”]. Jīngxué yánjiū lùncóng 13: 1–42. Yeung Man Shun 楊文信. 2000. “Shìlùn Yōngzhèng, Qiánlóng niánjiān Guǎngdōng de ‘zhèngyīn yùndòng’ jí qí yǐngxiǎng” 試論雍正、乾隆年間廣東“正音運動” 及其影響 [A take on the “correct pronunciation movement” and its influence in Guangdong in the Yōngzhèng and Qiánlóng periods]. In Dì qī jiè guójì Yuè fāngyán yántǎo huì lùnwén jí, edited by Dān Zhōuyáo and Lù Jìngguāng, 118–36. Beijing: Shāngwù yìnshūguǎn. Yuán Yīzhōu 袁一州. 1729. Guānyǔ xiángbiān 官語詳編 [Careful compilation of the language of officials]. Held at the Institute of Oriental Manuscripts, Russian Academy of Sciences in Saint Petersburg with the call number C 723. Zhōngguó Dìyī Lìshǐ Dǎng’àn Guǎn 中国第一历史档案馆, ed. 2011. Qiánlóng cháo Mǎnwén jìxìn dàng yìbiān 乾隆朝满文寄信档译编 [Manchu court letters of the Qiánlóng reign, edited and translated]. Changsha: Yuèlù shūshè.

3

Romanizing Southern Mǐn Missionaries and the Promotion of Written Chinese Vernaculars Don Snow

Abstract This chapter analyzes the promotion of written vernaculars in southeastern China by Western missionaries during the late nineteenth century. For example, in the city of Xiamen, the missionaries’ ideological support of vernacular literacy was translated into an ambitious project of teaching an alphabetic orthography of the regional vernacular to local illiterates. The promotion of written vernaculars by missionaries was a strategy that radically broke with Chinese cultural conventions. The analysis of the language planning strategies and the underlying language ideologies is based on a close reading of missionary writings such as mission journals, letters, and prefaces to dictionaries. The chapter concludes that missionary language planning, despite great successes during its period of greatest application, only left a small historical legacy. Moreover, it is shown that outsider groups such as missionaries played an important role as early promoters and users of written vernaculars in China.

Introduction One of the most significant changes in the linguistic landscape of East Asia in recent centuries has been the decline of Classical Chinese and the rise of written vernaculars. Pre-modern East Asia was a sinophone world in which Classical Chinese was the predominant written language, and serious texts were rarely written in the vernacular. Gradually, however, vernacular styles of writing developed in China, Korea, Japan, and Vietnam, and modernist reforms that accelerated beginning in the late nineteenth century led to a dramatic shift in language roles, with Classical Chinese being replaced by once-denigrated vernacular styles of writing. Of course, a similar shift occurred in many other societies around the world, with “traditional diglossia” (Snow 2012) being displaced by new patterns in which nations promoted standard national languages with written forms that adhered closely to the national spoken language. One interesting feature of the rise of written vernaculars in the sinophone world was the pioneering role often played by missionaries. An early example of this is the use of relatively vernacular writing styles by early Buddhist

Romanizing Southern Mǐn  61 missionaries in China. They believed that Buddha advocated using the vernacular, and that directly quoting the master’s oral teaching emphasized the reliability of the message; they were also influenced by Indian cultural traditions, which were relatively accepting of vernacular writing. Many of their practices, such as chanting, recitation, and spreading the word through storytelling, also encouraged use of the vernacular, as did the fact that they were often reaching out to people from lower levels of society who were less literate. Finally, many started learning Chinese as adults and presumably knew the vernacular better than they did Classical Chinese (Mair 1994; see also Hanan 1981). Another example is the role Jesuit missionaries played in creating written Vietnamese. In the seventeenth century Jesuits Romanized Vietnamese as part of their effort to learn Vietnamese for missionary work,1 and in the nineteenth century colonial government schools began teaching this system. While Romanized Vietnamese (Quoc Ngu) was initially disparaged by Vietnamese nationalists as collaborationist, eventually even they adopted and promoted it because it was easy to teach and use (DeFrancis 1977; Le and O’Harrow 2007). Yet another example is the projects of Protestant missionaries in China in the late nineteenth century and early twentieth century to translate the Bible and other texts into regional Chinese vernaculars. Some of these written vernaculars used Chinese characters; for example, Chinese-character vernacular versions of the Bible were published not only in Mandarin but also in Shanghainese, Hakka, Cantonese, and Fuzhounese. In other areas such as Ningbo, Hainan, Taizhou, and Wenzhou, missionaries created Bibles in Romanized versions of the regional vernacular (Cài 2012; Woodin 1890; Yóu 2003). The largest of these Protestant Romanized vernacular projects was carried out by mission groups in southeast China who actively collaborated to create and promote Romanized forms of Southern Mǐn vernaculars— Amoy, Swatow, and Taiwanese.2 This project was the most sustained Protestant missionary effort to popularize a written local Chinese vernacular, lasting more than a century. It was also the most successful such project; at its peak tens of thousands of people were literate in some form of Romanized Mǐn. Finally, it was the most ambitious project, with missionaries investing copious amounts of time into translating and writing books and periodicals in different varieties of Romanized Mǐn. Given the well-known commitment of Protestant missionaries to producing local language Bible translations, it is easy to overlook how ambitious the Romanized Mǐn project was. Of course, most of these missionaries did not expect Romanized forms of Mǐn to replace the existing written Chinese language––though a few did harbor such dreams––but they did assume that these new vernacular-based written languages in the Roman alphabet would be embraced by Mǐn-speaking Chinese Christian communities even despite the fact that China already had a well-established written tradition.

62  Don Snow In this regard China was quite different from most places where Protestant missionaries created new Romanized written vernaculars. In places such as Tahiti and Hawai’i, no previous written language existed, so missionarycreated written vernaculars quickly gained wide acceptance in part because there was no indigenous competitor. In contrast, China’s written language tradition was ancient and well-established, and its writing system was not only omnipresent in Chinese society but also respected and even venerated. In light of these realities, the choice made by Protestant missionaries to invest so much time and effort in Romanized Mǐn vernaculars––and to expect success in this effort––becomes quite striking. One goal of this chapter is simply to tell the story of how Protestant missionary groups in southeast China developed and promoted Romanized forms of Mǐn vernaculars. English Presbyterian Church missionaries were active in all three of the relevant areas––southern Fujian, northeastern Guangdong, and Taiwan––and hence played an especially important role in this project. However, important roles were also played by missionaries from the Reformed Church in America, the London Missionary Society, and eventually Canadian Presbyterians. Missionaries from these groups also became influential advocates of the Romanization strategy and were deeply involved in missionary debates about how best to make the Bible and other Christian texts available in China’s many regional languages. A second goal is to examine why these missionaries were willing to invest so much time and effort in such an ambitious project––attempting to popularize a new written language using an alien script in a society where an indigenous written language was well-established. Fortunately for our purposes, many of the missionaries involved in the project wrote extensively, and we can learn much about their motives from what they stated explicitly and also what we can read between the lines.

The Romanized Mǐn Project Amoy Protestant mission work in Amoy (Xiamen)3 was begun in the early 1840s by missionaries from the Reformed Church in America, the London Missionary Society, and the English Presbyterian Church. Right from the start, these three groups, similar in theology and church governance, worked closely together as “practically one mission” (Pitcher 1893: 93). The Amoy missionaries were vexed by the problem of illiteracy among their converts, so in 1850 Presbyterian missionary James Young and Reformed Church missionary John Van Nest Talmage began experimenting with Romanized Amoy. As Talmage wrote in a letter home: Some of us are now trying an experiment whether by means of the Roman alphabet the Sacred Scriptures and other religious books may not

Romanizing Southern Mǐn  63 be given to the Christians and to any others who cannot read, but who take enough interest in Christianity to desire to read the Scriptures for themselves…. Dr. James Young, an English Presbyterian missionary, has commenced teaching the colloquial, as written with the Roman alphabet, in his school…. We are of the opinion that almost every member of the church can soon learn to read by this system. (Fagg 1894: 106f.) The Amoy missionaries soon concluded that Romanized Amoy was an effective literacy tool, and set to work producing Christian literature in it. By 1852 Talmage had produced a primer for Romanized Amoy, and the next year he finished a reader. Meanwhile he and others began translating portions of the Bible into Romanized Amoy, and by 1873 there was a Romanized Amoy version of the entire New Testament, with the Old Testament completed in 1883 (Pitcher 1893). In 1888 the three missionary groups also began publishing a monthly church periodical in Romanized Amoy (White 2017). In addition to the Bible, eventually approximately 50 works were published in Romanized Amoy, including not only Christian works such as Pilgrim’s Progress, the Heidelberg Catechism, and the Life of Paul but also textbooks on physiology, geography, astronomy, arithmetic, algebra, Chinese history, and ancient Egyptian history, and even Confucian classics such as the Great Learning and Doctrine of the Mean (Pitcher 1893, 1912; Xu and Li 1999). The Amoy missionaries also devoted much of their time to educational work, establishing schools for both boys and girls and training classes for “Bible women,” and in these schools Romanized Amoy was taught (DeJong 1992). Granted, some mission schools also taught students Classical Chinese, but the missionary schools in this region taught graduates to read and often also write Romanized Amoy. Formosa In the mid-1860s the English Presbyterians extended their mission work to Formosa (Taiwan), where Southern Mǐn was also widely spoken. In 1865 James and Mary Maxwell moved to the island and soon concluded that widespread illiteracy was a major problem and Romanized Amoy was the answer. Maxwell described their strategy as follows: The primer already printed at Amoy was further simplified and abbreviated, so that two small leaves of cardboard, four pages in all, contained a perfectly sufficient arrangement and number of lessons from which to proceed to the reading of the Gospel. This primer could be carried about in the breast or sleeve without the slightest difficulty, and was at hand for use in the house or in the workshop, in the fields or on the hillsides, when any spare moments presented themselves. … Patients in

64  Don Snow hospital also, whose hands are tolerably empty, are encouraged to use the opportunity of mastering the art of reading in their own language; and not a few have without difficulty succeeded in the attempt. So simple is the system that one native Christian can teach another and many of the readers in Formosa have been so taught. (Gibson 1888: 38–39) Promotion of Romanized Amoy gained additional impetus with the arrival of Presbyterian missionaries William Campbell (1841–1921) in 1871 and Thomas Barclay (1849–1935) in 1875. Barclay soon became an especially zealous advocate of Romanized Amoy, writing: Soon after my arrival in Formosa I became firmly convinced of three things…. The first was that if you are to have a healthy, living Church it is necessary that all the members, men and women, read the Scriptures for themselves; second, that this end can never be attained by the use of the Chinese character; third, that it can be attained by the use of the alphabetic script, the Romanized Vernacular. (Band 1936: 67) In 1885 Barclay established a Romanized church newspaper, the Kau-Hoe-po (Church News); he was also a major contributor to a revised version of the Romanized Amoy New Testament in 1916 and the Old Testament in the 1930s (Band 1936, 1972; Ion 1993). Swatow In 1858 English Presbyterian missionaries also expanded their work in Mǐn-speaking regions by establishing a mission in Swatow (Shantou), where the spoken language is closely related to the vernaculars of southern Fujian. While American Baptists also began work in Swatow at approximately the same time, and translated the Bible and other Christian texts into the Swatow vernacular, they chose to take a different approach and promoted a form of written Swatow in Chinese characters (Cài 2012; Snow and Chen 2015). By 1874 Presbyterian missionary William Duffus was working on a Romanization system for the Swatow vernacular, editing a Romanized Swatow dictionary, and translating the Gospel of John (Giedt 1946; Hood 1986). However, the arrival of a new missionary, John Campbell Gibson (1849– 1919), kicked the project into a higher gear. Even before arriving in Swatow, Gibson had learned about the Romanization project from Amoy missionary Carstairs Douglas (1830–77), author of the Dictionary of the Amoy Vernacular. Gibson’s enthusiasm for the project was further stoked by a visit to Douglas in Amoy in 1875, at a time when Thomas Barclay was also present while en route to Formosa (Hood 1986).

Romanizing Southern Mǐn  65 Gibson and Duffus published a primer for Romanized Swatow in 1875, and Duffus’ translations of the Gospels of John and Luke followed in 1877. During the next decade Gibson devoted much of his time to producing a dictionary, A Swatow Index to the Syllabic Dictionary of Chinese, which was based in part on Douglas’ Dictionary of the Vernacular of Amoy, and he also taught Romanized Swatow to students in various schools run by the Presbyterian mission (Hood 1986: 117). In part due to encouragement from the English Bible Society, in 1887 a renewed translation effort was begun by Gibson, Duffus, and others such as Hur MacKenzie and Patrick Maclagen, and in the next few years Romanized Swatow translations of Genesis, I– III John, Acts, Mark, Matthew, and other books were published; the entire New Testament appeared in 1905 (Cài 2012; Wong 2015). Also, as in Amoy and Formosa, the mission published a Romanized church newspaper. Gibson’s role as a leading advocate of Romanization began with the publication of two pamphlets, Native Christian Church in China and Learning to Read in South China. In the first of these, Gibson reminds his audience of the important role the vernacular Bible translations of Wycliffe and Tyndale played in the life of the Scottish Church and argues that similar vernacular versions of the Bible and other Christian literature are needed in southeast China (Hood 1986). The second pamphlet argues that it is essential that church members become literate, and that the only way this could happen in southeast China is through use of Romanized vernaculars. It was due to this second pamphlet that he was invited to write a major report, Review of the Various Colloquial Versions and the Comparative Advantages of Roman Letters and Chinese Characters, for the General Conference of the Protestant Missionaries of China in Shanghai in May 1890. The Decline of Romanized Mǐn By the early twentieth century, much had been published in Romanized Amoy and, to a lesser extent, Romanized Swatow. Considerable progress had also been made toward building a community of people who could read Romanized forms of Mǐn; as of the early 1900s Pitcher estimates that at least several thousand people could read Romanized Amoy (1912), and this number continued to grow as churches taught it through schools and literacy campaigns (Xǔ and Lǐ 1999). Even as of the late 1930s missionaries were still running short-term literacy classes to teach Romanized Mǐn (de Velder n.d.). Eventually as many as 100,000 people may have been literate in some form of Romanized Mǐn (Klöter 2005; Xǔ and Lǐ 1999). However, on the Chinese mainland, by the 1920s, the high tide of the Romanized Mǐn project had passed and churches were growing less zealous in their promotion efforts (Xǔ and Lǐ 1999). One major reason was that by the 1920s the Chinese government was actively promoting mass education and literacy in Mandarin, and while the new national written language was based on a northern vernacular that differed considerably from the Mǐn

66  Don Snow vernaculars, it was still easier to learn than Classical Chinese. As more people learned to read Mandarin, there was less need for them to learn to read in a local vernacular; in fact, for people who already knew Mandarin, learning to read a Romanized vernacular became an additional burden rather than a shortcut to literacy. The early twentieth century also saw the publication of the Union Version of the Bible in Mandarin, which was vigorously promoted by churches and Bible societies. This reduced both need and enthusiasm for promoting Bibles in other Chinese vernaculars (Cài 2012; Mak 2017). Also, during the twentieth century, there was growing opposition to imperialism and a growing sense of nationalism and pride in Chinese identity. In this new climate, churches increasingly promoted literacy in Mandarin not only as a social service but also as a way for churches to identify with the Chinese nation (Mak 2017); also, the Western writing system used in Romanized Mǐn was increasingly stigmatized (Hood 1986; Xǔ and Lǐ 1999). On the mainland the death knell for Romanized Mǐn was sounded in the 1950s when the new Communist government began vigorously promoting Pǔtōnghuà, and actively discouraging use of local vernaculars in writing. These new policies did not cause Romanized Mǐn to vanish overnight, and even in the 1990s Romanized Mǐn Bibles could sometimes be seen in churches in southern Fujian (Xǔ and Lǐ 1999). However, this was a relic of the past rather than the wave of the future. In Taiwan4 the story of Romanized Mǐn unfolded quite differently. After the 1894–95 Sino-Japanese War, Taiwan was ceded to Japan, and over the next few decades one main goal of the Japanese colonial school system was to teach Taiwanese people to speak and read Japanese, in part to encourage a sense of loyalty to Japan (Heylen 2012; Kerr 1974; Tsurumi 1977; see also Chen, this volume). Initially the new colonial schools continued to teach pupils to read at least some Chinese, and even taught basic Romanized Taiwanese so that students could read labels printed on products like sugar (Tsurumi 1977). In the 1930s, however, the government took a harder line, promoting Japanese more vigorously (Ts’ai 2009). Presbyterian churches in Taiwan continued to function mainly in Taiwanese in preference to Japanese. In a society where the colonial power was Japan rather than a Western country, the foreignness of a Romanized (Western) writing system was not a problem to the extent it was on the mainland. Instead, in the 1930s as the Japanese government put increasing pressure on churches to use Japanese and to stop using Romanized Bibles and hymnbooks, Romanized Taiwanese gradually came to be an embattled in-group language, associated not only with Christian identity but also increasingly with local Taiwanese identity (Lin 1999). This association of Romanized Taiwanese with Taiwanese identity continued to strengthen after 1945 under the Nationalist Party (KMT) government. Until the end of martial law in 1987, the KMT actively promoted Mandarin through the school system and the media, and also discouraged use of both spoken and written Taiwanese (Klöter, this volume). Children

Romanizing Southern Mǐn  67 were not allowed to speak Taiwanese in schools, and there were strict limits on radio and television broadcasting in Taiwanese. Furthermore, the teaching of Romanized Taiwanese was banned, Romanized Taiwanese texts could not be published, and Romanized Bibles were sometimes seized by the government. In the minds of many Taiwanese such measures cemented the association between Romanized Taiwanese and Taiwanese identity, and even outside church circles many Taiwanese became positively disposed toward Romanized Taiwanese (Lin 1999); this association was heightened by the fact that during the years of martial law, the Presbyterian Church was frequently the most organized and visible advocate of Taiwanese identity. By the 1960s Romanized Taiwanese was increasingly viewed as a symbol of local Taiwanese identity rather than a reminder of foreign religions or nations. In short, during the twentieth century Romanized Mǐn did not decline nearly as much in Taiwan as it did in mainland China. Of course, as mass education led to dramatically higher literacy rates, its value as a tool for reducing illiteracy declined. However, in stark contrast to the mainland, where Romanized Mǐn declined in part because of its foreign associations, in Taiwan it came to be associated with local identity and culture.

Why Did Missionaries Promote Romanized Mǐn? Gibson succinctly states the case for Romanized Mǐn as follows: 1) The bulk of the Chinese people cannot read their Book language, and experience shows us, from the nature of the case, they never will. 2) No people has ever learned to read except in their own Vernacular. 3) The Christian Church must teach all of its members to read (unless disabled by age or infirmity), and must therefore use the Vernaculars. 4) No people has ever learned a hieroglyphic system, and the native written character, being of this class, is too cumbrous to be made the medium for writing the Vernaculars of South China. 5) The Roman alphabet gives an easy and complete solution of the difficulty; and experience has shown that even uneducated people easily learn to read and write their mother-tongue by means of it. (Gibson 1888: 22) Gibson’s final point sums up the most fundamental reason why most Protestant missionaries in southeast China were willing to invest so much time and effort in promoting Romanized Mǐn—Romanized vernaculars were easy to learn to read. The ease with which one could become literate in Romanized Mǐn was not just an advantage for personal study; it also made it easier for people to read Bible passages aloud during church services. Early Chinese Bible translations were written in Classical Chinese, and when read aloud in

68  Don Snow church, especially to congregations in which many were not literate, the text had to be translated on the spot into the local vernacular. Given that many preachers and evangelists did not have a strong command of Classical Chinese, problems and mistakes were evidently quite common. “Even natives who are fairly good scholars, fail greatly in this most difficult task. Their translations are sometimes loose, sometimes inconsecutive, often stiff and obscure, frequently incorrect, and sometimes wholly meaningless” (Gibson 1890: 69). Using Bibles in Romanized Mǐn eliminated, or at least greatly reduced, such difficulties. Romanized Mǐn was also easy to learn to write. Here the advantage was not simply that the writing system took less time to master, but also that using the Romanized vernacular freed writers from conventions strongly associated with written Chinese. As Woodin explains: Easy and accurate communication of thought by letter correspondence is evidently far more difficult in the Chinese character than in the Romanized, for all but the scholars, and they are constrained to use old forms of sentences that must be both translated, and then explained, before they convey the intended meaning. The Romanized, starting in the new present, need not be confined to the expressions handed down for scores of generations, which constantly hamper the Chinese writer. (Woodin 1890: 97) Writing in the Romanized vernacular allowed people to express thoughts using the same words they would naturally use in oral communication, which made the process of going from thought to word easier and smoother. Also, as Gibson points out, the written vernacular “reaches the heart direct” (1890: 70). Finally, Romanized Mǐn was easier for foreign missionaries to learn, and decreased the time they needed to become literate in Chinese (DeJong 1992). This is not to say that missionaries in the region only learned Romanized vernaculars. As newly arrived Reformed Church missionary Tena Holkeboer (1895–1965) writes in a letter to her family in 1921, …now we need not depend on the character, but it is still used so much in China that we have to learn the characters also. …. If it were not for these old characters, I’d soon be ready for my first examination. (Holkeboer 1921; emphasis as in the original) For Holkeboer—and many others like her—the experience of learning Chinese characters left her keenly aware of how much easier the Romanized was. Here we should note that the missionaries were well aware that Chinese society did not share their enthusiasm for writing in the vernacular; instead, Chinese expected serious texts to be written in what missionaries called “the

Romanizing Southern Mǐn  69 book language”—Classical Chinese—and Chinese generally viewed texts written in vernacular styles as being quite vulgar (cf. Söderblom Saarela, this volume). By and large Chinese people associated vernacular writing styles with low prestige kinds of texts. As J. S. Burdon writes: In most dialects there are probably many colloquial productions printed and sold, but they are all of a very low character from a moral point of view. They are said to consist of low songs or other rhyming compositions that appeal only to depraved minds…. (1890: 99) As a result, there was often opposition to Romanization. Pitcher describes this opposition as follows: Among those who wish to be classed as literary it has, to be sure, never found a warm reception. To them it is poor style. To devote any time to it is a waste of energy over childish things. To those who have no claim to being literary in any sense whatever, it has not always appealed as one might have expected it would. Rather than being seen reading it, or learning to read it, they prefer to remain ignorant, and so give it a wide berth. (Pitcher 1912: 208–9) Furthermore, “[s]ome were actually ashamed to be seen reading it” (Pitcher 1893: 194). Gibson was also keenly aware of local resistance and pleads with his fellow missionaries: Let us not do anything to encourage the idea which is apt to find favour among half-educated Chinese, that Romanized Vernacular is good enough for women and children, and for newly arrived missionaries, but quite beneath men who have reached the higher level of a smattering of the Book language. (1888: 23; see also Fagg 1984; “Vernacular Translations” 1901) Another source of resistance to Romanized vernaculars was Chinese reverence for written Chinese characters—and prejudice against foreign writing systems. As missionaries knew, Chinese people had enormous respect for Chinese characters, tending to equate mastery of Chinese characters with learning itself. As Gibson notes, one main obstacle in the path of Romanized Mǐn was “… deep-rooted blind reverence for the native character, which makes many think stumbling and stammering unintelligibly over the characters a far greater attainment than the most fluent and intelligent reading of their vernacular in Roman letter…” (1890: 88; see also 1888; Band 1936). Due to this positive association in the Chinese mind between learning and Chinese characters, many missionary-sponsored schools in Mǐn-speaking

70  Don Snow regions had to offer at least some training in Chinese characters and Classical Chinese in order to attract students. Chinese people also viewed Romanization as alien; as Pitcher notes: “Some despised [Romanized Mǐn] simply because it was too foreign” (1893: 194). Use of a Western alphabet clearly marked Romanized vernaculars as a “foreign importation” (Band 1936: 69), and even Gibson expressed concern that “… use of a foreign method of writing gives a foreign aspect to our religious teaching…,” going on to note that “…a purely native system, if such a thing were possible, would be preferable” (1888: 19). Even worse, some Chinese viewed Romanization as “… an attempt to extirpate the book language from the Christian community, which meant as much as cutting off the converts from their political and social connections…,” and sometimes converts used it more because the missionaries wanted them to than because they wanted to do so themselves (Woodin 1890: 94, 96). While keenly aware of prejudices against vernacular writing, most missionaries did not share them. Coming from societies in which classical written languages like Latin had largely been replaced by modern written vernaculars—a change in which Protestant Christianity had played a major role—Protestant missionaries tended to have positive views of vernacular-based writing styles, and to feel that Chinese society would generally be improved by adopting such styles. For example, writing of the advantages of writing in Romanized colloquials, S. F. Woodin notes: [G]etting out of these deep and terribly jolting ruts [of classical writing conventions] may tend in due time to awaken and give an immense impulse to the Chinese mind. In connection with the free knowledge of the Bible, perhaps it may cause an awakening of the Chinese intellect, like that which the literary use of the English, and the other European languages, instead of Latin, caused in Europe. (Woodin 1890: 97) Similarly, Gibson claims that “…substitution of Roman Vernacular for the Book language in letter-writing would go far to abolish in that department the artificiality and falseness by which Chinese society is honeycombed in all directions” (1888: 29). So, while the missionaries knew that writing serious texts in the vernacular was a dramatic break from Chinese tradition, many felt this was a good thing rather than a problem. Missionaries also tended to emphasize the disadvantages of the Chinese writing system. The most lurid portrayal comes from Pitcher, who describes written Chinese as one of the most difficult languages in the world. Whether it be supposition or fact that Satan was the author of the Chinese language or not, it is nevertheless true that there is no other nation that has been for so long and so completely under sway as China. The

Romanizing Southern Mǐn  71 language has been one of the highest and strongest walls that has surrounded this nation. (Pitcher 1893: 116) However, even more even-keeled promoters of Romanization like Gibson tended to highlight the challenges posed by the traditional Chinese writing system, arguing that a Hieroglyphic system, of which the Chinese Book language is the most perfect representative, has never produced a reading people. …. The arts of reading and writing in any such system have always and everywhere been the exclusive possession of a literary caste, and China is no exception to this rule…. This difficulty is not to be overcome by increased attention to education, nor by better educational methods. …. No people has ever learned such a system, and it is safe to prophesy that no people ever will. (Gibson 1888: 13) Also, responding to concerns that Romanized Mǐn was simply too foreign, missionaries tried to emphasize the fact that even though its foreign writing system made Romanized Mǐn appear alien, it was actually a better representation of the local spoken language than Classical Chinese was. Gibson argues: The Romanized Vernacular is foreign, it is true, to the eye, but not to the ear. When it is read aloud it is more racy of the soil than the Book language itself, and a Chinaman is delighted at hearing the familiar sounds of his own tongue read out to him from a foreign-looking medium. His conclusion is that there is not such a gulf as he supposed between what is native and what is foreign. He is gratified to find that his native tongue, despised as it is by his own literary men who yet speak no other, and ignored as it is by the native magistrates, is so esteemed by these ingenious foreigners, that they have actually applied to it one of their many skillful inventions. The Romanized Vernacular readily lends itself as a healer rather than an irritant of international prejudice. (Gibson 1888: 19) Gibson clearly hopes that over time the local language represented through the medium of Romanized Mǐn would become more salient in Chinese minds than the foreign nature of the writing system in which it is embodied. Finally, the missionaries professed confidence that eventually any prejudice toward Romanized Mǐn would be overcome by its obvious superiority. As Gibson writes,

72  Don Snow The Chinese have a keen eye for what is useful, whether it be new or old… We may be sure that in time the Christian people will recognize the boon that is offered them in this simple embodiment of their mother tongue. (1888: 15–6; see also Gibson 1890: 81; Woodin 1890: 98) In the late nineteenth century and even early twentieth century, missionary optimism about gradual acceptance of Romanized Mǐn is not hard to understand. In a time of transition when China was beginning to question its traditional culture and gradually accept elements of Western culture it would not have seemed entirely unreasonable to hope that China’s affection for its writing system would gradually fade, and so would objections to a new and easier writing system—even an imported one. Gibson’s optimism in this regard was encouraged by a visit to Japan where he discovered interest in Romanized forms of Japanese and even a newly established Roman Alphabet Association (Gibson 1888: 23–24). One final objection to Romanized vernaculars came primarily from the Protestant missionary community. In summarizing arguments against Romanized vernaculars, Gibson notes that the objection “…which will perhaps be felt of greater weight than any other” is that “a book in the Romanized Vernacular of any district can only reach that limited region in which the dialect is spoken…” (1888: 190). Conceding the seriousness of this problem, Gibson responds by pointing out that the regions in question are large enough in size and population that they deserve their own written languages; he also points out that the regional nature of spoken vernaculars in Europe did not prevent the replacement of Latin by written languages based on vernaculars such as French, German, and English. Reminding his readers of the imperative to carry the Christian gospel into the whole world, Gibson claims that it is essential to have “…the Word of God in the Vernacular of every country. China, in this regard, is not one country, but many” (1888: 31, emphasis as in the original). Underlying this argument is a historical awareness that Europe had experienced transitions similar to those which seemed to be taking place in China. The Holy Roman Empire existed as a reminder that once Europe had been united politically under Roman rule and culturally under Latin and (Catholic) Christianity. However, especially after the Reformation, it evolved into separate nation states in which new vernacular-based national written languages replaced Latin. From the perspective of a Westerner in China in the late nineteenth century or early twentieth century, it would not have seemed unthinkable that China might travel a similar road. The linguistic differences between regions, especially in the southeast, were fully as great as those in Europe, and the weakness of the central government during the final decades of the Qing dynasty and the first decades after the 1911 Revolution also made China’s future as a united nation seem less than certain, especially to outside observers who were

Romanizing Southern Mǐn  73 not firmly wedded to the idea that China would inevitably remain a single nation.

Discussion The reasons why the southeast China missionaries were willing to invest so much effort in Romanized Mǐn, and why they believed it had a good chance of success despite the substantial opposition the project faced, appear to fall into two general categories. Put another way, there are two general attributes of these foreign missionaries which help explain why they were willing to invest in a project that was so contrary to China’s established language conventions. The first attribute is that these missionaries were foreign, cultural outsiders who did not fully share the beliefs and conventions concerning language that would have been found among educated Chinese of the period. Keep in mind that in the late nineteenth century and early twentieth century many Chinese were also questioning the assumptions that Classical Chinese should be China’s dominant written language and that literacy should be limited to a relatively small elite; thus missionaries were not alone in considering new approaches. What distinguished these missionaries was the extent to which they were investing in a solution which represented a dramatic break with Chinese tradition. As we have seen, these missionaries were from cultures where it had long since been considered quite normal to write more or less as one speaks. In the West older diglossic patterns of societal multilingualism had long since vanished, and while Latin was still often taught in schools it was no longer assumed that Latin was the only language suitable for educated written discourse. From the cultural perspective of these missionaries, the accusation that writing in a local vernacular like Mǐn was “vulgar” simply would not have carried much weight, and it is not surprising they believed Chinese people would eventually come to realize there was nothing wrong with written languages based on spoken languages, especially when it became clear how much easier it was to learn vernacular-based written languages. As outsiders these missionaries also owed less loyalty to China’s civilization and written tradition. They had not been socialized from childhood into the norms and practices of education in Chinese society, and did not owe their social position in Chinese society to mastery of the conventions of China’s literary tradition, so they had fewer reasons to abide by or uphold the conventions than would a Chinese person with similar levels of education (Snow 2013). No doubt some missionaries were actively hostile to the Chinese written tradition. For example, Pitcher writes of the Chinese literary tradition: After wading through volume after volume filled with the deeds of rulers and princes….it will be discovered that there are no such treasures

74  Don Snow of thought, no such storehouses of knowledge, of philosophy, of science, and of travel, etc., as will be found in the fields of Western literature. Chinese literature is like a great wide ocean of books—books everywhere—yet with comparatively little to quench or satisfy the thirst for knowledge and truth. (Pitcher 1912: 172) Clearly Pitcher is not describing a tradition toward which he feels much affection. Also, as we saw earlier, learning to read the Chinese written language was a considerable struggle for many missionaries, and this would also have predisposed many of them to take a somewhat jaundiced view of it. However, many missionaries had a much less negative view of the Chinese written tradition. For example, Gibson writes of Classical Chinese: “As a literary system, the book language is a marvel of ingenuity, of subtle versatility, and condensed force. It attains a surprising variety of expression by the use of the most unpromising materials…” (1888: 17). In fact, in addition to working on Bible translations into Romanized vernaculars, Gibson also worked on a committee to produce a new translation of the Bible into “easy Wen-li” (Campbell 1996: 362–63; Hood 1986: 124). But, while Gibson and many other missionaries had a relatively positive view of Classical Chinese, we can safely assume few if any missionaries had the level of loyalty to the Chinese written tradition felt by educated Chinese. Furthermore, they probably under-estimated the cultural pride which Chinese people felt toward Chinese characters. Granted, the nature of this pride changed over time. In the nineteenth century this was more likely to be manifested in what the missionaries referred to as “reverence” for Chinese characters, but over time it evolved into pride in Chinese characters as a symbol of national culture. What did not change was the reality that Chinese did not see their written language simply as a communication tool that could readily be discarded if a better one came along. Instead they saw it as a proud symbol of Chinese culture. In contrast, the foreign missionaries tended to see the Chinese written language mainly in practical communication terms, and to assume that eventually the practical advantages of Romanized Mǐn—and the fact that the underlying spoken language was a variety of Chinese—would overcome resistance to use of a writing system that was so evidently foreign. Looking more broadly to other cases of missionary involvement in pioneering written vernaculars, it seems safe to assume that the foreign background of the missionaries was generally part of the reason why they were willing to make relatively heavy use of vernacular writing styles. Referring back to the beginning of this chapter, clearly one reason Buddhist missionaries were more favorably inclined to vernacular writing styles was that they were from cultures outside China in which writing styles were normally closer to the spoken language; also, elements of this foreign faith and its practice, such as the high value placed on faithful preservation of oral sayings, encouraged greater use of the vernacular. The case of Romanized

Romanizing Southern Mǐn  75 Vietnamese is more complicated in that the Jesuit missionaries only developed Romanized Vietnamese in order to facilitate their own efforts to learn Vietnamese; Romanized Vietnamese was not more widely promoted until the nineteenth century when the French colonial government set up primary schools in the south, in part to undermine the influence of traditional Chinese culture. In this case it was the French colonial government which not only flouted traditional Vietnamese written language conventions but also actively sought to destroy them, making use of a missionary-designed system. The second relevant attribute of the Protestants who promoted Romanized Mǐn was precisely that they were missionaries; in other words, they were trying to promote a new belief system, and to do so in a society that was not very receptive to their agenda. The challenge they faced had several key components. First, few people were willing to listen to their message, and of those few most were poorly educated or completely illiterate. Second, Protestants believed it was important for Christians to read the Bible for themselves, so literacy was very important. Third, the Chinese written language—Classical Chinese—was difficult to learn partly because of its writing system and because it differed so much from the spoken Chinese of the Mǐn region. These realities motivated missionaries to ignore Chinese conventions regarding the written language and adopt an approach that would have been unthinkable to most of their Chinese contemporaries. Put another way, the desire to produce a literate Christian community as quickly as possible outweighed any reservations they might have had about the relatively audacious strategy of creating and trying to promote a new written language in a society that already had its own written language. When we compare the case of Protestants and Romanized Mǐn to other instances of missionaries pioneering written vernaculars, evidence appears to support the generalization that the missionary nature of these groups was an important factor in explaining their pioneering role. Looking again at early Buddhist missionaries in China, it seems quite clear that they perceived use of the vernacular as a useful strategy in their desire to reach as many as possible with their message. In contrast, the case of Jesuits in Vietnam actually does not fit the pattern particularly well in that Jesuits did not employ Romanized Vietnamese as part of a strategy to reach the masses. Significantly, during this period Jesuit missionaries in China were taking exactly the opposite approach, conforming to local language conventions and mastering Classical Chinese and its literary tradition as part of a strategy to win a hearing among China’s elite. However, another “missionary” example does tend to support the generalization. In the Dialect Literature Movement in Hong Kong from 1947 to 1949, Communist Party members and their allies advocated the use of written Cantonese, Hakka, and Chaoshanese as a strategy to arouse the masses to drive the Nationalist Party from power. They also experimented to an unusual degree with writing texts in a style that adhered closely to the

76  Don Snow spoken form of these vernaculars, and in using regional vernaculars in new genres (Lo 1987; Snow 2004). While the activists of the Dialect Literature Movement were not religious missionaries—nor were they foreign—they were certainly missionaries in a broader sense. Based on the cases mentioned earlier, the pattern which emerges is not that missionary groups necessarily promote use of written vernaculars. Rather, the pattern is that in situations where missionaries are attempting to reach the masses, and the dominant written language is difficult to learn because it is very different from the local spoken language, missionaries tend to pioneer vernacular writing styles because they feel this makes it easier to reach the masses.

Conclusion Clearly missionaries are still often in the forefront of creating vernacularbased written languages for Bible translation purposes. However, most such projects involve spoken languages that have no written form—as was true of past cases like Tahiti and Hawai’i. Is it likely that we will again see something like the Romanized Mǐn project, where missionaries attempt to promote a new vernacular-based written language in a society that has an established written language? For a number of reasons this is less likely than in the past. First, fewer societies now are diglossic in the traditional sense. Today nations increasingly promote literacy through schools and the media—and teach modern written languages rather than ancient classical languages that have few similarities to modern speech. For example, while in China there are still many areas where the local variety of Chinese differs considerably from Mandarin (Pǔtōnghuà), most young people in these areas learn to speak and read Mandarin in school. Whereas in the past a regional vernacular may have been easier for the illiterate to learn, today most people are already literate in Mandarin and learning to read a regional variety of Chinese would be an extra effort. Also, around the world smaller languages are disappearing and being replaced by larger national languages. For example, in places like Suzhou where the local vernacular was once quite prestigious, many young people only speak Mandarin. Here, even though a written form of Suzhounese exists, it is not widely used (Snow, Zhou, and Shen 2018); the only place in the sinophone world where the written form of a regional Chinese vernacular is widely used is Hong Kong (Snow 2004; Li and Tong, this volume). We also need to remember, however, that in many places there is a considerable gap between the language taught in schools and the vernaculars that are still spoken daily at home; examples in China would include Cantoneseand Shanghainese-speaking areas. While young people in these areas can read and write in Mandarin, and written vernaculars are not needed for reducing illiteracy, the local spoken language also retains a considerable

Romanizing Southern Mǐn  77 degree of vitality (Snow 2004; Snow, Shen, and Zhou 2018). In these areas written vernaculars play a useful role because reading and writing as one speaks not only has a communication function but also an identity function. We should also remember that “missionaries” are not confined to the religious type; there are many kinds of organized groups that seek to win converts and promote agendas—whether these agendas be religious, ideological, or even commercial. As long as there is a significant gap between a society’s spoken language and the official language taught through schools and the media, we are likely to see missionaries of various types experimenting with writing styles closer to the way people normally speak—though perhaps less because these are easier to learn then because they “reach the heart” in ways that a language only learned in school does not.

Notes

References Band, Edward. 1936. Barclay of Formosa. Tokyo: Christian Literature Society. ———. 1972. Working His Purpose Out: The History of the English Presbyterian Mission, 1847–1947. Reprint of 1947 edition. Taipei: Ch’eng Wen Publishing. Burdon, John Shaw. 1890. “Colloquial Versions of the Chinese Scriptures.” In Records of the General Conference of the Protestant Missionaries of China Held at Shanghai, May 7–20, 1890, 98–105. Shanghai: American Presbyterian Mission Press. Cài Jǐntú 蔡锦图. 2012. “Zài yuánshǐ jiù yǒu dào: Shàntóuhuà shèngjīng de fānyì hé liúchuán” 在元始就有道：汕头话《圣经》的翻译和流传 [In the beginning was the word: The translation and dissemination of the Bible in the Chaoshan colloquial]. In Cháoshàn shèhuì yǔ jīdūjiào shǐlùn, edited by Xíng Fúzēng and Lǐ Línghàn, 136–56. Shantou: Shàntóu Dàxué chūbǎnshè. Campbell, William. 1996. Sketches from Formosa. Originally published in 1915. Taipei: SMC Publishing. DeFrancis, John. 1977. Colonialism and Language Policy in Vietnam. The Hague: Mouton Publishers. De Jong, Gerald. 1992. The Reformed Church in China, 1842–1951. Grand Rapids, MI: Eerdmans. De Velder, Walter. n.d. “A Missionary Journey over Nine Decades.” Unpublished memoir. Held at the Joint Archives of Holland at Hope College in Holland, Mich.

78  Don Snow Fagg, John. 1894. Forty Years in South China: The Life of Rev. John van Nest Talmage, D.D. New York: Reformed Church in America. Gibson, John Campbell. 1888. Learning to Read in South China: Being a Plea for the Use of Romanized Vernacular in Mission Work. London: Hazell, Watson, and Viney. ———. 1890. “Review of the Various Colloquial Versions and the Comparative Advantages of Roman Letters and Chinese Characters.” In Records of the General Conference of the Protestant Missionaries of China Held at Shanghai, May 7–20, 1890, 62–89. Shanghai: American Presbyterian Mission Press. Giedt, Emanuel. 1946. “Early Mission History of the Swatow Region Brought Down to the Present for the American Baptist Mission.” Unpublished manuscript. Held at Yale Divinity School. Hanan, Patrick. 1981. The Chinese Vernacular Story. Cambridge, MA: Harvard University Press. doi:10.4159/harvard.9780674418462 Heylen, Ann. 2012. Japanese Models, Chinese Culture and the Dilemma of Taiwanese Language Reform. Wiesbaden: Harrassowitz. doi:10.2307/j.ctvbqs7ng Holkeboer, Tena. 1921. Unpublished Correspondence. Held at the Joint Archives of Holland at Hope College in Holland, Michigan. Hood, George. 1986. Mission Accomplished? The English Presbyterian Mission in Lingtung, South China. Frankfurt: Peter Lang. Ion, Hamish. 1993. The Cross and the Rising Sun, Vol 2: The British Protestant Missionary Movement in Japan, Korea, and Taiwan, 1865–1945. Waterloo, ON: Wilfrid Laurier University Press. Kerr, George. 1974. Formosa: Licensed Revolution and the Home Rule Movement, 1895–1945. Honolulu: University Press of Hawaii. doi:10.2307/j.ctv9zckv6 Klöter, Henning. 2005. Written Taiwanese. Wiesbaden: Harrassowitz. ———. 2011. The Language of the Sangleys: A Chinese Vernacular in Missionary Sources of the Seventeenth Century. Leiden: Brill. doi:10.1163/9789004195929 Le, Ming-Hang, and Stephen S. O’Harrow. 2007. “Vietnam.” In Language and National Identity in Asia, edited by Andrew Simpson, 415–41. Oxford: Oxford University Press. Lin, Christine. 1999. “The Presbyterian Church in Taiwan and the Advocacy of Local Autonomy.” Sino-Platonic Papers 92: 1–136. Lo Wai-luen 盧瑋鑾. 1987. Xiānggǎng wénzòng 香港文縱 [Hong Kong writers from the north]. Hong Kong: Huáhàn wénhuà shìyè gōngsī. Mair, Victor. 1994. “Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages.” Journal of Asian Studies 53, no. 3: 707–51. doi:10.2307/2059728 Mak, George Kam Wah. 2017. Protestant Bible Translation and Mandarin as the National Language of China. Leiden: Brill. doi:10.1163/9789004316300 Pitcher, Philip. 1893. Fifty Years in Amoy: Or, a History of the Amoy Mission, China. New York: Reformed Church in America. ———. 1912. In and About Amoy. 2nd ed. Shanghai: Methodist Publishing House. Snow, Don. 2004. Cantonese as Written Language: The Growth of a Written Chinese Vernacular. Hong Kong: Hong Kong University Press. doi:10.1075/japc.16.2.13bou ———. 2012. “Revisiting Ferguson’s Defining Cases of Diglossia.” Journal of Multilingual and Multicultural Development 34, no. 1: 61–76. doi:10.1080/01434632.20 12.699531

Romanizing Southern Mǐn  79 ———. 2013. “Towards a Theory of Vernacularization: Insights from Written Chinese Vernaculars.” Journal of Multilingual and Multicultural Development 34, no. 6: 61–76. doi:10.1080/01434632.2013.786082 ——— and Chen Nuanling. 2015. “Missionaries and Written Chaoshanese.” Global Chinese 1, no. 1: 5–26. doi:10.1515/glochi-2015-1001 ———, Shen Senyao, and Zhou Xiayun. 2018. “A Short History of Written Wu, Part II: Written Shanghainese.” Global Chinese 4, no. 2: 217–46. doi:10.1515/ glochi-2018-0011 ———, Zhou Xiayun, and Shen Senyao. 2018. “A Short History of Written Wu, Part I: Written Suzhounese.” Global Chinese 4, no. 1: 143–66. doi:10.1515/ glochi-2018-0007 Ts’ai, Hui-yu Caroline. 2009. Taiwan in Japan’s Empire Building: An Institutional Approach to Colonial Engineering. London: Routledge. doi:10.1163/15685314-12341276 Tsurumi, Patricia. 1977. Japanese Colonial Education in Taiwan, 1895–1945. Cambridge, MA: Harvard University Press. doi:10.4159/harvard.9780674434080 “Vernacular Translations of the Bible.” 1901. The Chinese Recorder and Missionary Journal 32, no. 11: 563–64. White, Chris. 2017. Sacred Web: The Social Lives and Networks of Minnan Protestants, 1840s–1920s. Leiden: Brill. doi:10.3366/swc.2018.0219 Wong, Simon. 2015. “Bible Translation of Non-Mandarin Han Fangyan (Dialects) in Mainland China: The Case of Swatow.” The Bible Translator 66, no. 1: 24–44. doi:10.1177/2051677015569712 Woodin, Simeon F. 1890. “Review of the Various Colloquial Versions and the Comparative Advantages of Roman Letters and Chinese Characters.” In Records of the General Conference of the Protestant Missionaries of China Held at Shanghai, May 7–20, 1890, 89–98. Shanghai: American Presbyterian Mission Press. Xǔ Cháng’ān 许长安, and Lǐ Xītài 李熙泰. 1999. Xiàmén huàwén 厦门话文 [Written Amoy]. Xiamen: Lùjiāng chūbǎnshè. Yóu Rǔjié 游汝杰. 2003. Shèngjīng fāngyán yìběn kǎoshù 圣经方言译本考述 [A study of dialect translations of the Bible]. In Zhùmíng zhōngnián yǔyánxuéjiā zìxuǎn, edited by Yóu Rǔjié, 265–94. Hefei: Ānhuī jiàoyù chūbǎnshè.

4

Interactions across Englishes in Mainland China, Hong Kong, Macao, and Singapore Christiane Meierkord

Abstract This chapter provides an historical overview of the role of English as the major international language of the sinophone world. Taking a comparative look at China, Hong Kong, Macao, and Singapore, it examines the historical development of varieties of English and past and present-day interactions that are conducted using English across these different varieties. It is shown that throughout the sinophone world, the ubiquitous presence of English has made it the main contact variety and contributor to emerging creole languages. Within the sinophone world, Singapore stands out as the only place with a new trans-ethnic English variety. In other areas, English has not had a sufficiently strong status for new varieties to stabilize.

Introduction: The Multilingual Ecologies of the Sinophone World The sinophone world is home to a vast number of languages belonging to diverse language families, and the linguistic ecologies of those nations which belong to the sinophone world have long been characterized by the co-existence of languages that are not mutually intelligible. However, through the trading activities of the various seafaring powers, further languages, including Arabic, Germanic, and Romance languages—i.e. Portuguese, Spanish, Dutch, and eventually English—were introduced to the area from the sixteenth century onward. Today, it is particularly English which has come to stay as a significant addition to the linguistic ecologies of the various countries in the area. More recently, migration and colonization have added to the picture and have led to the introduction of yet further languages, e.g. from the African continent. In many cases, Chinese languages and dialects, Austronesian and Tibetan languages, European languages, Malay, and languages originating on the Indian subcontinent have contributed to the multilingual character of individual speech communities and countries. To this multilingual ecology, English has contributed in various ways: either as the first language of British settlers (as in the case of Hong Kong)

Interactions across Englishes  81 or as a second language (in the case of Singapore) or as a foreign language (mainland China, Macao). This has led to interactions between speakers of different varieties of English, who often use English in Interactions across Englishes (IaEs) (Meierkord 2012). For example, in Singapore, speakers of Malay, Tamil, and Chinese interact in English; in Hong Kong speakers of Cantonese and English interact in English, and in China speakers of various Chinese languages communicate with an international workforce in English. This chapter takes a look at the history of English in these four places,1 the historical development of varieties of English in the area, and past and present-day interactions that are conducted using English across these different varieties. I shall outline how these interactions compare throughout the sinophone world and describe who uses English as a lingua franca with whom and for what purpose and how this is (dis)similar in the individual nations. In addition, I shall discuss how multilingual language planning and policy in the various areas influence the use of English and to what extent the use of English as a lingua franca has resulted in changes in the varieties of English in the area.

A Short History of English(es) in the Sinophone World British Presence in the History of the Sinophone World While there had been earlier sporadic visits to the sinophone world, regular and prolonged visits by Europeans took place from the sixteenth century onward. At that time, the island of Singapore had already served as a trading port for Arabs, Chinese, Malays, and traders from the Indonesian islands and other parts of Southeast Asia (Chew 2013). The Portuguese were the first to successfully negotiate territorial rights at Macao in 1557 (Wade 2001). Bolton (2003: 125) assumes that “a small number of British travelers may have visited Macao in the period 1580–1620.” However, the nature of British involvement differs considerably between the four areas discussed in this chapter. The first British traders came to China in 1637, establishing trading posts in the Pearl River delta in 1672 (ten years after the Dutch East India Company had been expelled from Taiwan by the Chinese; cf. Graddol 2013) and a “factory” (harbor-side business premises and residences) in 1685 in Guangzhou (Canton). The British East India Company’s ships arrived from India and the Company was initially allowed to trade in Xiamen (Amoy), Zhoushan Island, and Guangzhou, but from 1757 onward the Chinese government imposed severe restrictions on foreign traders, and their activities were confined to the port and city of Guangzhou. There, they would live for the duration of the trading season each year, from August to March. At the end of the season, the westerners would either return to Europe or the United States or move to Macao at the mouth of the Pearl River. (Bolton 2003: 147)

82  Christiane Meierkord where their families lived. In fact, until the establishment of Hong Kong as a British possession in 1842 “Macau was the primary ‘centre’ of English in China” (Moody 2019: 82). In the eighteenth century, the Americans joined the trading community together with the French and others. Only 13 trading companies (hongs) were allowed at any one time, and all trading had to be pursued via cohongs, Chinese merchants franchised by the government. Direct trade with the Chinese was forbidden. The British initially traded British wool and Indian cotton for Chinese tea, porcelain, and silk. Eventually, however, a trade imbalance between China and Britain resulted in a shortage of silver to pay for tea, and the British discovered opium as an export good. Trade was restricted to this area until 1842, following China’s defeat in the Opium War (1839–42), when other trading ports were opened to Westerners and contact between them and the local population was allowed in Shanghai and Xiamen, Fuzhou, and Ningbo. Earlier, in 1819, Sir Stamford Raffles (1781–1826) had leased the island of Singapore from the Sultan of Johore, strengthening British influence in the area over that of the French and Dutch. The island then came under the control of the East India Company, and later, in 1867, became an official British colony. During World War II, between 1942 and 1945, Singapore was under Japanese rule, but then became British again. It gained independence in 1965. Colonization involved British administration of the area and the use of English for this purpose. However, English was used by the colonial government and then the independent government primarily for administrative purposes. Singaporeans in those days communicated across ethnic lines largely in Bazaar Malay, a form of pidginised Malay, or in a simplified form of Hokkien. (Bolton and Ng 2014: 309)2 Hong Kong was ceded to Britain as a result of the First Opium War, in 1842. Here, English largely spread through the education system. Between 1851 and 1876, a number of mission schools were established, in which Chinese language and literature was taught alongside English, thus creating a category of school referred to as ‘Anglo-Chinese’, a term which survives until today, referring to schools where English is the declared teaching medium and the printed medium for most textbooks. (Bolton 2000a: 267) The 1860s witnessed the establishment of the first mission schools not only in Hong Kong but also in Guangzhou, Xiamen, Shanghai, and Ningbo. This means that English was available in a British standard form, taught by Protestant educators. English was also the medium of instruction (MOI) at the University of Hong Kong, which was established in 1911 as the first university in the

Interactions across Englishes  83 territory. It was not until 1963 that teaching in Chinese was established here. In Macao, the situation was different due to the fact that it served as a summer residence for those Europeans who were trading in Guangzhou. “The history of the British and English language in Macau was recorded as early as 1637 through the arrival of English traders in the region at the time, although English travelers may have visited Macau earlier than this” (Botha and Moody 2020). In the eighteenth century, when companies from various European countries and from America were pursuing trade in the area, an Anglo-American community developed (Da Silva 1998). The genteel, decaying Portuguese colony became the Ascot, the Monte Carlo, the Riviera, and even the home of tired traders after a busy tea season at the factories. At Macao dwelt the wives and families, as did the mistresses and the occasional ladies who brightened the hours of relaxation between the strenuous sessions at Canton. (Downs 1997: 49)

Englishes in the Histories of China, Hong Kong, Macao, and Singapore As indicated earlier, English came to the area through traders and sailors, and later through administrative, missionary, and military personnel. Most of these were British, but some were American. During the initial period of contact, it was illegal for foreigners to learn Chinese, and at the same time, of course, English was not taught to the Chinese so that, as a result, interpreters were employed. However, due to the early presence of the Portuguese in the area, a variety of Portuguese had already become the lingua franca in the area (Indo-Portuguese; cf. Yule and Burnell 1886). This variety of Portuguese also contained borrowings from the various languages the Portuguese had been in contact with, e.g. Arabic, Cantonese, and Malay-Javanese, and many of these were subsequently integrated into what is referred to as AngloIndian. Some of these lexical items eventually became part of the Standard English word stock; others such as chop for ‘seal, stamp’ are confined to the area. Besides, Chinese Pidgin English most likely developed between English-speaking foreign traders and the Chinese merchants of the Co-hong and their staff; second, to local employees including the comprador and his subordinates; third, to personal servants within the factory; fourth, with local shopkeepers and tradesmen; and fifth, as a means of communicating with the Tanka girls aboard the ‘flower boats’ on the Pearl River. (Bolton 2003: 157) As a result, because neither the British nor the Chinese aimed at learning each other’s language, Chinese Pidgin English (Hall 1944), also earlier

84  Christiane Meierkord referred to as China Coast English, developed between the 1720s and 1830s in Guangzhou (Bolton 2003: 156; Hall 1944). Its existence is documented in the accounts of sailors, merchants, and, later, missionaries (cf. Bolton 2000b). Pidgin was apparently spoken in two variants, which are recorded as phrases and dialogues in memoirs and travelogues written by Westerners as well as in phrasebooks. The two varieties spoken by Chinese on the one hand and by English and other Europeans on the other hand seem to differ systematically (cf. Ansaldo, Matthews, and Smith 2011). The variety used by Chinese speakers is represented in the phrasebooks, while the European lect is represented in the memoirs and similar Western texts. The examples below, taken from Matthews and Li (2013), illustrate the Chinese (example 1) and the European (examples 2 and 3) lect: 1 希合吿簡當 2 She now no can go to Whampoa as before time. 3 Bum bye you kum my housy second teem. In Macao, the situation was more complex, with the population consisting of a hybrid mix of Portuguese, other Europeans, Asians, and Chinese. Bolton (2003: 158) assumes that by the end of the eighteenth century, Portuguese creole would have been losing ground to English and that a pidginized variety would have been used by Europeans with Chinese traders and merchants, with local shopkeepers and tradesmen, and with domestic staff. These accounts and the fact that pidgin assumed two distinct lects indicate that English has always been used in different forms or varieties in the area, even in this early situation. As a result, speakers of these different varieties, stable or not, have interacted with each other in what I call Interactions across Englishes (IaEs, Meierkord 2012).

Interactions across Englishes The core assumption of this model is that different Englishes potentially merge in interactions that involve speakers of different first languages, and that this, also potentially, results in the development of new emergent varieties. Given the heterogeneity of the functions which English performs as a

Interactions across Englishes  85 lingua franca and the diversity of the possible constellations of such interactions in terms of the speakers and their different Englishes, there will not be one stable or even codified variety, but rather a heterogeneous array of new linguistic systems. Figure 4.1 illustrates how speakers contribute to these interactions. Individuals engaging in IaEs are speakers of different varieties of English, which may be any of the full range of the Englishes, i.e. first language Englishes (e.g. spoken by many Singaporeans), second language Englishes (e.g. spoken by many Hong Kong Chinese), immigrant Englishes (such as those of the Filipina domestic workers in Hong Kong), foreign-language Englishes (spoken by most mainland Chinese). Figure 4.1 illustrates the processes for speakers of just three different varieties. Obviously, there may be more or less varieties involved in any given instance of IaEs. The various oval shapes represent these varieties. The blocks in the small arrows represent the different features which are associated with the different Englishes and which are potentially brought into the interaction, as the arrows symbolize. The rectangular shape in the middle is a potential “feature pool” (Mufwene 2001), which results from the participants using the features of their various varieties in the interaction(s) and thus contributing them to the “feature pool.” The various features in that pool are then available to all speakers to use, i.e. a Malay speaker of English in Singapore can “decide” (most likely, this is not really a conscious decision) to pick up a feature that is characteristic of the English of a Chinese speaker in Singapore.

Figure 4.1 Interactions across Englishes. Source: Meierkord (2012: 62), reprinted with permission by Cambridge University Press.

86  Christiane Meierkord It is in fact in Singapore that the regular interaction between speakers of various first languages in English has resulted in the development of a pan-Singaporean variety of English (see below). While different varieties of English also exist in Hong Kong, there is no account of a pan-Hong Kong variety emerging, potentially based on the English spoken by Chinese of various regional, ethnic, and national backgrounds, the Filipinas, Indonesians, and L1 speakers from the United Kingdom, America, and Australia. As the explanations for Chinese Pidgin English illustrate, IaEs existed even in these early times. Obviously, speakers with different first languages had different Englishes and different varieties of the pidgin existed.

Interactions across Englishes in the Sinophone World Today Today, uses of Englishes in the sinophone world are of three types: as a first language, as a second language, and as a foreign language. Second language varieties differ from foreign-language varieties in that they are used for intra-national communication, particularly in administration, politics, as a MOI, and in the media. Kachru (1985) modeled these different uses of English in his three circles model, which places countries into three different circles: an inner circle in which English is the first language for the majority of citizens (e.g. the United Kingdom, the United States, or Australia), an outer circle in which English has become an important language for intra-national communication in administration, education, politics, and the media and where English is used as a second language by large parts of the population (e.g. India, Nigeria, Singapore, or Hong Kong), and an expanding circle in which English is mainly used for international communication and remains a foreign language (e.g. China and Macao). This section describes the different uses of English in Singapore, Hong Kong, Macao, and mainland China, pointing out how IaEs are evidently taking place. It finishes with a brief description of the international use of English in the area. Intra-national Interactions across Englishes English(es) in Singapore In Singapore, English is an official language, alongside Tamil, Mandarin, and Malay (for details, see Goh and Fong, this volume). It is the language of the law courts, government administration, and education (McArthur 2002: 338). Since 1987, it has been the only MOI in Singapore’s primary and secondary schools and at universities (ibid., 339). According to Bolton and Ng (2014), a formal education policy was introduced in the late 1950s in Singapore with emphasis on the four official languages. In its original form, the language policy stated that the four official languages were also the media of instruction. In the time following independence in 1965, while this was

Interactions across Englishes  87 the case in most schools, there were also a number of Tamil, Malay, and Mandarin-medium schools. However, by 1987 all of these were closed because of falling student numbers (Tan 2007). This change reduced Mandarin, Tamil, and Malay to being taught as second languages in primary and secondary schools, and English has since dominated the country’s education system (Tan 2012; Tan and Goh 2011). Unlike what is the case in Hong Kong (see below), however, English is spoken as a first language by a considerable number of Singaporeans. Based on the 2010 census data, Lewis et al. (2015) estimate 1,100,000 first language and another two million second language speakers of English in Singapore. Similarly, Tan (2014: 319) notes that the “latest 2010 census however reveals that over 30 per cent of Singaporeans report English to be the primary language used in the home, an increase from about 20 per cent in 2000.” A detailed, fine-grained, analysis of this development is available in Buschfeld (2019), who also looks at cross-ethnic differences. As Tay (1991: 322) explains: “[o]f all varieties of English spoken in Southeast Asia, Singaporean English performs the widest range of uses” and mentions six characteristic uses of English that have been described in the literature: “(1) official language, (2) language of education, (3) working language, (4) language of interand intra-ethnic communication, (5) language for the expression of national identity, (6) international language.” Two more functions, i.e. (7) language of religion and (8) home language, had been identified by Bloom (1986: 388). It is significant that in Singapore English is used not only in the public domains of transactions, employment, education, media, government, law, and religion, but also in the more private domains of family and friendship (Platt and Weber 1980). However, focusing on a “young and highly educated segment of the Malay population” (2014: 376), Chong and Seilhamer (2014) document that ethnic Malay proudly maintain Malay as their first language and as a language of identity, while English “best defines them as Singaporeans” (2014: 363). English is, reportedly, predominantly used by the participants in the everyday activities of thinking/speaking to oneself, counting, diary-writing, watching television, reading books/magazines and reading newspapers, whereas Malay was found to be predominantly used in the everyday religious domain activity of praying and was just slightly more likely than English to be used when telling short stories to fellow Malay friends. (Chong and Seilhamer 2014: 369) Furthermore, Malay was overwhelmingly used with grandparents, parents, and uncles or aunts. Singapore English, as an indigenized variety, developed through Englishmedium education (Platt 1991: 376), having initially been taught by the British to an elite part of the local population, who had either one of the

88  Christiane Meierkord Chinese dialects, and Indian language, or Malay as their home language, from approximately the 1830s. Eventually, English has “also become the lingua franca for inter-ethnic communication, especially among the younger and more educated, particularly in more formal settings, effectively replacing the ubiquitous Bazaar Malay and Hokkien which performed the service in previous generations” (Lim and Foley 2004: 5–6). Present-day Singapore English has both a standard variety, which is taught in schools, and a colloquial one, “which has a distinctive phonology, syntax and lexicon which shows a high degree of influence from other local languages such as Hokkien, Cantonese, Malay and Tamil” (Lim and Foley 2004: 7), and Singapore English varies at the lectal continuum between these two varieties. As a result of the diverse ethnic backgrounds of Singapore English speakers, “features particular to certain ethnic groups are also observable” as well as “pan-Singaporean features which all Singaporeans do share” (Lim 2004: 19). The fact that Singapore English had become increasingly similar across ethnicities had been pointed out earlier by Platt and Weber (1980: 46). At the level of phonology, the pan-Singaporean variety seems to be mainly influenced by Hokkien and Malay. Tamil, Cantonese, and Mandarin, all spoken by fewer people, are not reflected. For example, “egg rhymes with vague but not with peg for nearly everyone, regardless of their ethnic background” (Deterding 2007: 5). Chinese languages influence pitch patterns of Singapore English and interestingly “exhibit a maintenance of their tone as in their language of origin …, even though they are being used in a variety of English” (Lim 2004: 48; see also Wasserfall, this volume). As regards grammar, differences as regards the use of the be copula have been observed across speakers with Chinese dialects, Malay, and Tamil as their first language (Tay 1991: 378). On the other hand, the use of will to indicate regular events and would to indicate tentativeness is again a pan-Singaporean feature. At the level of the lexicon, borrowings into pan-Singaporean English originate from various local languages, and include the Hokkien item kiase (‘afraid to lose out’) and the Malay word makan (‘eat’) (cf. Deterding 2007: 5). “As the ethnically Chinese majority constitutes 76 per cent of the population, the influence of Chinese dialects, especially of the dominant dialect Hokkien, is the main one” (Platt 1991: 377). However, this influence was not always a direct one, originating from the speakers’ first languages. In many instances, the influence came through Baba Malay, a creole spoken by the Peranakan or “Straits-born” Chinese. “This partly explains why Malay lexical items such as makan ‘eat, food’ became a part of colloquial Singapore English along with Chinese influenced morpho-syntactic structure and semantics as in verbs of direction” (Platt 1991: 377). We can therefore not speak of the result of IaEs in these “Malay-mediated” long-term developments. In fact, Ansaldo explains the pan-Singaporean features as follows: [T]he existence of an early English variety pre-dating SE, probably modelled on the Peranakan’s English, is plausible from a careful

Interactions across Englishes  89 interpretation of the historical and literary documents of the early years of Singapore. This also explains why it is ultimately irrelevant to establish whether the substrate origins are to be found in Hokkien or Malay: they are more likely to be found in Bazaar Malay, i.e. typological convergence had already taken place before the formation of SE, which explains even better the hybrid structures observed so far. (Ansaldo 2004: 143) Finally, geographical factors play an important role in fostering IaEs and a pan-Singaporean variety: Singapore’s densely populated geographical space involves neighborhoods in which ethnicities live in close contact, which is also fostered by the government (Deterding 2007: 5). English(es) in Hong Kong Today, Chinese and English are the two official languages of Hong Kong. However, the fact that the Hong Kong area has traditionally been Cantonesespeaking, whereas the central government of China uses Pǔtōnghuà, has resulted in a language policy that favors trilingualism (see also Li and Tong, this volume). Primary education is largely in Cantonese. Secondary education is either English or Cantonese medium, with Mandarin being taught as well. Two of the city’s newspapers are in English. However, English is “prominent in the media and advertising, omnipresent visually alongside Chinese, and crucial as the lingua franca of business, but is not widely used as an everyday spoken medium” (McArthur 2002: 359). In fact, English seems to be yielding its status as the dominant second language to Pǔtōnghuà, despite the steady growth of competence that Graddol (2013) attests. The number of Hong Kong residents who, in the latest census (Census and Statistics Department 2017), reported that they can speak Pǔtōnghuà stands at 48 percent. In the 2001 Hong Kong census, only a third of Hong Kong’s residents had reported the ability to speak Pǔtōnghuà, and this recent rise is probably due to the fact that China has been promoting Pǔtōnghuà, particularly in Hong Kong’s schools, since 1997. At the same time, 46 percent reported to be able to speak English. However, English is spoken as a first language by only 3.5 percent of Hong Kong’s population, while 89.5 percent speak Cantonese. Interestingly, the majority of individuals who reported using English as their usual3 language are Filipinas (and Filipinos) rather than citizens originating from countries associated with English as a first language, such as Great Britain, the United States, or Australia.4 As Table 4.1 illustrates, English comes in many different varieties. For the majority of Hong Kong’s population, however, it remains a foreign language. The linguistic impact of either of these varieties is negligible, given their small size. The status of English in Hong Kong is highly disputed. Tay (1991) describes it as being a learner variety, due to the fact that it is not used for informal purposes. As such, uses of English in Hong Kong seem to differ from

90  Christiane Meierkord Table 4.1 First language and nationality in Hong Kong

Chinese Place of domicile— Hong Kong Place of domicile— other than Hong Kong Indonesian Filipino British Indian Pakistani American Australian Nepalese Thai Japanese Others Total

Cant.

PTH

Other Ch.

Engl.

5,877,564

70,956

260,095

30,469

64,079

17,242

11,978

1,853

97,105 6,081 6,408 1,654 780 6,171 4,647 358 10,830 732 18,804 6,095,213

3,659 225 72 – 7 275 69 – 47 135 1,712 94,399

340 28 215 – – 125 135 – 357 – 472 273,745

19,987 111,227 22,145 8,189 1,305 8,743 9,204 866 773 2,111 21,416 238,288

Others

Total

5,195 6,244,279 93

95,245

16,115 137,206 16,571 134,132 2,282 31,122 15,088 24,931 13,070 15,162 227 15,541 145 14,200 13,509 14,733 2,079 14,086 9,851 12,829 12,563 54,967 106,788 6,808,433

Source: Census and Statistics Department (2017). Cant. = Cantonese; PTH = Pǔtōnghuà; Ch. = Chinese; Engl. = English.

those in other parts of Asia, and Tay points out that “Hong Kong, consisting of the island of Hong Kong itself, Kowloon and the New Territories, fits less easily into the region” (1991: 319). This is due to the fact that linguists as well as educators traditionally held that Hong Kong English adhered to an exonormative, British standard, and that it “was never ‘nativized’” (Bolton 2000a: 263). Bolton finds that Hong Kong differs from other Asian societies due to the “relative longevity of the colonial era in the ‘territory’” (2000b: 265). However, he reports stable pronunciation as well as lexical features which to him do indicate nativization. At the same time, English in Hong Kong is not a homogeneous variety. It is spoken by a number of British expatriates, but predominantly by Chinese with Cantonese as their first language. As Bolton (2003: 204) points out, Hong Kong differs considerably from Singapore in that “in Singapore English is used as an interethnic lingua franca between the majority Chinese population and sizeable minorities of Singaporean Malays and South Asians”— conditions that do not exist in Hong Kong. However, “significant numbers of Indians, Parsees, and Eurasians still live in the territory, as do Indonesians, Filipinos, Japanese, Malaysians, Nepalis, Pakistanis, and Thais” (Bolton 2000b: 276).5 As a result, many different Englishes add to Hong Kong’s linguistic ecology, so that IaEs abound in the city, but precise descriptions of these diasporic varieties and of interactions across them do not yet exist.

Interactions across Englishes  91 English(es) in Macao Macao is home to various ethnic groups, “comprising Chinese (who are themselves of various ethnic groups), Portuguese, and other smaller communities of people, mostly from Southeast Asia” (Botha and Moody 2020).6 Besides, there exists “a group of people who are of mixed ancestry, through the intermarriage of Portuguese, Chinese, Malay and Indians, among others” (ibid.). The colonial heritage of Macao is reflected in its linguistic ecology today. It is home to five languages, according to Lewis et al. (2015). As is the case in Hong Kong, the vast majority of Macao people speak Cantonese as their first language. In addition, Southern Mǐn and Mandarin speakers add to the linguistic ecology of the city as do approximately 2,800 speakers of Portuguese and a very small group of 50 speakers of the Portuguese-based creole Macanese (spoken mainly by those people of mixed ancestry; see also Li and Tong, this volume). English is used as an immigrant language by approximately 2,700 speakers (cf. Lewis et al. 2015, see also Yan and Moody 2010). For the majority of its users, however, English is a foreign or, at best, a second language. It does not have any legal status in the Macao Special Administrative Region, whose linguistic ecology is dominated by Cantonese, and whose two official languages are Chinese7 and Portuguese. However, “there is a widespread use of English within the government at various levels” (Moody 2008: 4), e.g. in legal texts, in government agencies, and as an additional working language of the government. In fact, 58.6 percent of civil servants speak English, compared to only 16.6 percent of the total resident population. English furthermore “maintains de facto official status within the Macau educational system” (Moody 2008: 8), which does not stipulate the MOI for fully private schools. However, only 13.4 percent of Macao students attend English-medium schools. Tertiary education largely takes place outside of Macao, frequently in English. Mandarin is also the language that dominates business and tourism. The latter relies largely on the gambling sector and attracts a huge number of tourists from mainland China who are often unable to use Cantonese or English. English is furthermore a MOI in the education sector, though it is forbidden in government schools, where the official languages Chinese and Portuguese are used for this purpose. At the same time, “private schools that offer free subsidized education may use English as their medium of instruction” (Moody 2019: 90), and in fact 12 subsidized schools did so, both at primary and secondary level, in the 2015/16 academic year (ibid.). English is also used as a MOI in tertiary education, though the extent to which this applies varies. Botha (2013) reports for Macao University of Science and Technology, a private university founded in 2000, and thus the newest institution of higher education, that there is “visibly higher exposure to English in the faculty of management and administration and the faculty of hospitality and tourism compared with the other faculties” (Botha 2013:

92  Christiane Meierkord 470). However, the largely mainland Chinese and Macao students consider English important for their future career, particularly to interact with Asian and foreign tourists and fellow workers from the Philippines or Thailand (Botha 2013: 472–73). Access to English is also via the media. However, it is not the dominant language in the media, with The Macau Post Daily, launched in 2004, being the only local newspaper published in English (Young 2006). While in Macao itself official media are in Chinese or Portuguese, the proximity to Hong Kong allows easy access to English-language media. Bolton (2003) claims that the second language Englishes spoken in Macao and Hong Kong are very similar in terms of their phonology and syntax. Moody (2008) however points out that there seem to be differences in the lexicon, with Macao English containing “a number of Portuguese-influenced loan-words that, although they are often included as HKE lexemes, may appear more frequently in Macau than they do in Hong Kong” (Moody 2008: 11–12). English in Mainland China While Singapore and Hong Kong share a history of British colonialism and a linguistic ecology that has been shaped, among other factors, by the British policy of making English the official language of these territories, the situation differs considerably in mainland China. As Bolton and Botha explain: In many respects, China is a classic Expanding Circle society, where English has long had the status of a foreign language, although the effects of globalisation, including increased migration and travel and access to the Internet and international media have also had an impact. Nevertheless, a fuller understanding of the status and functions of English in contemporary China also involves an understanding of the internal sociolinguistic dynamics of the nation. Indeed, at an official level, issues relating to the English language have typically been seen as far less important than other language policy issues, not least the promotion of Putonghua as a national language throughout a geographically vast and ethnically diverse society. (Bolton and Botha 2015a: 169) The recent history of English in China has gone through various stages. At the beginning of the twentieth century it was, together with German and Russian, appreciated as a language of philosophy, Western ideas, and intellectual revolution and was also used as a language of diplomacy. In fact, starting from the late nineteenth century, several Christian colleges and universities were established and became popular in the 1920s and 1930s among middle class parents. Several of these were English-medium establishments

Interactions across Englishes  93 (cf. Bolton and Botha 2015b: 191). Things changed considerably after World War II. From 1949, Russian was the first foreign language learnt in China, and English was used for science and technology only. However, following the deterioration of China-Russia relations in the late 1950s, English witnessed a first renaissance as a language of modernization and international understanding up until the Cultural Revolution (1966–76), during which English speakers were mistrusted. From 1982 onward, English has been strongly promoted in school curricula and held to be highly desirable (Kirkpatrick 2007: 145f.). According to Bolton (2003), the number of English teachers has increased from an estimated 850 in 1957 to well over half a million today. Nevertheless, proficiency in English is low. Wei and Su’s (2015) survey, conducted between 1998 and 2001, focused on the urban areas Beijing, Shanghai, Tianjin, Guangzhou, Shenzhen, Chongqing, and Dalian. It revealed that “overall, the frequency of the use of English fell between ‘seldom’ and ‘sometimes’, leaning toward the ‘seldom’ end” (Wei and Su 2015: 179). As they explain, this is due to the fact that English does not have official status in the country. In general, i.e. for the survey at large and at the national level, respondents’ self-ratings for reading proficiency “fell between ‘Able to recognize a few words’ and ‘Able to understand simple sentences’, leaning toward the ‘Able to understand simple sentences’ end” (ibid., 181). Only 3.26 percent of all respondents said they were able to read English books and periodicals fluently. Respondents in the seven cities scored “slightly above the ‘Able to understand simple reading passages’ level” (ibid., 181). Respondents’ self-rated spoken proficiency was generally lower, with the national average close to “Able to say some greetings.” In the cities self-ratings were higher but nowhere close to “Able to conduct daily conversations.” While Alvaro (2015) reports that there are a considerable number of English-language Chinese newspapers and websites, these are predominantly geared to an international audience with the foreignlanguage press having “performed a major role in China’s propaganda system” (2015: 269). Interactions across different varieties of English largely take place when they involve speakers from different national backgrounds, since communication between speakers of different Chinese dialects is typically conducted in Pǔtōnghuà. International Interactions across Englishes in Asia While the earlier sections have documented how English has been developing into an intra-national interethnic lingua franca in the sinophone world, it fulfills this role in competition with Mandarin. The role of English is, however, much more pronounced in interactions that result from the political and economic relations between the sinophone world and its Asian neighbors, particularly the ASEAN countries8 (of which Singapore is a member itself), and from recent migration. ASEAN’s sole official language is English. At the risk of gross oversimplification, the joint cultural heritage

94  Christiane Meierkord of the area seems to foster the development of common norms for interactions conducted in English, which differ from the conventions in traditional English-speaking countries. As Kirkpatrick explains, the major role of English in ASEAN is as a lingua franca and it is the sole working language of the group. While there are, of course, cultural differences between people from the ten different nations of ASEAN, they also share a number of pragmatic norms. For example, most of the cultures of ASEAN are much more comfortable with deflecting rather than accepting compliments. Most are more comfortable with prefacing a request with reasons for it, than making it up front. Most are more comfortable with allowing a speaker to finish a turn than interrupting it. (Kirkpatrick 2011: 220) Kirkpatrick has collected one million words of naturally occurring spoken interactions conducted in English as a lingua franca in The Asian Corpus of English (ACE). All speakers are “English-knowing” multilinguals from the ASEAN countries as well as mainland China, Hong Kong, Taiwan, Korea, and Japan. L1 Singaporean or Filipino speakers were not included. They were recorded when giving interviews, during press conferences, service encounters, seminar discussions, working group discussions, workshop discussions, meetings, panels, question-and-answer sessions, and during casual conversations. While these interactions are seemingly successful, this is not necessarily the case when interlocutors have very different cultural backgrounds and speak varieties of English that are very different from the ones learnt by Chinese in the classroom. Following the most recent migration trends, the sinophone world today is furthermore host to a large number of immigrants from very many different regional, ethnic, and linguistic backgrounds. West African immigrants are a case in point. Bodomo explains that Afro-Asian relations officially started in 1955 and that “African communities have become a reality in cities such as Shanghai and Hong Kong” (Bodomo 2010: 693). In his field study of Guangzhou, Bodomo finds that most Africans, the majority being Nigerians, work as businessmen. He explains that “English is the main lingua franca in the community for interaction with non-Africans, along with Chinese and French” (2010: 701). The low levels of English proficiency on the side of the Chinese, however, frequently result in communication problems, which are then resolved by using an interpreter or a calculator to type in prices for negotiation.

Conclusions English has added to the linguistic ecologies of the sinophone world since the late sixteenth century. Throughout its history in this region, English has

Interactions across Englishes  95 been used as Englishes, i.e. in various varieties, spoken by speakers of different first languages. Thus, the area has always been characterized by IaEs or the use of English as a lingua franca. In Singapore this has led to a new trans-ethnic English. In other areas, English does not have a sufficiently strong status as an L2 for new stable varieties to emerge. These areas are likely to show tendencies that have been found to characterize international IaEs such as heterogeneity, simplification, leveling, and accommodation (see Meierkord 2012: 158–94).

Notes

References Alvaro, Joseph James. 2015. “Analysing China’s English-Language Media.” World Englishes, 34, no. 2: 260–77. doi:10.1111/weng.12137 Ansaldo, Umberto. 2004. “The Evolution of Singapore English: Finding the Matrix.” In Singapore English: A Grammatical Description, edited by Lisa Lim, 127–49. Amsterdam: John Benjamins. doi:10.1075/veaw.g33.08ans ———, Stephen Matthews, and Geoff P. Smith. 2011. “The Cantonese Substrate in China Coast Pidgin.” In Creoles: Their Substrates and Language Typology, edited by Claire Lefebvre, 289–301. Amsterdam: John Benjamins. doi:10.1075/ tsl.95.17ans Bloom, David. 1986. “The English Language and Singapore: A Critical Survey.” In Singapore Studies: Critical Surveys of the Humanities and Social Sciences, edited by Basant K. Kapur, 337–458. Singapore: Singapore University Press. Bodomo, Adams. 2010. “The African Trading Community in Guangzhou: An Emerging Bridge for Africa-China Relations.” The China Quarterly 203: 693–707. doi:10.1017/S0305741010000664

96  Christiane Meierkord Bolton, Kingsley. 2000a. “The Sociolinguistics of Hong Kong and the Space for Hong Kong English.” World Englishes 19, no. 3: 265–85. doi:10.1111/1467-971X.00179 ——— 2000b. “Language and Hybridization: Pidgin Tales from the China Coast.” Interventions: International Journal of Postcolonial Studies 2, no 1: 35–52. doi:10.1080/136980100360788 ———. 2003. Chinese Englishes: A Sociolinguistic History. Cambridge: Cambridge University Press. ———, and Werner Botha. 2015a. “Researching English in Contemporary China.” World Englishes 34, no. 2: 169–74. doi:10.1111/weng.12131 ———, and Werner Botha. 2015b. “English in China’s Universities: Past and Present.” World Englishes 34, no. 2: 190–210. doi:10.1111/weng.12133 ———, and Bee Chin Ng. 2014. “The Dynamics of Multilingualism in Contemporary Singapore.” World Englishes 33, no. 3: 307–18. doi:10.1111/weng.12092 Botha, Werner. 2013. “English-Medium Instruction at a University in Macau: Policy and Realities.” World Englishes 32, no. 4: 461–75. doi:10.1111/weng.12053 ———, and Andrew Moody. 2020. “English in Macau.” In The Handbook of Asian Englishes, edited by Kingsley Bolton, Werner Botha, and Andy Kirkpatrick. Oxford: Wiley. Buschfeld, Sarah. 2019. Children’s English in Singapore: Acquisition, Properties, and Use. London: Routledge. doi:10.4324/9781315201030 Census and Statistics Department (Hong Kong SAR). 2017. “2011 Population Census.” Accessed February 1, 2019 at http://www.census2011.gov.hk. Chew, Phyllis Ghim-Lian. 2013. A Sociolinguistic History of Early Identities in Singapore: From Colonialism to Nationalism. London: Palgrave Macmillan. doi:10.1057/9781137012340 Chong, Euvin Loong Jin, and Mark F. Seilhamer. 2014. “Young People, Malay and English in Multilingual Singapore.” World Englishes 33, no. 3: 363–77. doi:10.1111/ weng.12095 Da Silva, Beatriz. 1998. Cronologia da História de Macau: Século XIX. Macao: Fundação Macau. Deterding, David. 2007. Singapore English. Edinburgh: Edinburgh University Press. Downs, Jacques M. 1997. The Golden Ghetto: The American Commercial Community at Canton and the Shaping of American China Policy, 1784–1844. Bethlehem, PA: Lehigh University Press. Graddol, David. 2013. Profiling English in China: The Pearl River Delta. Cambridge: Cambridge English Language Assessment. Hall, Robert A., Jr. 1944. “Chinese Pidgin English Grammar and Texts.” Journal of the American Oriental Society 64, no. 3: 95–113. doi.org/10.2307/594237 Kachru, Braj. 1985. “Standards, Codification and Sociolinguistic Realism: The English Language in the Outer Circle.” In English in the World, edited by Randolph Quirk, and Henry G. Widdowson, 11–30. Cambridge: Cambridge University Press. Kirkpatrick, Andy. 2007. World Englishes: Implications for International Communication and English Language Teaching. Cambridge: Cambridge University Press. ———. 2011. “English as an Asian Lingua Franca and the Multilingual Model of ELT.” Language Teaching 44, no. 2: 212–24. doi:10.1017/s0261444810000145 Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig, eds. 2015. Ethnologue: Languages of the World. 18th ed. Dallas: SIL International. Accessed June 1, 2019 at http://www.ethnologue.com.

Interactions across Englishes  97 Lim, Lisa. 2004. “Sounding Singaporean.” In Singapore English: A Grammatical Description, edited by Lisa Lim, 19–56. Amsterdam: John Benjamins. ———, and Joseph A. Foley. 2004. “English in Singapore and Singapore English: Background and Methodology.” In Singapore English: A Grammatical Description, edited by Lisa Lim, 1–18. Amsterdam: John Benjamins. doi:10.1075/veaw. g33.03lim Matthews, Stephen, and Michelle Li. 2013. “Chinese Pidgin English.” In The Survey of Pidgin and Creole Languages. Vol. I: English-Based and Dutch-Based Languages, edited by Susanne Maria Michaelis et al., 206–13. Oxford: Oxford University Press. McArthur, Tom. 2002. The Oxford Guide to World English. Oxford: Oxford University Press. Meierkord, Christiane. 2012. Interactions across Englishes: Linguistic Choices in Local and International Contact Situations. Cambridge: Cambridge University Press. doi:10.1017/CBO9781139026703 Moody, Andrew. 2008. “Macau English: Status, Functions and Forms.” English Today 24, no. 3: 3–15. doi:10.1017/s0266078408000242 ———. 2019. “Educational Language Policy in Macau: Finding Balance between Chinese, English and Portuguese.” In The Routledge International Handbook of Language Education Policy in Asia, edited by Andy Kirkpatrick and Anthony J. Liddicoat, 76–96. London: Routledge. doi:10.4324/9781315666235-6 Mufwene, Salikoko. 2001. The Ecology of Language Evolution. Cambridge: Cambridge University Press. doi.org/10.1017/cbo9780511612862 Platt, John. 1991. “Social and Linguistic Constraints on Variation in the Use of Two Grammatical Variables in Singapore English.” In English around the World, edited by Jenny Cheshire, 376–88. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511611889.026 ———, and Heidi Weber. 1980. English in Singapore and Malaysia: Status, Features, Functions. Kuala Lumpur: Oxford University Press. Tan, Jason. 2007. “Schooling in Singapore.” In Going to School in East Asia, edited by Gerald A. Postiglione and Jason Tan, 301–19. Westport, CO: Greenwood Publishing Group. Tan, Ying-Ying. 2012. “To r or not to r: Social correlates of /ɹ/ in Singapore English.” International Journal of the Sociology of Language 218, no. 1: 1–24. doi:10.1515/ ijsl-2012-0057 ———. 2014. “English as a ‘Mother Tongue’ in Singapore.” World Englishes 33, no. 3: 319–39. doi:10.1111/weng.12093 ———, and Irving Goh. 2011. “Politics of Language in Contemporary Singapore Cinema: The Films of Jack Neo, or Politics by Cinematic Means.” Interventions: Journal of Postcolonial Studies 13, no. 4: 610–26. doi:10.1080/13698 01X.2011.628144 Tay, Mary V. J. 1991. “Southeast Asia and Hong Kong.” In English Around the World: Sociolinguistic Perspectives, edited by Jenny Cheshire, 319–32. Cambridge: Cambridge University Press. Wade, Geoff. 2001. “The Portuguese as Represented in Some Chinese Sources of the Ming Dynasty.” In Portugal e a China: Conferências nos encontros de história luso-chinesa, edited by Jorne M. dos Santos Alves, 263–316. Lisbon: Fundação Oriente.

98  Christiane Meierkord Wei, Rining, and Jinzhi Su. 2015. “Surveying the English Language across China.” World Englishes 34, no. 2: 175–89. doi:10.1111/weng.12132 Yan, Xi, and Andrew Moody. 2010. “Language and Society in Macao: A Review of Sociolinguistic Studies on Macao in the Past Three Decades.” Chinese Language and Discourse 1, no. 2: 293–324. doi:10.1075/cld.1.2.07xi Young, Ming Yee Carissa. 2006. “Macao Students’ Attitudes toward English: A Post-1999 Survey.” World Englishes 25, no. 3–4: 479–90. doi:10.1111/j.1467971x.2006.00468.x Yule, Henry, and Arthur Coke Burnell. 1886. Hobson-Jobson: Being a Glossary of Anglo-Indian Colloquial Words and Phrases and of Kindred Terms. London: John Murray.

Part II

Language Planning

5

One Legacy, Two Legislations Language Policies on the Two Sides of the Taiwan Strait Henning Klöter

Abstract This chapter examines (dis)continuities in language planning from Republican China to modern language planning in the People’s Republic of China (PRC) and Taiwan after the middle of the twentieth century and the political split of the two polities. From the perspective of language ideology, the 1920s marked a transitional period during which a former inclusive approach to language standardization, aimed at the integration of different regional features, gave way to an exclusive approach that defined correctness in terms of the pronunciation of the Beijing area. This shift has since then been upheld consistently on both sides of the Taiwan Strait, despite the fundamental ideological differences between the two polities. This continuity implied a growing marginalization of topolects through neglect (PRC) and suppression (Taiwan). In both cases, the spread of standard Mandarin has been successful, not only in terms of actual language use but also in terms of attitudes toward the standard.

Introduction Within sinophone studies, notions of multilingualism are usually accompanied by a critique of the dominance of Mandarin vis-à-vis other Sinitic varieties. This critique is, in turn, based on an understanding of Mandarin having a northern base, with China’s capital Beijing as its center. To be sure, with the exception of Hong Kong and Macao (see Li and Tong, this volume), all polities within the sinophone world have cemented the dominance of Mandarin in their sociolinguistic hierarchies. Whenever this position is challenged nowadays, the competitor is not a Sinitic alternative but English as the most important international language (see the chapters of Chen, Su, Meierkord, and Wasserfall, this volume). From a historical perspective, this dominance of Mandarin is remarkable, because just 100 years ago, modern Mandarin in its present spoken and written forms was still in its infancy and hardly known among the common people throughout China’s vast territory. Instead, they made use of the linguistic resources of their immediate environments, i.e. regional dialects or non-Sinitic languages of different ethnic groups. Due to a lack of reliable data, especially for the early twentieth

102  Henning Klöter century, it is not possible to capture the massive shift toward Mandarin in exact numbers and figures. Nonetheless, from various accounts it is clear beyond reasonable doubt that the victory of Mandarin must be considered the most striking feature of China’s sociolinguistic development during the twentieth century and beyond. Narratives of the history of Mandarin typically suffer from terminological ambiguities, the main cause of which is the lack of a clear Chinese equivalent. Whatever is used in Chinese to refer to Mandarin can have different linguistic or social connotations. The same is true for “China,” which can refer to quite distinct political entities over the past 100 years. In brief, three polities need to be distinguished here: The Republic of China (ROC) on the Chinese mainland, the ROC in Taiwan,1 and the PRC. Between 1912 and 1949, the ROC was under the leadership of the nationalist party Kuomintang (KMT). With regard to policy implementation in general and language policy in particular, it is important to emphasize that due to the continuing influence of warlords and the Japanese invasion, the KMT’s control over China remained limited during most of its reign. In 1949, after defeating the nationalist troops in the Chinese civil war, the Chinese Communist Party (CCP) came to power and founded the PRC. The ROC government relocated to the island of Taiwan which, until 1945, had been a colony of Japan. In Taiwan, the ROC gradually evolved into a multiparty democracy in which the KMT was no longer the sole ruling party. At the same time, members of the younger generation lost their identification with China. Thus, ever since 1949, the PRC and Taiwan have existed side by side, geographically separated by the Taiwan Strait and politically separated by ideological rivalry and antagonism. Against this backdrop, the main purpose of this chapter is to elaborate on two hypotheses. First, if we want to understand the dominance of Mandarin in its historical evolution, we have to take a close look at the 12 years following the foundation of the ROC in 1912, a period that arguably brought about a conceptual paradigm shift in modern language planning and can therefore be thought of as establishing the historical legacy of modern Mandarin. Second, the paradigm shift that occurred during this 12-year period was so fundamental that its effects persisted beyond political separation and still dominate language policies on both sides of the Taiwan Strait today.

Prologue: Defining Sinophone Multilingualism The critique of the dominance of Mandarin vis-à-vis other Sinitic varieties rests on the presupposition that Sinitic topolects like Mǐn, Wú, or Yuè can be regarded as languages distinct from Mandarin.2 In other words, these varieties meet criteria of “languageness,” a term used by Jaffe to explain the claim that Corsican “had internal unity and structure, and was clearly differentiated from other linguistic codes” (1999: 41). More importantly, “languageness” presupposes the existence of written forms, notably of literature. Significantly, in sinophone studies, the status of non-Mandarin

One Legacy, Two Legislations  103 Sinitic varieties has also been placed in the context of local-language literature (cf. Ng 2013: 88). Throughout the history of modern language planning, the notion of “languageness” has never been applied to regional Sinitic varieties. Instead, conceptually and institutionally, ever since the foundation of the ROC in 1912, language planning has always made a clear distinction between languages of ethnic minorities and Sinitic languages. Minority languages have been subject to planning, whereas topolects have not, one reason being the well-established link between language and ethnicity of the speakers. In contemporary terminology, Mandarin and topolects alike are subsumed under the term Hànyǔ, lit. ‘language of the Hàn people’, which has both linguistic and ethnic connotations (Wiedenhof 2015: 8). Thus, since speakers of topolects and speakers of Mandarin share one ethnicity, their unity also in linguistic terms seems to be taken for granted. Vice versa, due to their ethnic distinctiveness, the linguistic distinctiveness of members of non-Hàn ethnicities is readily recognized. Therefore, ever since the foundation of the ROC in 1912, the planning of Sinitic and non-Sinitic languages has been on separate institutional and ideological agendas. After the foundation of the PRC in 1949, the “right of minority peoples to use and develop their native languages and writing systems has been guaranteed constitutionally” (Zhou 2000: 130). Nothing in this respect was developed for topolects. Quite tellingly, as I will discuss presently, the promotion of Mandarin is only mentioned in one sentence of the constitution of the PRC, whereas three articles specify the language rights of ethnic minorities. In the ROC (Taiwan) constitution, only one article relates to language. Additional Article 10 says: “The State affirms cultural pluralism and shall actively preserve and foster the development of aboriginal languages and cultures” (online via https:// english.president.gov.tw).

From Integration to Exclusion: The Birth of a Legacy Haugen distinguishes four phases of a “path ‘underdeveloped’ languages must take to become adequate instruments for a modern nation” (1966: 931): (1) selection of a norm, (2) codification of form, (3) elaboration of function, and (4) acceptance by the community (ibid., 933). The first phase, selection of a norm, is self-evident. Codification refers to “the form of a language, i.e. its linguistic structure, including phonology, grammar, and lexicon” (ibid., 931), and also its spelling. Elaboration of function refers to the application of a norm in various functional domains, one important target being “utilization in writing” (ibid.). Finally, acceptance of a norm, “even by a small but influential group, is part of the life of the language” (ibid., 933). This four-phase model provides a handy set of descriptors that may help us to distinguish different aspects of language planning after the foundation of the ROC in 1912. From an institutional perspective, two types of agency in language planning and policy can be distinguished, i.e. governmental and

104  Henning Klöter ­

One Legacy, Two Legislations  105 path taken in imperial days. To be sure, there had been a concept of “correct pronunciation” (正音; see Söderblom Saarela, this volume), but correctness had never been determined narrowly in regional terms. The first standard dictionary of the national language, the Dictionary of National Pronunciation (國音字典) of 1919, continued this practice by codifying an artificial language that incorporated features of different Mandarin dialects. As Simmons has pointed out, this mixed language, known as “bluegreen Mandarin” (藍青官話), was not without precedent in Chinese history. It can be seen to have been heir to a practice in traditional times that struck a compromise between competing forms of prestige Mandarin koinés, including both a southern type and a northern type. (2017: 64) Equally important, he argues that “we can discern a popular preference for a composite standard that allowed for variation as well as including a set of features thought to be indispensable in the aggregate, but not found collectively in any single spoken dialect” (ibid.). In the words of Simmons, in imperial days the standard was a vague set of conventions for speaking and writing that the educated literati generally held to be most preferred and most useful for oral and written communication throughout China’s vast territory and across the many dialects and varieties of spoken Chinese. (this volume) In short, there was not one single manifestation of one Mandarin; instead, the concept of a composite standard to be loosely followed prevailed. There were different reasons for initially continuing to use a composite standard, the most important of which brings us back to Haugen’s aforementioned four-phase model. He writes that codification is unlikely to succeed “unless the community can agree on the selection of some kind of a model from which the norm can be derived” (1966: 932). He further argues that “where there are socially coordinate groups of people within the community, usually distributed regionally or tribally, the choice of any one [norm] will meet with resistance from the rest” (ibid.). One could almost be tempted to argue that Haugen was looking at language planning in the founding years of the Chinese Republic when he wrote these lines. In China, there was broad agreement that a national pronunciation norm had to be selected. There was also agreement that this norm should be codified with a phonographic script indicating the reading of characters. However, from the very beginning, the “community” was unable to agree on a specific norm, and resistance between different regional factions was strong. The reason was simple: The proposed selection of Beijing as the regional base of

106  Henning Klöter the new national language would have created a diglossic hierarchy to the disadvantage of all other regional languages. In Haugen’s terms, the conference delegates had to make some embarrassing decisions. To choose any one vernacular as a norm means to favor the group of people speaking that variety and to disfavor others. It gives one group of speakers prestige as norm-bearers and a headstart in the race for power and position. (1966: 932) In addition to the quandary of political decision making, the initial pursuit of an integrative approach to language planning also had an obvious ideological dimension (cf. Weng 2018). The idea of having one national language reflecting different local varieties of Mandarin was more compatible with contemporary ideals of national unity than an exclusive approach to language standardization. Quite tellingly, language specialists and language planners alike played down the dimensions of variation within Mandarin. For example, the linguist Li Chin-Shi 黎錦熙 (Lí Jǐnxī 1890–1978) wrote that “unification of pronunciation merely implies adjusting the minor differences among a great bulk of similarities, by fixing a standard” (Li 1922: 35). Against this backdrop, the first years after the founding of the ROC can be described as a hybrid period that was characterized by two competing approaches: an inclusive and abstract (i.e. unspoken) national language standard versus an exclusive and region-based standard. At some point during the year 1924, the tide turned toward the latter (for details see Simmons 2017: 79). What motivated the shift from an integrative to an exclusive approach in language planning? Most importantly, the composite standard of “bluegreen Mandarin” simply did not work when it came to its dissemination. Anecdotal evidence suggests that the newly codified national pronunciation had exactly one speaker: the Chinese-American linguist Yuen Ren Chao 趙 元任 (1892–1982), who later recalled ironically that “for thirteen years I was the sole speaker of this idiolect, meant to be the national language of 4, 5, or 600 million speakers” (1961: 175). In a similar vein, Simmons points out that teachers in China, who were by definition not native speakers of the standard, turned out to have problems in deciphering the complex system of indicating pronunciation in the Dictionary of National Pronunciation. He argues that the complexity of the dictionary’s presentation seriously diminished the prospects of effectively establishing the Guóyīn standard. Those who were called upon to teach the National Pronunciation surely found it highly difficult to determine the complete pronunciation of the characters in the dictionary, the average user probably even more so. (2017: 78)

One Legacy, Two Legislations  107 To be sure, cumbersome lexicographic devices were only one part of a bigger problem. More importantly, there was a lack of sufficient infrastructure that would have allowed for efficient acquisition planning, i.e. a planning toward “an increase in the users of a language or language variety” (Cooper 1989: 33). Although these reasons are certainly convincing, it seems hard to conceive that the paradigm shift from an inclusive to an exclusive approach in language planning was made for practical reasons only. In the paragraphs that follow I shall propose additional explanations, which are to be considered working hypotheses that need further empirical testing on the basis of historical documents. As has been pointed out earlier, national language debates in China followed the track beaten by language ideologies in Japan during the late Meiji period (1868–1912) (cf. Chen 2007: 145). But there were also subtle but nonetheless important differences between Japan and China. In Japan, some ten years earlier than in China, the linguist Ueda Kazutoshi 上田萬年 (1867–1937) had played a decisive role in Japanese language standardization (Heinrich 2012: 64). Ueda’s linguistic thinking had been inspired by notions of language purism that were rooted in German discourses of the late nineteenth century. As Heinrich points out, “Ueda closely followed the debate surrounding language purification in the name of Germanization (Verdeutschung), as promoted by the General German Language Association (Allgemeiner Deutscher Sprachverein)” (Heinrich 2012: 61). In a similar vein, “Ueda strongly supported the idea that Japanese be liberated from what he perceived as the yoke of Chinese loanwords” (ibid., 64). The purified standard promoted by Ueda was to be based on the “Tokyo language variety of educated speakers” (ibid., 68). Ueda’s approach can rightly be considered a textbook example of language purism combined with standardization and prescriptivism. Language purism typically reflects “a desire to preserve what speakers believe to be the ‘pure’ or ‘true’ form of the language” (Ameka 2016: 74) which comes along with attempts to protect it from outside influences (Kickham 2015: 50). Purism cannot be separated from attempts to impose the belief of what is pure and true on the language practice of others, ideally the entire language community. Imposing norms on the behavior of others requires the prescription of norms and standards. I thus concur with Brunstad’s claim that “both purism and language standardisation may be regarded as aspects of the same processes of norm homogenization” (2003: 53). Against this backdrop, a crucial difference between language planning in Japan and China at the beginning of the twentieth century can be described in terms of purism. In contrast to the Japanese national language, its Chinese counterpart was initially neither an existing language variety nor something that was deemed worthy of protection from outside influence. It was only with the decision in 1924 to define correctness on the basis of the pronunciation of the capital Beijing that ROC language planning resembled its puristic Japanese model more closely. Prior to the shift toward an exclusive

108  Henning Klöter approach, i.e. during the “blue-green period,” notions of protection from outside influence apparently did not receive noteworthy attention in China. As a matter of fact, linguistic nationalism had risen amid large-scale outside influence. Masini has pointed out “that about one third of the basic vocabulary of modern Chinese, or Modern Chinese lexicon, was created during the nineteenth century and the early decades of the twentieth century” (2017: 65). This was paralleled by changes in written Mandarin of Chinese prose which experienced grammatical and stylistic innovations during the late nineteenth and early twentieth centuries (Gunn 1991: 36). This openness toward the outside does not imply, however, that purism was absent from Chinese linguistic thinking in those days. The very fact that attempts at establishing a standard were made implies a notion of exclusiveness in terms of what is correct and what is incorrect that had been unknown in imperial times, most importantly in the area of pronunciation. It is this notion of exclusiveness that showed that the days of laissez-faire of the imperial epoch were over. Instead, the entire process of standardization was accompanied by fierce debates on correctness. This does not contradict my aforementioned reference to the premodern concept of “correct pronunciation.” Whereas the latter can be perceived as a rather loosely defined guideline for the spoken practice of court officials, the modern national language was closely bound to a strict notion of prescriptivism that was supposed to permeate the entire population. In two respects, “blue-green Mandarin” can thus be described as a hybrid. There was hybridity in terms of its codification, i.e. the combination of northern and southern elements. And there was also hybridity in terms of “standardness,” i.e. exclusiveness vis-à-vis non-recognized varieties combined with inclusiveness in terms of recognizing forms originating from different regions. From the perspective of an individual speaker, this meant that some forms of their own variety were correct but only when used in combination with forms from other varieties. It seems hard to conceive that such a standard is enforceable. One could therefore argue that it was this double hybridity that ultimately co-contributed to the failure of “blue-green Mandarin.” This would imply that once standardization as a process has been implemented, purism inevitably gains acceptance. The growth of purism benefited from yet another factor that was specific to China, i.e. the decision to officially use a phonetic script as a supplement to Chinese character writing. To be sure, the late nineteenth century had witnessed several attempts to create phonetic notation systems (cf. Tsu 2010: 18–48). In the early twentieth century, phonetic writing came to be closely associated with Western notions of science and logical thinking (cf. Wang 2011: 118). Disputes raged over two issues, i.e. the choice of a set of graphemes for the indication of pronunciation and, as discussed earlier, the choice of a regional base for the new standard pronunciation. The former was decided in favor of a phonetic notation system known as Mandarin Phonetic Symbols (注音字母; for details, see Wippermann 2017a). Therefore, with the

One Legacy, Two Legislations  109 decision to introduce a phonetic script, language planning could no longer evade a commitment toward unambiguousness in pronunciation. As a consequence, linguists and educators like Cài Yuánpéi 蔡元培 (1868–1940), Wú Zhìhuī 吳稚暉 (1865–1953), Wáng Zhào 王照 (1859–1933), or Lú Zhuàngzhāng 盧戇章 (1854–1928) were faced with the challenge to devise a scheme that would unambiguously indicate “what sounds were to be represented” (DeFrancis 1950: 57).

The Effects of Language Planning before 1949 On the basis of previous research, it can be concluded that the spread of written Mandarin or báihuàwén was the greatest success of language planning. After the founding of the ROC, báihuàwén quickly won wide recognition as the new language of literature. Its rapid spread benefited from cultural, technological, and social factors. From a cultural perspective, it must be emphasized that language planning in the Republican period had always been a symbiosis of spoken language standardization and the formation of a new literary language. This symbiosis takes center stage in the well-known articles by Hú Shì 胡適 (1891–1962), one of the intellectual masterminds of the new literature movement. His call for the creation of “Literature in the National Language and a literary National Language” (國語 的文學，文學的國語, Hú 1918), according to Li, “signified the convergence between the national language movement and the literary revolution” (Li 1934: 136; quoted and translated by Liu 2013: 33). In other words, in the case of China, codification and elaboration of a function (i.e. use of the new standard in literary works) were two sides of the same coin. The successful spread of written Mandarin also benefited from the emergence of print media. According to Snow, by 1919 “at least 400 newspapers across China [were] published in the vernacular, and this number was to increase” (2010: 138). Production of textbooks accelerated the spread of Mandarin. A ministerial decree of 1920 declared written Mandarin as standard in first- and second-year classes in lower primary schools (cf. Culp 2008). This decision, in the words of Culp, is “often cited as a threshold moment in the movement for vernacular language” (2008: 4). According to Culp, with “mass production of primary- and secondary-level vernacular textbooks during the early 1920s, the reading public for modern baihua grew quickly and exponentially as the nation’s roughly 6.8 million primary and secondary students all became potential readers” (ibid., 26–27). Much less is known about the spread of the spoken standard, i.e. the national pronunciation. However, there is good reason to assume that language planning in this field was much less successful. Kaske points out that Mandarin Phonetic Symbols “remained virtually unknown” (2008: 417). We can only speculate about how successfully a spoken standard could have spread without its annotation system. Some relevant information with regard to the spread of spoken Mandarin can be found in newspaper comments released

110  Henning Klöter in the 1930s, during the “Mass Language Movement” (大眾語運動). To give one example, on June 29, 1934, a person named Chén Yì 陳弈 published the following statement in the newspaper Evening News (大晚報): I have personally been in many places like steamboats, trains, docks, stations, inns, restaurants, and fairs, but I have never heard this Common Language that is supposed to have spread to every place. When you come to a place, you hear the language of that place. For example, when you come to the boats, trains, docks, stations, and fairs of Hong Kong or Canton, then you always hear Cantonese or even dialects from the inner regions. (Wén Yì 1934: 206) This letter to the editor was just one of many contributions to a debate on the language of the common people in China. Although solid sociolinguistic data of this period are not available, we can assume, beyond reasonable doubt, that the earlier quotation adequately depicts China’s southeastern provinces as a linguistically highly diverse and fragmented area. Interestingly, however, many participants in the Mass Language Movement, most of whom were leftist intellectuals, expressed sympathy for the notion of peasants and workers from different regions being able to communicate with each other. For these intellectuals, such a lingua franca would coexist with regional dialects. As a matter of fact, quite a few intellectuals expressed sympathy for the cultivation of regional languages or “dialects.” The most prominent example was Qū Qiūbái 瞿秋白 (1899–1935), a former correspondent to Moscow, who “argued for the legitimate existence of dialect scripts” (Liu 2013: 41). Qū’s proposal was later supported by some of China’s famous literary writers, including Lǔ Xùn 魯迅 (1881–1936) and Máo Dùn 茅盾 (1896–1981). Despite this support, it would go too far to posit the existence of “a leftist alternative to a single national standard,” as claimed by Ramsey (1987: 14). Nonetheless, it is certainly true that opposition to the policy to impose a single standard on the whole of China came from leftist intellectuals, who were typically driven by the ideal to support the illiterate masses of workers and peasants. To the best of my knowledge, however, leftist critique of national language policies has never materialized into a concrete counterproposal. In short, then, the ideal of establishing a national pronunciation remained an ideal. The significance of the national language movement therefore lies in its ideological force. As I shall discuss in the paragraphs that follow, it not only endured the regime change in 1949 but also reached Chinese communities in far-away places like Singapore, where the new standard played an important role in language planning (see Goh and Fong, this volume). The limited spread of the new standard within China implies that the topolects of southeastern China occupied various domains and were used as the main language of inter-generational communication and thus enjoyed a high degree of linguistic vitality. Written Mandarin or báihuàwén, on the

One Legacy, Two Legislations  111 other hand, had successfully replaced Literary Chinese in literary composition, textbook, and dictionary compilation (cf. Klöter 2019). Failures in the spread of the national language are often attributed to the weak political system as a whole and are therefore not attributable to language planning as an isolated field of political decision making. Yet it has to be emphasized that the period beginning with the foundation of the ROC until the end of the Mass Language Movement marks the decisive years of modern language planning in China. The core language ideology of one national language for one unified Chinese nation had been born and defended against oppositional voices from the left. As I will also argue, after 1949, the core ideals of language planning that had emerged during the Republican period became the shared language ideology of two mutually hostile Chinese governments.

Weathering Separation: The Long-Standing Effects of a Legacy Language Legislation and the Spread of Mandarin in the PRC With the establishment of the PRC in 1949 and the relocation of the ROC government to the island of Taiwan in the same year, the ideological, political, and institutional conditions of language planning in the sinophone world changed fundamentally. The resulting political divide of China and Taiwan provides a unique example of two antagonistic polities “inheriting” the same package of language ideologies and the same approach in language planning. Most importantly, Mandarin language planning on either side of the Taiwan Strait is a direct continuation of the “critical period” 1912–24. The comparison of post-1949 language policies on both sides of the Taiwan Strait reveals a high degree of continuity with regard to the dominance of Beijing-based Mandarin as the uncontested national language—despite the conflicts surrounding its introduction some decades earlier. In its general political rhetoric, the new ruling party of China, the CCP, tried to distance itself as much as possible from its predecessor and political enemy, the KMT. In language planning and policy, however, there was considerable continuity before and after 1949. After a series of conferences on script reform (文字改革) and the standardization of Modern Chinese (現代 漢語規範化), the State Council declared in February 1956: Pǔtōnghuà is the standard form of Modern Chinese with the Beijing phonological system as its norm of pronunciation, and Northern dialects as its base dialect, and looking to exemplary modern works in báihuà ‘vernacular literary language’ for its grammatical norms. (translated by Chen 1999: 24) This formal declaration was preceded by similar statements by then Minister of Education Zhāng Xīruò 張奚若 (1889–1973) and the official newspaper People’s Daily (人民日報) (cf. Simmons 2017: 66). Thus, as has been pointed

112  Henning Klöter out by Ramsey, “this was the very program that had been advocated by supporters of the National Language, pursued by the Kuomintang government, and bitterly opposed by the left” (Ramsey 1987: 14–15). As a matter of fact, according to Simmons, it was only after the foundation of the PRC that the issue of using the pronunciation of the Beijing region as the national norm “was finally definitely settled” (2017: 66). The focus of PRC national language planning was not placed on the definition of the standard, but on its implementation, i.e. the effective spread of the national language in all public domains throughout the country. This process of implementation was to be accompanied by measures in other areas of language planning, most importantly the simplification of the Chinese script (Bökset 2006; Chen 1999: 148–63; Zhao 2008), standardization of the modern lexicon and the development of lexicography (Chen 2001; Kholkina 2017; Klöter 2019; Lee 2014; Tang 2017), introduction of Hànyǔ Pīnyīn as a supplement to Chinese character writing (Chen 1999: 186–89; Wippermann 2017b), and, more recently, large-scale internationalization of Chinese language teaching (Zhou 2018: 207–48). From today’s perspective, national language planning in the PRC has been extremely successful. Official reports praise “brilliant achievements” and argue: In the past six decades, the following major events were accomplished in the field of national language management: The formulation and implementation of the fundamental policies of the national language and writing system, including the equality and co-existence of all ethnic languages; […] the national promotion of Putonghua for general use and the simplification of Chinese characters; the modernization and standardization of Chinese characters; the formulation and promotion of Chinese Pinyin; and the further standardization of contemporary Chinese as well as the maintenance of minority national languages. (Xie and Chen 2015: 33) According to the UNESCO Institute for Statistics, the overall adult literacy rate in China had reached 95 percent by 2010 (online via data.worldbank. org). In 2017, 60 years after the release of the first guidelines for its promotion, some 70 percent of the population were able to speak Pǔtōnghuà (Lǐ 2017). Language competencies are one part of the story; attitudes of speakers toward the standard are another. Generally speaking, there is a lack of comparable empirical data that would give clear-cut answers to the question to what extent speakers within the political boundaries of the PRC have accepted the dominance of Pǔtōnghuà. Official reports are rather blunt when they argue that “Putonghua as the language of official business has … become deeply rooted in the hearts of the people” (Wáng and Yuán 2013: 32). In a similar vein, Zhou cites research reports that point to the broad acceptance of Pǔtōnghuà by speakers, regardless of the actual language or

One Legacy, Two Legislations  113 dialect they use in daily life (Zhou 2018: 152). Blum likewise concludes that “Standard Putonghua is good to hear, even if for most of China’s 1.3 billion people it is not necessarily good (easy, appropriate) to speak” (2004: 139). Yet on the other hand, widespread anger against the proposal that one TV channel should switch from Cantonese to Pǔtōnghuà and the resulting “Protecting Cantonese Movement” in 2010 tell quite a different story (Gao 2017; Qu 2015; Zhang and Guo 2012). Blum’s earlier quoted claim that most people in China do not speak Pǔtōnghuà well may seem to contradict the suggested rate of 70 percent speakers. One explanation is that Blum’s analysis is based on actual usage which she contrasts with the “pure” standard used in the state media. Since actual usage by Mandarin speakers is typically influenced by regional features (for details, see Lin, this volume), there is often if not always a gap between actual usage and standard Mandarin. However, even if we consider an approximation of the standard as “more or less standard” Pǔtōnghuà, Blum’s claim would not be compatible with the official success story of language planning. Another explanation could be that the systematic implementation of Mandarin acquisition planning only gained ground after the beginning of the reform era and was significantly intensified after the turn of the twenty-first century. The PRC constitution of 1982 for the first time contains a sentence on the promotion of Mandarin according to which “The state promotes the nationwide use of Putonghua” (國家推廣全國通用的普通話, online via en.pkulaw.cn). Interestingly, only the official English translation explains, in parentheses, “common speech based on Beijing pronunciation” (ibid.). In 1986, the principal task of language planning was re-defined as the promotion of Pǔtōnghuà (Rohsenow 2004: 31). Subsequently, the following goals were set for the end of the twentieth century: (1) Putonghua to become the language of instruction in all schools nationwide; (2) Putonghua to become the working language of government at all levels; (3) Putonghua to be the language used in radio and television broadcasting, and in cinemas and theatres; (4) Putonghua to become the common language used among speakers of various local dialects. (ibid.) These and other targets were reconfirmed in the “Law of the People’s Republic of China on the National Commonly Used Language and Script” (中華人民共和國國家通用語言文字法) that passed the Standing Committee of the National’s People’s Congress in 2000 (CPG 2005; Rohsenow 2004: 36). However, amid this new dedication to the promotion of Mandarin, some sobering figures were released in 2004, when an official language report mentioned that only 53.06 percent of the population were able to communicate in Pǔtōnghuà (cf. Wáng and Yuán 2013: 36). As a consequence, since the beginning of the twenty-first century, official language planning agencies have taken a number of measures aimed at intensifying the spread

114  Henning Klöter of Mandarin. For example, in regions that are known for the widespread use and popularity of regional varieties, notably wealthy urban centers like Guangzhou and Shanghai, Mandarin became the sole medium of instruction (for Guangzhou see Liang, this volume). Furthermore, the Ministry of Education and its State Language Commission (國家語言文字工作委員會) have systematically evaluated the linguistic performance of entire cities by conducting inspection tours (for details, see Klöter 2016: 62f.). In addition to inspecting and testing, the promotion of Mandarin has relied on campaigning. Ever since 1998, the Ministry of Education has organized an annual “National Publicity Week for Pǔtōnghuà Promotion” (全国推广普通话宣传 周), which was held for the 21st time in the city of Tianjin in September 2018 (Wáng and Yuán 2013, MOE press release via http://www.moe.edu.cn). Language Legislation and the Spread of Mandarin in Taiwan From the perspective of the people in Taiwan, the regime change of 1945 brought fundamental changes. From the perspective of the ROC government, on the other hand, the fact that the small island of Taiwan was added to its vast territory did not entail any ideological, legislative, or terminological changes. Since Taiwan now fell under the jurisdiction of the ROC, existing laws and regulations were simply extended to the new territory, including the core agenda of language planning—promotion of Mandarin as the “national language” (國語) and the use of traditional characters and Mandarin Phonetic Symbols. For those on the receiving end of language planning, the speakers in Taiwan, it needs to be emphasized that Mandarin in 1945 was virtually a foreign language, since most people spoke a regional Sinitic variety, either Southern Mǐn or Hakka, and also Japanese. Institutional preparations for the promotion of Mandarin started immediately after the surrender of Japan in 1945 (for details, see Tsao 2008: 249f.). Most importantly, in 1946, with the establishment of the Taiwan Provincial Council for the Promotion of the National Language (臺灣省國語推行委員 會) the organizational foundation for the spread of Mandarin was laid. Initially, people in Taiwan were even encouraged to use their local dialects, but this was merely considered a temporary measure “as a first step toward the acquisition of Mandarin by the local people” (Feifel 1994: 72; see also Tsao 2008: 250). In the short period between 1945 and 1949, the political situation underwent significant changes. The KMT government was defeated by the Communist troops during the Chinese civil war (1945–49) and, as a consequence, relocated its government to Taiwan. Ideologically, despite this relocation, the KMT continued to consider itself the sole legitimate ruler of China. In reality, however, its political sphere of influence came to be restricted to Taiwan, the Penghu archipelago, and the offshore islands of Kinmen 金 門 and Matsu 馬祖. When the ROC government relocated to Taiwan, it encountered a speech community that was much smaller and therefore much

One Legacy, Two Legislations  115 more manageable in terms of number of speakers, number of languages, and size of the territory. Furthermore, as a result of Japanese colonial policies, Taiwan was much more advanced in terms of infrastructure and public education. In other words, the conditions for successful language planning were much better than they had ever been in ROC history. If the sole purpose of post-1949 language planning had been to facilitate successful communication, Taiwanese Southern Mǐn would have been the obvious candidate for a standard language. Nonetheless, despite the fundamental differences in terms of linguistic ecology of the Chinese mainland and Taiwan, the KMT government continued to pursue the course in language planning it had embarked upon during the 1920s. The Mandarin that was promoted was unmistakably and uncompromisingly northern Mandarin based on the Beijing region. Rather ironically, therefore, the new national pronunciation that had resulted from ROC language planning was only endorsed when the standard had fully lost touch with Beijing as its concrete regional base. The promotion of Mandarin had two major target domains: the media and education. From 1946 onward, Qí Tiěhèn 齊鐵恨 (1892–1977), a member of the Mandarin Promotion Council, conducted daily “model Mandarin” radio programs aimed at the dissemination of correct national pronunciation (cf. Fèi 1997: 101). In 1976, the ROC government promulgated a Radio and Television Act (廣播電視法) which, until 1993, limited the use of languages other than Mandarin in radio and television programs (cf. Chang and Holt 2015: 33; Huang 1993: 354–73). In the education domain, after an initial period of “linguistic improvising” due to a lack of competent Mandarin teachers (see Lin, this volume), Mandarin was firmly established as the sole medium of instruction. The implementation of this measure is arguably the most extreme documented example of restrictive language policies in the sinophone world. Many Taiwanese who attended school after the 1950s report about physical punishment and other harsh penalties for using local languages in school. Things changed after the late 1980s, when Taiwan’s society developed toward pluralism, and the political system toward democracy. When former restrictions on the use of local languages were successively lifted, Mandarin had been firmly established as Taiwan’s dominant language, despite measures aimed at the protection of Taiwan’s linguistic diversity (for details, see Chen, this volume). The first of these were initiated in the early 1990s, with the introduction of a mother tongue education program. This program was initially directed at aboriginal languages and started in 1990 in the northern district of Wūlái 烏來, with a program for children of the Atayal (泰雅族) ethnic group at elementary and junior high school levels (cf. Liao 2000: 168). A decade later, similar programs were launched at the elementary level for Sinitic mother tongues other than Mandarin. It has been argued that the introduction of mother tongue education and the draft of a language equality law (語言平等法) in the early twenty-first century mark a fundamental paradigm change in sinophone language planning. As Chen (this volume) points

116  Henning Klöter out, however, the introduction of compulsory mother tongue education at the primary school level has not been accompanied by a systematic set of related language planning measures. In addition, the planned language equality law has not been passed. The same is true for the latest initiative, the National Language Development Law (國家語言發展法) which was in the process of formal approval at the time this chapter was written. According to a press release, the draft bill “guarantees the right of citizens to use any of Taiwan’s national languages and establishes the equality of all languages and cultures” (Executive Yuan 2018). It can therefore be stated that under different names and with different ideological wrappings, the planning of Mandarin has followed very similar pathways on both sides of the Taiwan Strait. As in the PRC, the planning of Mandarin in Taiwan has been extremely successful. It is spoken by the overwhelming majority of the population and continues to be the dominant language in education and the media. Moreover, it has increasingly entered private domains. Against this backdrop, I would add a note of caution with regard to Zhou’s claim that language planning on the two sides of the Taiwan Strait has been developing in opposite directions. He argues: In this new language order, the PRC requires every citizen to learn and use Putonghua while limiting the use of their mother tongues. On the other hand, the ROC has evolved in the opposite direction […]. Ideologically, the ROC has shifted from monolingualism to multilingualism in its democratic process since the early 1990s. The ideological evolution leads to the development of a new multilingual order in Taiwan. (Zhou 2018: 256) Whereas it is certainly true that sinophone polyphonicity prevails in Taiwan, regional languages continue to face marginalization. As I have argued before (Klöter 2006), language planning with regard to Taiwan’s Sinitic topolects has been largely symbolic and has not aimed at imposing significant changes on an order that goes back to the 1920s. Recent changes in language planning, notably the introduction of mother tongue education and legislative initiatives, have not fundamentally altered the historical legacy. Significantly, recent research has pointed out that primary and junior high school students seldom use Taiwanese Southern Mǐn, and “the intergenerational transmission of their mother tongue is impeded” (Tiun 2018).

Conclusion The 12-year period 1912–24 had a deep impact on language planning in China. Nurtured by nationalism and purism, language planning gradually abandoned its initial inclusive approach and adopted an exclusive strategy by defining the pronunciation of the Beijing region as the sole yardstick for correct language use. This strategic change arguably brought about the most fundamental paradigm shift in modern language planning and

One Legacy, Two Legislations  117 therefore created a lasting legacy of modern Mandarin. Ever since, the basic principles of defining standard in language have remained unchallenged in two polities that are committed to the spread of Mandarin: the PRC and Taiwan. It is only in recent years that language planning in Taiwan has cautiously taken a more supportive approach toward diversity in the planning of different Sinitic languages. Due to a lack of systematic and integrated planning, however, this approach can be expected to have limited effects only. On both sides of the Taiwan Strait, the victory of Mandarin has created a diglossic configuration in which Mandarin is the uncontested high variety, dominating over different low varieties. As a result, aside from spoken language use, Chinese literacy on both sides of the Strait is by definition Mandarin literacy. This contrasts with Hong Kong and Macao, where vernacular literacy in Cantonese still prevails (see Li and Tong, this volume). High literacy rates on both sides of the Strait are also clear indicators of the success of language planning. What, then, are the side effects of language planning with regard to Sinitic topolects? It is an obvious tendency on both sides of the Taiwan Strait that the use of topolects is rapidly decreasing. As a matter of fact, the vitality of topolects has been severely weakened, a phenomenon which has been referred to as “dialect crisis” (Gao and Shao 2018: 303). Commenting on this development, Guo writes that the “question was only whether to eradicate dialects artificially […] or to allow them to die out naturally as Putonghua spread to every domain of language use” (Guo 2004: 47). My claim that topolects have never been subject to systematic language planning, however, needs to be put in perspective. The emphasis lies on official language planning. Beyond the official level, there have been various attempts at protecting and even cultivating Sinitic topolects. The aforementioned “Protecting Cantonese Movement” is just one very recent example. Other examples such as the Mass Language Movement in China (Lee 2016) or the Taiwanese nativist literature movement (documented in Nakajima 2003) are positioned at the interface between language and literature studies. A comparative sociolinguistic analysis of these and other movements awaits future research.

Acknowledgments The author wants to thank Jeroen Wiedenhof and Richard VanNess Simmons for their valuable comments on an earlier draft of this chapter.

Notes 1 For stylistic simplicity and without political implications, this polity will simply be referred to as “Taiwan” in the remainder of this chapter. 2 For an overview of Sinitic dialect groups, see Norman (1988, chapters 8 and 9). 3 As Söderblom Saarela points out, báihuàwén (literally ‘plain speech writing’) as an “alternative written language” had been firmly established by the eighteenth century. Before the Republican period, however, texts written in báihuàwén had never been recognized as a part of the literary canon (this volume).

118  Henning Klöter

References Ameka, Felix. 2016. “The Uselessness of the Useful: Language Standardization and Variation in Multilingual Contexts.” In Prescription and Tradition in Language: Establishing Standards across Time and Space, edited by Ingrid TiekenBoon van Ostade and Carol Percy, 71–87. Clevedon: Multilingual Matters. doi:10.21832/9781783096510-007 Blum, Susan D. 2004. “Good to Hear: Using the Trope of Standard to Find One’s Way in a Sea of Linguistic Diversity.” In Language Policy in the People’s Republic of China, edited by Minglang Zhou and Sun Hongkai, 123–42. doi:10.1007/1-4020-8039-5_7 Bökset, Roar. 2006. “Long Story of Short Forms: The Evolution of Simplified Chinese Characters.” PhD diss., Stockholm University. Brunstad, Endre. 2003. “Standard Language and Linguistic Purism.” Sociolinguistica 17, no. 1: 52–70. doi:10.1515/9783110245226.52 Chang, Hui-Chang, and Richard Holt. 2015. Language, Politics and Identity in Taiwan: Naming China. London: Routledge. doi:10.4324/9780203362044 Chao, Yuen Ren. 1961. “What is Correct Mandarin?” Journal of the American Oriental Society 81, no. 3: 171–77. doi:10.2307/595651 Chen, Ping. 1999. Modern Chinese: History and Sociolinguistics. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139164375 ———. 2001. “Development and Standardization of Lexicon in Modern Written Chinese.” In Language Planning and Language Policy: East Asian Perspectives, edited by Nanette Gottlieb and Ping Chen, 49–73. Richmond, Surrey: Curzon. ———. 2007. “China.” In Language and National Identity in Asia, edited by Andrew Simpson, 141–67. Oxford: Oxford University Press. Cooper, Robert L. 1989. Language Planning and Social Change. Cambridge: Cambridge University Press. CPG (Central People’s Government of the PRC). 2005. “Zhōnghuá Rénmín Gònghéguó guójiā tōngyòng yǔyán wénzì fǎ.” 中华人民共和国国家通用语言文字法 [Law of the people’s republic of China on the national commonly used language and script]. Accessed March 08, 2020 via http://www.gov.cn Culp, Robert. 2008. “Textbook Publishing and the Production of Vernacular Language and a New Literary Canon in Early Twentieth-Century China.” TwentiethCentury China 34, no. 1: 4–41. doi:10.1353/tcc.0.0011 De Francis, John. 1950. Nationalism and Language Reform in China. Princeton, NJ: Princeton University Press. Executive Yuan, Department of Information Services. 2018. “Press Release: Executive Yuan Approves Draft National Language Law.” Accessed November 21, 2019 via https://english.ey.gov.tw. Fèi Jǐnchāng 费锦昌, ed. 1997. Zhōngguó yǔwén xiàndàihuà bǎinián jìshì 中国语文 现代化百年记事 [A chronicle of one hundred years of language modernization in China]. Beijing: Yǔwén chūbǎnshè. Feifel, Karl-Eugen. 1994. Language Attitudes in Taiwan: A Social Evaluation of Language in Social Change. Taipei: Crane. Gao, Xuesong. 2017. “Linguistic Instrumentalism and National Language Policy in Mainland China’s State Print Media Coverage of the Protecting Cantonese Movement.” Chinese Journal of Communication 10, no. 2: 157–75. doi:10.1080/175 44750.2016.1207694

One Legacy, Two Legislations  119 ——— and Qing Shao. 2018. “Language Policy and Mass Media.” In The Oxford Handbook of Language Policy and Language Planning, edited by James W. Tollefson and Miguel Pérez-Milans, 299–317. Oxford: Oxford University Press. doi:10.1093/oxfordhb/9780190458898.013.19 Gunn, Edward. 1991. Rewriting Chinese: Style and Innovation in Twentieth-Century Chinese Prose. Stanford, CA: Stanford University Press. Guo, Longsheng. 2004. “The Relationship between Putonghua and Chinese Dialects.” In Language Policy in the People’s Republic of China: Theory and Practice since 1949, edited by Zhou, Minglang and Hongkai Sun, 45–54. Dordrecht: Kluwer Academic Publishers. doi:10.1007/1-4020-8039-5_3 Haugen, Einar. 1966. “Dialect, Language, Nation.” American Anthropologist 68: 922–35. doi:10.1525/aa.1966.68.4.02a00040 Heinrich, Patrick. 2012. The Making of Monolingual Japan: Language Ideology and Japanese Modernity. Bristol: Multilingual Matters. doi:10.21832/9781847696588 Huang Shuan-fan 黃宣範. 1993. Yǔyán, shèhuì yǔ zúqún yìshí: Táiwān yǔyán shèhuìxué de yánjiū 語言社會與族群意識: 臺灣語言社會學的研究 [Language, society and ethnic identity: A sociolinguistic study of Taiwan]. Taipei: Crane. Jaffe, Alexandra. 1999. “Locating Power: Corsican Translators and their Critics.” In Language Ideological Debates, edited by Jan Blommaert, 39–66, Berlin: De Gruyter Mouton. doi:10.1515/9783110808049.39 Kaske, Elisabeth. 2008. The Politics of Language in Chinese Education, 1895–1919. Leiden: Brill. doi:10.1163/ej.9789004163676.i-537 Kholkina, Liliya. 2017. “Lexicography, Modern.” In Encyclopedia of Chinese Language and Linguistics, vol. 2, edited by Rint Sybesma et al., 594–600. Leiden: Brill. Kickham, Elizabeth. 2015. “Purism, Prescriptivism, and Privilege: Choctaw Language Ideologies and Their Impact on Teaching and Learning.” PhD diss., University of Oklahoma. Klöter, Henning. 2006. “Mandarin Remains More Equal: Changes and Continuities in Taiwan’s Language Policy.” In What Has Changed? Taiwan before and after the Change in Ruling Parties, edited by Dafydd Fell, Henning Klöter, and Bi-yu Chang, 207–23. Wiesbaden: Harrassowitz. ———. 2016. “‘What is Correct Chinese?’ Revisited.” In Prescription and Tradition in Language: Establishing Standards across Time and Space, edited by Ingrid Tieken-Boon van Ostade and Carol Percy, 57–70. Clevedon: Multilingual Matters. doi:10.21832/9781783096510-006 ———. 2019. “China from c. 1700.” In The Cambridge World History of Lexicography, edited by John Considine, 317–39. Cambridge: Cambridge University Press. doi:10.1017/9781316827437.016  Lee, Siu-Yau. 2014. “Defining Correctness: The Tale of the Contemporary Chinese Dictionary.” Modern China 40, no. 4: 426–50. doi:10.1177/0097700413491468 Lee Tsz-him 李子謙. 2016. “1930 niándài Zhōngguó ‘dàzhòng yǔ’ yánjiū” 1930年代 中國“大眾語”研究 [A critical study of the “language of the masses” in China in the 1930s]. PhD diss., University of Hong Kong. Li, Chin-Shi. 1922. Chinese Phonetic System and Language. Trans. Alex. R. Mackenzie. Chinese original published in 1919. Shanghai: Commercial Press. ——— 黎錦熙. 1934. Guóyǔ yùndòng shǐgāng 國語運動史綱 [A history of the national language movement. Beijing: Shāngwù yìnshūguǎn. Lǐ Yǔmíng 李宇明. 2017. “Tuīguǎng Pǔtōnghuà: Pǔjí yǔ tígāo bìngxíng” 推广 普通话：普及与提高并行 [The popularization of Pǔtōnghuà: Spread and

120  Henning Klöter enhancement]. In Zhōngguó yǔyán shēnghuó zhuàngkuàng bàogào 2017, edited by the State Language Commission, 3–6. Beijing: Shāngwù yìnshūguǎn. Liao, Chao-Chih. 2000. “Changing Dominant Language Use and Ethnic Equality in Taiwan Since 1987.” International Journal for the Sociology of Language 143, no. 1: 165–82. https://doi.org/10.1515/ijsl.2000.143.165 Liu, Jin. 2013. Signifying the Local: Media Productions Rendered in Local Languages in Mainland China in the New Millennium. Leiden: Brill. doi:10.1163/9789004259027 Masini, Federico. 2017. “Modern Lexicon, Formation.” In Encyclopedia of Chinese Language and Linguistics, vol. 3, edited by Rint Sybesma et al., 65–72. Leiden: Brill. Nakajima Toshio 中島利郎, ed. 2003. 1930 niándài Táiwān xiāngtǔ wénxué lùnzhàn zīliào huìbiān 1930年代台灣鄉土文學論戰資料彙編 [A collection of materials on the Taiwanese xiāngtǔ literature debate in the 1930s]. Kaohsiung: Chūnhuī chūbǎnshè. Ng, Kim Chew. 2013. “Sinophone/Chinese: ‘The South Where Language Is Lost’ and Reinvented.” Trans. Brian Bernards. In Sinophone Studies: A Critical Reader, edited by Shu-mei Shih, Chien-hsin Tsai and Brian Bernards, 74–89. New York: Columbia University Press. Norman, Jerry. 1988. Chinese. Cambridge: Cambridge University Press. Qu, Shaobing. 2015. “Reflections on the Incident of the ‘Preservation of Cantonese’ in Guangzhou.” In The Language Situation in China, edited by Yuming Li and Wei Li, vol. 3, 211–23. Berlin: De Gruyter. doi:10.1515/9781501503146-021 Ramsey, S. Robert. 1987. The Languages of China. Princeton, NJ: Princeton University Press. Rohsenow, John S. 2004. “Fifty Years of Script and Written Language Reform in the P.R.C.: The Genesis of the Language Law of 2001.” In Language Policy in the People’s Republic of China: Theory and Practice since 1949, edited by Minglang Zhou and Hongkai Sun, 21–43. Dordrecht: Kluwer Academic Publishers. doi:10.1007/1-4020-8039-5_2 Simmons, Richard VanNess. 2017. “Whence Came Mandarin? Qīng Guānhuà, the Běijīng Dialect, and the National Language Standard in Early Republican China.” Journal of the American Oriental Society 137, no. 1: 63–88. doi:10.7817/ jameroriesoci.137.1.0063 Snow, Don. 2010. “Diglossia in East Asia.” Journal of Asian Pacific Communication 20, no. 1: 124–51. doi:10.1075/japc.20.1.10sno Tang, Jingchun. 2017. “Xiàndài Hànyǔ cídiǎn 現代漢語詞典.” In Encyclopedia of Chinese Language and Linguistics, vol. 4, edited by Rint Sybesma et al., 608–9. Leiden: Brill. Tiun Hak-khiam 張學謙. 2018. “Cóng guóxiǎo hé guózhōngshēng de Táiyǔ shǐyòng kàn Táiyǔ bǎocún” 從國小和國中生的臺語使用看臺語保存 [Assessing Taiwanese language maintenance from primary and junior high school students’ Taiwanese language use]. Paper presented at the 12th International Symposium on Taiwanese Languages and Teaching (ISTLT-12), Kaohsiung, National Sun Yat-sen University, 26–27 October. Tsao, Feng-fu. 2008. “The Language Planning Situation in Taiwan.” In Language Planning and Policy in Asia, Vol. 1: Japan, Nepal, Taiwan and Chinese Characters, edited by Robert B. Kaplan and Richard B. Baldauf Jr., 237–84. Bristol: Multilingual Matters. doi:10.21832/9781847690968-007

One Legacy, Two Legislations  121 Tsu, Jing. 2010. Sound and Script in Chinese Diaspora. Cambridge, MA: Harvard University Press. doi:10.2307/j.ctvjnrskk Wang, Hui. 2011. The Politics of Imagining Asia. Cambridge, MA: Harvard University Press. doi:10.4159/harvard.9780674061354 Wáng Huī, and Yuán Zhōngruì. 2013. “The Promotion of Putonghua (Mandarin Chinese): An Overview.” In The Language Situation in China, edited by Yuming Li and Wei Li, vol. 1, 27–39. Berlin: De Gruyter. doi:10.1515/9781614512530.27 Wén Yì 文逸. 1934. Yǔwén lùnzhàn de xiàn jiēduàn 語文論戰的現階段 [The current stage of the debate on language and literature]. Shanghai: Shànghǎi shūdiàn. Weng, Jeffrey. 2018. “What is Mandarin? The Social Project of Language Standardization in Early Republican China.” The Journal of Asian Studies 77, no. 3: 611–33. doi:10.1017/s0021911818000487 Wiedenhof, Jeroen. 2015. A Grammar of Mandarin. Amsterdam: John Benjamins. doi:org/10.1075/z.197 Wippermann, Dorothea. 2017a. “Transcription Systems: Zhùyīn fúhào 注音符號.” In Encyclopedia of Chinese Language and Linguistics, vol. 4, edited by Rint Sybesma et al., 436–43. Leiden: Brill. ———. 2017b. “Transcription Systems: Hànyǔ pīnyīn 漢語拼音.” In Encyclopedia of Chinese Language and Linguistics, vol. 4, edited by Rint Sybesma et al., 424–32. Leiden: Brill. Xie, Junying, and Zhangtai Chen. 2015. “Sixty Years’ Work with Spoken and Written Languages.” In The Language Situation in China, edited by Yuming Li and Wei Li, vol. 3, 33–39. Berlin: De Gruyter. doi:10.1515/9781501503146-005 Zhang, Qing. 2016. Language and Social Change in China: Undoing Commonness through Cosmopolitan Mandarin. London: Routledge. doi:10.4324/9781315886251 Zhang, Xiaoling, and Zhenzhi Guo. 2012. “Hegemony and Counter-Hegemony: The Politics of Dialects in TV Programs in China.” Chinese Journal of Communication 5, no. 3: 300–15. doi:10.1080/17544750.2012.701421 Zhao, Shouhui. 2008. “Chinese Character Modernization in the Digital Era: A Historical Perspective.” In Language Planning and Policy in Asia Vol 1: Japan, Nepal and Taiwan and Chinese Characters, edited by Robert B. Kaplan and Richard B. Baldauf, 38–101. Bristol: Multilingual Matters. doi:10.21832/9781847690968-003 Zhou, Minglang. 2000. “Language Policy and Illiteracy in Ethnic Minority Communities in China.” Journal of Multilingual and Multicultural Development 21, no. 2: 129–48. doi:10.1080/01434630008666398 ———. 2018. Language Ideology and Order in Rising China. Singapore: Palgrave Macmillan. doi:10.1007/978-981-13-3483-2

6

Language Policy and Practice in Taiwan in the Early Twenty-First Century Su-Chiao Chen

Abstract Due to the call for globalization and glocalization since the late 1990s, Taiwan has been promoting both English language policy to develop English communicative proficiency and local language policy to foster Taiwan identity, following decades of a “Mandarin-only” strategy. The simultaneous promotion of English, local languages, and Mandarin is aimed at presenting Taiwan as a multilingual place in which people can develop the communicative competence required in a globalized world. This chapter examines the effectiveness of policy implementation in terms of people’s language proficiency and language use in different domains. The results show that government commitment to efficiently implementing the local language policy has been rather weak. In contrast, implementation of the English language policy has been more systematic and extensive, although the frequency of English use is still very low.

Introduction Political democratization and economic globalization are two trends that have strongly influenced the formulation of language policies in Taiwan in the past two decades. The establishment of formal ties between the United States and the People’s Republic of China in 1979 served as a catalyst for the political democratization of Taiwan. This event led to the rise of the Taiwanese Nativization Movement which sought the right to use Taiwan’s ethnic languages and to promote a Taiwanese national identity. This movement grew in the 1980s and resulted in the establishment of the first opposition party in 1986, namely the Democratic Progressive Party (DPP) which was composed predominantly of Holo (i.e. Taiwanese Southern Mǐn) speakers. Another more important consequence is the lifting of the long-imposed martial law in 1987. Taiwan launched a policy of political liberalization and began to advocate the language rights of the people. This is especially true after Lee Teng-hui 李登輝 (b. 1923) became the first Taiwanese-born president in 1988. He actively promoted Taiwanese identity as distinct from Chinese identity. Since then, discussion of ethnic identity, research on Taiwanese culture, and demands for using Taiwanese dialects have flourished

Language Policy and Practice in Taiwan  123 (Chang 1994: 97). The call for a new era of “de-Sinicization of Taiwan,” generally known as “Taiwanization,” became a widespread political ideology, contrasting to the “Sinicization of Taiwan” that had been practiced since the Nationalist Government of the ROC was defeated by communists in Mainland China and fled to Taiwan in 1949 (also see Klöter, this volume). Contrary to the ideology of the Sinicization of Taiwan, which contributed to the formulation and practice of the Mandarin-only National Language Policy and promoted the idea that monolingualism in Mandarin was desirable (Hsiau 1997), the de-Sinicization of Taiwan advocated localization and triggered the Mother Tongue Movement, which called for all the major local languages—i.e. Holo, Hakka, and Austronesian languages, along with Mandarin—to be recognized as national languages. Parallel to the promotion of multilingualism and the right to use Taiwanese ethnic languages was an attempt to advocate English learning at an earlier age in order to improve oral English proficiency and to establish English as a medium of instruction (EMI) in higher education. This “English language movement” developed partly as a result of the impact of economic globalization since the early 1990s, just a little later than the emergence of the Mother Tongue Movement (Chen 2006; 2010). English was considered vital for Taiwan to be able to participate in the global economy and to facilitate the internationalization of its businesses (see also Su, this volume). In order to improve the population’s English proficiency, Taiwan’s existing English language policy was reformulated as a New English Language Policy (Chen 2006). Thus political democratization and economic globalization triggered the implementation of both the local language policy and the English language policy. Language policies, which embody and shape public attitudes toward language (McGroarty 1996), determine the allocation of resources and the direction of language change, which is crucial to multilingual development. However, how and to what extent these two language policies were formulated and implemented and what results they achieved in Taiwan as part of the sinophone world are yet to be investigated. These are the goals of this chapter. I investigate how languages policies in Taiwan reflect the ideologies of languages as perceived by linguistic communities, and what has been the impact on language practices, i.e. the ecology of languages (Spolsky 2004) in this sinophone society. It has been argued that the English language policy was given higher priority and was more effectively implemented than the local language policy by all parties involved in teaching at the elementary-school level. By contrast, the English language policy at the higher education level, where it was to be used as a medium of instruction, was not implemented effectively. This chapter begins with an overview of the language policy that was practiced before the turn of this century. This is followed by an analysis and comparison of the local language policy and the English language policy in terms of planning and implementation in the context of language planning

124  Su-Chiao Chen (Haugen 1983; Weinstein 1980). The discussion includes a description of the planning process and the implementation policy. The latter is then analyzed in terms of the Kaplan and Baldauf’s (2003) language-in-education policy framework in order to examine how each implementation step can reflect existing language values and impact on the effectiveness of the planning.

Overview of Sociolinguistic Background and Language Policy before 2000 Taiwan is geographically composed of one major island and several smaller islets and is separated from Mainland China by the Taiwan Strait. It has been the major remaining territory of the Republic of China (ROC) since 1949, when the Kuomintang (KMT), the then ruling party of the ROC, retreated from Mainland China and relocated to Taiwan. As of 2019, the current population of Taiwan is estimated to be about 23.7 million, based on the latest United Nations estimates (worldometers.info 2019). In the early 1990s, the population was composed of four main ethnolinguistic groups: Holo speakers (73.3 percent), Mainlanders (13 percent), Hakka (12 percent), and Austronesian aborigines (1.7 percent) (Huang 1993:21; see also Wasserfall, this volume). It was noted that the ratios of the four groups remained about the same in the 2000 population census (Chen 2010). The group of Taiwanese new immigrants (新住民) was not included in the 2000 census or in any other official survey, implying that the number of this group is insignificant. However, there has been a very substantial increase in this group since the turn of this century. According to the 2016 Statistical Yearbook of Interior (MOI 2016), the Taiwanese new immigrants (2.85 percent) now outnumber Austronesian language speakers (2.35 percent). In short, Taiwan is a sinophone ecology marked by ethnolinguistic diversity. The ethnolinguistic heritages in Taiwan are the result of four waves of immigration with immigrants taking over the lands of the existing Austronesian populations. Waves of immigration in Taiwan’s history have been paralleled by changes in language policies toward multilingualism (Chen 2010). The first wave of immigration came in the seventeenth century, when Taiwan was colonized by the Dutch in the south (1624–62) and the Spanish in the north (1626–42), followed by the Qing dynasty (1683) until Taiwan was ceded to Japan in 1895. A pragmatic language policy that was neither oppressive nor discriminatory was adopted by the colonists and later by the Qing government. As a large number of Holo speakers poured into Taiwan during this period, Holo became the dominant language of the island, followed by Hakka and Austronesian languages (Tsao 2008a: 240). The second wave of immigration began with the cession of Taiwan to Japan in 1895 and continued until the ROC army arrived from Mainland China and took over Taiwan in 1945 (immediately following the end of World War II). The main immigrants were Japanese. During Japanese colonization, Japanese was made the official language and spread via education.

Language Policy and Practice in Taiwan  125 The language policy implemented was oppressive and discriminatory in regard to Holo, Hakka, and Austronesian local languages. Local languages were considered lower-status. In terms of foreign language learning in this period, not much can be said. This is due to the fact that, during Japanese colonization, the aim of education at all levels in Taiwan was to “Japanize” the Taiwanese, and learning Japanese had priority over learning other languages, including Taiwanese local languages. Although English as a foreign language was offered as an optional subject for Taiwanese starting in secondary schools, only few people could receive secondary education, let alone gain access to an English class (Chen 2003). The third wave of immigration began right after the end of World War II, when Japan ceded sovereignty claims over Taiwan and the ROC government took control of the island in 1945. In 1949, the ROC army retreated from Mainland China and settled in Taiwan. This wave of immigrants, generally known as Mainlanders (lit. ‘extraprovincial person’, 外省人) as opposed to local people, brought an influx of a variety of Chinese dialects as well as the official language, Mandarin. In this period, the ROC government planned to Mandarinize the Taiwanese and thus, in 1946, enforced the implementation of National Language Policy, also known as the National Language Movement (國語運動). Mandarin became the national and official language of Taiwan. Furthermore, a series of actions were taken to discourage or prohibit the use of any language other than Mandarin. For example, in 1948, local languages were declared inappropriate for academic and cultural communication (Cheng 1979: 560). In 1956, Mandarin became the sole language of education; the use of local languages in schools was prohibited; bilingualism was discouraged; and the use of English was restricted (Chen 2003). As a result, Mandarin was effectively promoted as the sole national and official language in Taiwan at the expense of the development of Taiwan’s local languages (Chen 2006). The foreign-language-in-education policy during this period deserves some discussion. After the ROC government had relocated to Taiwan from China, it continued the foreign-language-in-education policy that had been implemented in Mainland China. This policy was in fact a pure Englishlanguage-in-education policy. The aims of English teaching have not changed substantially since the implementation of English instruction in 1912 (Tse 1987: 90). In 1968, English became the only required foreign language subject in middle schools, with the word “English” being parenthesized after the term “foreign language” in official documents (MOE 1971, 1972; cited in Tse 1987: 87). Until 1993, English was the only required foreign language in the first year of university. From then on, other foreign languages were also available, but were never widely taken up. The fourth wave of immigration began just before and after the new millennium. Most of the immigrants came through intermarriage from Mainland China and Southeast Asia, mainly from Vietnam, Indonesia, Thailand,

126  Su-Chiao Chen and the Philippines. The largest group of immigrants originated from Mainland China, followed by those originating from Vietnam and others (www. immigration.gov.tw). Though immigrants in this period were from different countries and spoke a variety of native languages, they are generally categorized as Taiwanese new immigrants. As mentioned earlier, the number of new immigrants now exceeds that of the local Austronesian aborigines. The language policy implemented in this period will be discussed in later sections. In summary, Taiwan has been open to immigrants of diverse cultural and linguistic backgrounds. As a nation of immigrants, Taiwan is linguistically diverse. However, the evidence mentioned earlier shows that the National Language Policy was harshly enforced. Consequently, the primacy of Mandarin has been firmly established, with its adoption as the national and official language, while local languages have been seriously stigmatized and their use reduced (for details, see Klöter 2004 and in this volume).

Local and English Language Planning and Implementation The past two decades have witnessed three challenging international trends: political democratization, economic globalization, and educational reform (Law 2004). Following these international trends, Taiwan’s sociopolitical environment has drastically changed during this period. As mentioned earlier, political democratization led to the call for “Taiwanization” and the formulation of local language policy, whereas economic globalization led to the call for the internationalization of Taiwan and the formulation of the new English language policy. These two policies, as was claimed by Tsao (2008b), were seen as indigenization-driven language planning moving from Mandarin-only to Mandarin-plus, and as internationalization- driven language planning moving from English for the elite to English for all, respectively. The two policies were implemented at the beginning of the 2001 school year. The year 2000 is generally regarded as a natural watershed for Taiwan language policy, as it is the year in which the KMT lost the presidential election after ruling for 55 years and the DPP came to power (Yang 2018). Language policies are commonly declarations of intent while less attention is given to the implementation of policy in practice (Shohamy 2006). Because the existence of language policy does not guarantee that it will be put into practice, a comprehensive study of language policy and planning should include an investigation of both the policy itself and how it is applied in practice. One indicator of the effectiveness of language planning is actual language practice, i.e. the extent to which the languages are used. Thus, in the following section, both planning and implementation are discussed first in regard to the local language policy and then to the English language policy. The two language policies are then compared in terms of how and to what extent those languages are practiced.

Language Policy and Practice in Taiwan  127 Local Language Policy Planning Language recognition has been an important component of democratization projects in many countries, e.g. Canada or South Africa. In Taiwan, however, the local languages were persistently marginalized. All local languages had been designated “dialects” (方言) and their use had declined in the 40 years since Mandarin had been imposed as de facto national language in the Mandarin-only National Language Policy before the lifting of martial law in 1987 (Ang 1992). Following democratization in the late 1980s and 1990s, the then opposition party DPP aggressively advocated “indigenization” (本土化). A turning point was in 1989, when You Ching 尤清 (b. 1942) was elected Taipei County magistrate on an electoral promise to include local language teaching in schools. After that, a series of actions were taken to claim the rights of local languages. Instruction in local languages was first included in the curricula of the elementary schools of Yilan County, followed by those in non-KMT-dominated counties. The Ministry of Education (MOE) then announced in 1998 that one period (40 minutes) of Native Culture Instruction (鄉土文化教學) should be offered nationwide to students from the third to the sixth grade subject to teacher availability. This class instruction focused on the development of students’ awareness and knowledge of cultural tradition and native language(s) practiced in their immediate community. Native Culture Instruction provided a solid foundation for the discussion of local language learning in schools in 2000 when the DPP gained the presidency. This led to the formulation of the local language policy in response to the emerging of a “Taiwan consciousness” (台灣意識) ideology and rallies for vitalizing local and ethnic identity and winning the right to use local languages. After heated discussion, all the mother tongues in Taiwan were formally recognized as local languages. All elementary school children in Taiwan have been required to study at least one local language at school since the beginning of the 2001 school year. At least one period of local language instruction was to be offered in elementary schools each week. These classes are known officially as “local (or vernacular) language education” (鄉土語言教育) and are generally referred to as “mother tongue education” (母語教育) by the public. The local language policy is clearly a bottom-up policy, initiated by a grassroots movement with a strong conviction of the need to halt the decline of local languages. This policy justified the need for governmental intervention in a proactive fashion. It contributed to changing Taiwan’s language policy from “Mandarin-only” to “Mandarin-plus” (P. Chen 2001; cited in Scott and Tiun 2007: 58). However, this local language policy has not yet accomplished its goal, as Mandarin remains the dominant language, and the use of local languages in most domains is still limited (Chen 2018: 97–98).

128  Su-Chiao Chen In 2003, a draft of a National Language Development Law (國家語言發 展法) was introduced. Instead of approaching the local language issue from the perspective of minority language rights, the Council of Cultural Affairs approached it from the perspective of cultural heritage. This denoted a change in focus from the concept of officialization of local languages to simply preserving the heritage of local languages and cultures (Council for Cultural Affairs 2017). The shift from claiming languages as a right to claiming languages as artifacts showed the political compromise of the DPP government and language revivalists. The functions of local languages were kept to a minimum. As a result, local languages were taught in elementary schools with students being required to choose any one local language course available. In secondary schools, however, although local languages had also been available since 2001 as an elective course, very few students opted for it (Tsao 2011). In 2008, when the KMT returned to power, little moved on Taiwan’s linguistic front (Dupre 2016). Implementation The implementation process was mainly focused on the allocation of resources for local language learning in elementary schools. My analysis is based on the language-in-education policy framework proposed by Kaplan and Baldauf (2003) which comprises access policy, personnel policy, methods and materials policy, and resourcing policy. Access policy refers to the status of the language in the curriculum, personnel policy to teachers’ professional training, and resourcing policy to the allocation of financial resources and facilities. All of them are crucially important for implementing a language policy. Access policy: From the perspective of access policy, it was found that the local language policy did not ensure broad access to all students. According to the curriculum, elementary school students are required to learn one of the local languages offered; the offerings depend on the ethnolinguistic background of the majority of students and the availability of teachers (Chen 2006). This shows that the policy contains an escape clause: access is subject to the availability of teachers (Bamgbose 1991). This escape clause has resulted in some schools simply ignoring the policy. For instance, Chen (2007) found that in Chiayi City, 19 out of the 20 schools teach Holo only, with only one school teaching both Holo and Hakka. The virtual dominance of Holo in the implementation of the local language policy to some extent resulted in the situation where “a movement intended to save a minority language may turn out to be a new form of oppression that threatens other minority languages” (Hsiau 1997: 313). In other words, the status of some local languages in the curriculum is lower than that of others. Personnel policy: The local language policy has been constrained by the lack of trained native language teachers. In the past, no local languages were used as a medium of instruction in schools, nor were they offered as

Language Policy and Practice in Taiwan  129 school subjects. The national government has attempted to increase the pool of teachers qualified to teach local languages at elementary schools in two ways: (1) by allowing universities to set up local language teaching departments or short-term training programs; and (2) by introducing a proficiency test for teacher recruitment and certifying qualified teachers. In 2004, local languages were taught by three types of teachers: (1) full-time teachers qualified to teach any local language; (2) substitute or part-time teachers, certified and uncertified, with a certain level of proficiency in one local language; (3) regular homeroom elementary school teachers assigned to teach a local language as well as other subjects. Of the three types of teachers, substitute or part-time and homeroom teachers dominate language teaching nationwide; and only a small number of well-qualified local language teachers occupy full-time teaching positions (Chen 2006). This situation remains unchanged, as more than 70 percent of local language classes are taught by homeroom teachers, and only a very small percentage of them are full-time and specialized local language teachers (Chan and Chen 2012). This indicates that professional knowledge and competence for local language teaching have not yet been given the priority stated in the policy. Methods and materials policy: The methods and materials employed for the implementation of the local-language-in-education policy were not officially specified. The production of materials was in the hands of largely non-official agencies, such as private or regional publishers, who were reluctant to take responsibility for the use of scripts, including transcription systems, because there was no consensus on these matters (Chen 2006). As teaching materials are involved in corpus planning activities and local languages exist only in oral form, the decision on graphization of local languages or vocabulary expansion requires government sanction. Without the approval of the appropriate authority, the corpus used in textbooks was left to the recommendation of experts, making language standardization problematic. The specification of teaching methodology also reflects the perceived functions of local languages. In the local language policy, the teaching methods were not formally specified in the guidelines, which led to different views on appropriate methods. The teaching methods chosen depend on individual teachers or school administrators. Some choose to teach listening and speaking but not writing, claiming that local languages should be learned mainly for the purpose of communication with older generations rather than for developing literacy in the language (Chen 2006). Others focus on teaching reading skills, but, generally speaking, very few teachers focus on the development of students’ literacy (Tsao 2011). The inconsistency in teaching methods and procedures can become a serious challenge to the effectiveness of teaching and the nationwide spread of local language education. In short, the production of pedagogically sound and well-designed textbooks for local language literacy remains problematic.

130  Su-Chiao Chen Resourcing policy: The resourcing policy refers to available resources such as teaching hours and financial aid to support a language education program. Based on the curriculum of local language learning, students from first grade through sixth grade have been required to learn one of the local languages for at least one period (40 minutes) per week since 2001. However, it was found that most schools chose to meet only this minimal requirement (Chen 2004, 2006). In addition, some 55 percent of the teachers report that they have encountered difficulties, e.g. a lack of teacher training classes, a lack of multimedia facilities for teaching local languages, and a failure to coordinate with other subjects in the curriculum (Chen 2006). Although it is hard to ask for more teaching hours or other facilities, schools are allowed to apply for funding by proposing projects to provide activities, e.g. summer camps related to the learning of local languages. One typical funding project is proposed to observe “Taiwan Mother Language Day” (台灣母語日), which was created by the Taiwanese Culture and Education Foundation (台灣文教基金會) in 2008, in order to celebrate International Mother Language Day on February 21. On Taiwan’s Mother Language Day, some activities, such as story-telling in local languages, were held but were criticized as superficial in contents and sometimes ineffective for local language learning (Tsao 2011). In short, although extra funding is available, on the whole the implementation mechanism for local languages needs to be more thoroughgoing. English Language Policy English Language Teaching in Elementary Schools The planning of English teaching in elementary schools was motivated by political democratization and economic globalization in the 1990s, which led to local governments gaining greater autonomy in deciding on education policy. Educational reform for English teaching was one of the outcomes (Law 2004). Taiwan entered an era of “decentralized language planning,” a term originally used by Tollefson (1981). Beginning in 1993, in response to calls for the internationalization of Taiwan, the Taipei City government allowed elementary schools to offer English as an optional course with flexible hours for fifth grade students. In 1998, the School Board of Taipei further decided to carry out citywide implementation from the third grade up. Taipei’s experience was then taken as a model for English teaching, planning, and implementation by other local governments, although some local governments (e.g. Tainan City) chose to implement the teaching of English to students who are one grade lower than those in Taipei and yet others (e.g. Hualien County) chose not to implement at all (Chen 2003). With no centralized language policy, it was found in 1998 that nearly half of Taiwan’s elementary schools offered English lessons on a voluntary basis only (Chou 2003). The decentralized English language teaching policy in elementary schools has resulted in a greater inequality of education opportunities for students

Language Policy and Practice in Taiwan  131 in different school districts, notably for students in urban and rural districts. Due to concerns that children may not be provided with an equal opportunity to receive quality English education, congressmen and citizens all exerted pressure on the central government to centralize the English language policy by introducing national standards. The MOE announced the implementation of a national policy, effective from the 2001 school year, that English as a subject must be included in the regular curriculum of elementary schools. Students at grade five and above were required to have at least one 40-minute period of English each week, but the final implementation of English language teaching at this level is decided upon by the local government. As a result, even after the English language policy had been formally set out by the MOE, the decentralized planning situation remained unchanged. The local governments structured their own English education framework with reference to the MOE’s suggestion as the minimal standard. It was found in 2003, for instance, that 70 percent of elementary schools were already teaching English in the first or second grades (Chou 2003). Many schools were found to be offering more hours of English teaching than required (Chen 2001). In response to this interest in learning English, the central government revised its policy in 2005 and extended the amount of English instruction, requiring third graders and above to attend two 40-minute periods of English per week. Some local governments even introduced English at lower grade levels or offered more hours, for instance, Taipei City, Tainan City, and Hsinchu City (Chen 2012). English as a Medium of Instruction in Higher Education As the pressure to compete internationally mounted, the attainment of global recognition has become one of the major benchmarks in evaluating university performance (Mok 2003). As elsewhere in the sinophone world— e.g. Hong Kong and Macao (Li and Tong, this volume)—Taiwan began to actively promote internationalization of universities under the “Promotion of Colleges’ Internationalization Scheme” (推動大學校院國際化), which was part of “Challenge 2008” (挑戰 2008), a six-year National Development Plan (國家發展重點計畫) (NDC 2005). The internationalization of universities had been proposed by the MOE in 2002, when Taiwan became a member of the World Trade Organization (WTO). The General Agreement on Trade in Services (GATS), which stipulated that educational services should be accessible to other WTO member countries, compelled the MOE to launch initiatives to internationalize higher education (Huang 2014). The internationalization of universities was further proclaimed in 2004, when the focus was on accomplishing global competitiveness and admitting more international students. Another proclamation followed in 2008, with the MOE issuing subsidy guidelines for international cooperation and projects to improve students’ foreign language proficiency at technical colleges/universities. The MOE launched a similar incentive program in 2011 to enhance

132  Su-Chiao Chen English-taught degree programs in institutions of higher education. In short, the international competitiveness of higher education institutions has been promoted robustly from 2003 to 2018 through variously named projects, e.g. Challenge 2008 from 2003 to 2008, the Top University and Excellent Research Center Project (邁向大學頂尖計畫) and the Program for Promoting Teaching Excellence in Universities (大學教學卓越計畫) from 2008 to 2018. The current project, known as the Higher Education Sprout Project (高教 深耕計畫), which will run from 2018 through 2023, also aims among other things to boost international competitiveness. In practice, the EMI curriculum can be divided into three levels: (1) campuswide: Almost all the courses on campus are taught in English; (2) programwide: A program is designed within which all courses are taught in English; (3) individual-EMI: EMI is offered by individual teachers in selected courses only (Huang 2011: 41). In reality, it was found that there is no campuswide EMI in Taiwan. Program-wide EMI is basically a program dominated by international students and is becoming increasingly prevalent. The individual-EMI is the most common type; encouraged by the government, instructors who offer EMI courses can earn various types of financial rewards through a range of support programs. Internationalization can be achieved through various other strategies, including encouraging students to study abroad, recruiting foreign students, increasing joint-degree programs, and strengthening cooperation with overseas universities (Kuo 2016). All of these strategies are geared to the need to develop English language proficiency. The recruitment of international students continues to be given top priority. For example, the MOE aims to attract between 130,000 and 140,000 foreign students by 2020, which means they would account for 10 percent of all students in higher education in Taiwan (MOE 2012). In order to attract students from the global market, many universities are encouraged to improve their global ranking. Accordingly, one of the first things to do is to offer EMI classes, which has been promoted as the mechanism through which universities can achieve internationalization. EMI, also called academic internationalization (Coleman 2006), has thus been popular in Taiwan as a way of enhancing students’ English language ability, while promoting the internationalization of the local education system in order to recruit foreign students, and to increase the competitiveness of Taiwanese educational institutes (Chen and Tsai 2012). Similar to the local language policy, this English language planning for higher education is a decentralized policy with different universities receiving funding for different projects, and thus with planning conducted locally. Implementation of the English Language Policy In this section both the implementation of English language teaching in elementary schools and of the use of EMI in higher education will be discussed with respect to access policy, personnel policy, methods and material policy, and resourcing policy.

Language Policy and Practice in Taiwan  133 Access policy: The English language policy for elementary schools, beginning in the school year of 2001, introduced mandatory English classes for students from grade five upward. This was lowered to grade three in 2005. The policy gives learners considerable exposure to English through instruction. As English is made accessible to all students, learning the language has become popular and is seen to bring personal benefits to students. In other words, the analysis of the mechanisms of access policy shows that government agencies are attaching greater importance to English than to the local languages in the early twenty-first century. On the other hand, the access policy in higher education is inadequate, and EMI is not mandatory. Nevertheless, the MOE encourages professors to offer EMI courses in return for various types of financial rewards, depending on financial resources. In some universities, those professors who offer an EMI course in their classes can earn 50 percent extra in terms of hourly salary. In other universities, as of 2010, professors could earn bonuses ranging from 10,000 to 30,000 Taiwan dollars per semester (Chen 2010). However, only those universities which are awarded the projects mentioned earlier can afford to pay extra money for those professors who offer EMI courses. In other words, EMI courses are basically available only at those universities which are sufficiently competitive to recruit international students or those which can raise extra funding from the MOE. This implies that EMI is not necessarily a de facto policy, but a de jure policy. The social prestige and economic benefits associated with English make EMI courses highly popular, but it is still controversial if EMI is an effective learning approach (Huang 2014). Personnel policy: The qualifications of English teachers in elementary schools can be divided into the following three levels: (1) trained, full-time English teachers with elementary English teaching certificates, most of whom have majored in TESL or in a similar field; (2) substitute or part-time teachers who may be trained or proficient in English to a certain degree but have no official elementary teaching certificate or are not offered a full-time position; and (3) regular homeroom teachers who have not received special training in English teaching but take such positions for other reasons (Chen 2006). There are currently sufficient qualified English teachers in elementary schools due to the fact that many English teaching programs have been established and many English teachers have been trained. Teachers in higher education are all qualified to teach a certain subject. Their professional training in that subject is without question. However, in EMI classes, teachers are required to use English to converse, to lecture, and to interact with students. This is a problem for many Taiwanese teachers (Huang 2014). Therefore, the MOE has sponsored a considerable amount of in-service training through Teaching and Learning Resources Centers established in some top universities or through the MOE’s regional Teaching and Learning Resource Centers. In-service training is mainly offered to train teachers to teach in EMI courses or programs, and is often held in summer (NAER 2013). The MOE also sponsors teachers to go abroad to

134  Su-Chiao Chen develop their English-only teaching skills. All of these training programs available are offered on a voluntary basis, but they are not mandatory for EMI teachers. Methods and materials policy: In the English curriculum of elementary schools, the government has set out clear guidelines for teaching methods and learning materials. The guidelines for teaching methods include: (1) the adoption of a communicative language teaching approach and the use of phonics rather than phonetic symbols (Yeh and Shih 2000); and (2) the use of English as the medium of instruction in order to increase exposure to the language and improve English proficiency. Oral proficiency should be developed for international communication. With regard to textbooks, the government has identified basic words for textbook compilers and teachers, but the adoption of a particular variety of English is not specified. Therefore, the English language policy has opened up the textbook market and resulted in different sets of textbooks and varieties of English. It was found that the textbooks were dominated by the use of American English, while the use of other varieties of English is very limited. All the English textbooks have to be approved by the National Institute of Compilation and Translation (國立編譯館) before publication. Teachers have the flexibility to choose from a wide selection of approved textbooks based on the needs of their students and available resources. Resourcing policy: In contrast to the resourcing policy of local language education, English language education has been allotted more resources from the government. Financial and other resources are provided through many different professional groups or government agencies. For example, in Taipei, the Administrative Planning Group helps English teachers to equip classrooms with hardware facilities and to purchase children’s picture books for supplementary reading. The Teacher Training and Consulting Group holds various types of in-service workshops, and the Curriculum and Instruction Group helps develop diagnostic tools for evaluating different versions of English teaching materials and holds teaching demonstrations and seminars (Chen and Tsai 2012). In addition, students are required to take English for at least two periods of 40 minutes per week beginning in 2001. Most schools meet or exceed the minimal requirements (Chen 2006). The comparison between the resourcing policy of local language education and that of English education shows that local language teaching is not given the same priority as English language teaching in the school curriculum.

Language Practices Language policy has different purposes; one is to maintain existing language practice, and the other is to bring about a major change (Spolsky and Shohamy 1999). An analysis of language policy should not be limited to the examination of official and declared statements; it should also examine

Language Policy and Practice in Taiwan  135 the de facto and concrete language policies, i.e. language practices. Language practices, defined as “the observable behaviors and choices” (Spolsky 2004: 4), can reflect the extent to which the policy has been effective. Effectiveness can be measured in terms of language use and language attitudes. The languages considered here are English, Mandarin, and local languages, including Holo, Hakka, and Austronesian languages. All of the data were collected through questionnaires in 2003 and 2013, and all the answers were self-rated. The comparison of language use between 2003 and 2013 is based on data presented in Chen (2010) and shows the extent to which language policy has influenced language use patterns in the early twenty-first century. Attitudes toward different languages are compared based on the data collected in 2013. The investigation reveals the extent to which people actually use these languages and the value attached to them in the early twenty-first century. The results can presage the changing patterns of language behaviors as a result of official language planning. Language Use Statistics on language use were collected using survey data. For each of the survey questions on language use, subjects were given three choices: frequent, occasional, and rare/never. In order to compare the frequency of language use in different domains, those data of frequent use were further analyzed in terms of two criteria: the frequency of overall use and the frequency of use in the domains family, friendship networks, religion, work/ school, and government. The comparison of language use in 2003 and 2013 shows that within the ten years of language planning, the overall use of English remained almost the same, while local language use increased from 31.38 to 35.37 percent. Although no deliberate efforts were made to promote Mandarin, the use of Mandarin increased more than that of other languages, namely from 81.44 to 87 percent. If we look at the domains of language use (Table 6.1), it was found that Mandarin dominates in all five, even in the family domain (home), a place where local languages were traditionally spoken. The high frequency of Mandarin use is followed by the use of local languages in all domains, but the gaps are very large. Less than half the population speaks Holo in the family and less than one-fifth of the population uses it in the official school/ government domain. A considerable imbalance exists therefore between the two language varieties even after the local language policy was implemented. With regard to English use, the results suggest that English is frequently used in the five domains by only 1 to 4 percent of the population, a much lower share than that of local languages. This corresponds with Ho’s (1998) study, which claims that English has no place in people’s daily lives because Mandarin fulfills all language functions in Taiwan. Nevertheless, the use of English increased in schools/government from 1.87 to 3.0 percent, indicating that the English language policy implemented in schools has had

136  Su-Chiao Chen Table 6.1 Frequency of language use in different domains Language

English

Mandarin

Local Language

Year

2003

2013

2003

2013

2003

2013

2.74 2.05 2.17 1.85 3.83

1.23 0.8 2.39 3.0 3.77

78.22 71.06 85.13 89.15 83.65

87.94 76.22 90.53 93.23 86.56

45.7 36.22 29.61 16.82 28.54

44.01 40.55 31.31 21.62 39.38

Domain Home Religion Friendship School/Govt. Work

an effect in the school domain. The low frequency of English use in Taiwan indicates that English is not used widely to fulfill pragmatic functions in daily life. This does not correspond with the very favorable attitude toward English in Taiwan, which will be discussed in the following section. Language Attitudes Attitude reflects the perceived value attributed to languages or language varieties by community members. In this study, respondents were asked to express their perception of the functional value of three language categories, namely, English, Mandarin, and local languages. These functional values were: (1) symbol of solidarity, (2) symbol of ethnicity, (3) mark of social status, (4) tool for global outlook, (5) tool for upward mobility, (6) tool for learning, (7) tool for communication. The results of the study are presented in Table 6.2. Table 6.2 shows that people in Taiwan have the most positive attitude toward English in terms of its function as a mark of social status, a tool for global outlook, and a tool for upward mobility, all of which are rated higher than for Mandarin (see Su, this volume). Similar to that in Hong Kong (see Li and Tong, this volume), the functional value of English is instrumental in facilitating upward and outward mobility. On the other hand, the most positive attitudes toward Mandarin were for its function as a symbol of solidarity, a symbol of ethnicity, a tool for learning, and a tool for communication. This analysis shows that Mandarin still dominates in all functions except those related to globalization and prestige. The local languages do not dominate in any language function, but are perceived as being more positive as symbols of ethnicity and group solidarity. This indicates that Taiwanese people perceive English to have greater instrumental value than Mandarin. The perceived value associated with English exceeds the actual role it plays. This pattern was not found for Mandarin or local languages. While it was shown that the shift toward English could be motivated by the perceived instrumental functions of the language, Mandarin is perceived to

Language Policy and Practice in Taiwan  137 Table 6.2 Attitudes toward English, Mandarin, and local languages Language

English

Mandarin

Local Language

4.59 3.91 58.26 70.34 71.26 27.03 4.63

49.72 55.65 35.95 27.61 6.30 68.43 85.81

42.11 33.43 6.57 3.66 4.994 6.20 11.82

Perceived value Symbol of solidarity Symbol of ethnicity Mark of social status Tool for global outlook Tool for upward mobility Tool for learning Tool for communication

be more suited to fulfilling integrative and communicative functions. Local languages have lost ground in all functional values.

Conclusions Political democratization and economic globalization prompted Taiwan to implement both a local language policy and an English language policy at the turn of this century. While the former focuses on the promotion of local language learning in elementary schools, the latter emphasizes the promotion of English learning in elementary schools and the use of EMI in higher education. The goals of the local language policy are to preserve indigenous languages and cultures. However, the local language access policies, personnel policies, method and materials policies, and resourcing policies reflect a weak commitment of the government to effectively implement those policies. At the same time, in light of socioeconomic changes, calls to internationalize Taiwan have made the lack of oral English proficiency and the underuse of English in higher education an issue of public concern. In order to resolve this problem, the government has committed to systematically planning and implementing the English language policy. This has contributed to the phenomenon that English education has recently been dubbed a “national movement” in Taiwan, driving the country to become more competitive in the globalized world (Chen and Tsai 2012). This phenomenon is observed in other Asian countries which see English as the key to survival in this global era. They, too, have started teaching English at elementary schools, lowering the starting age at which English is introduced (Spolsky and Moon 2012: xv). In short, the analysis shows that the Taiwanese government’s attitude toward local languages is neutral, but more positive toward the English language. However, people’s actual language practices are not necessarily consistent with language planning. By comparing language use frequency in 2003 with that in 2013, respondents’ self-rated language use frequency showed an increase in the overall use of Mandarin and local languages in all domains.

138  Su-Chiao Chen Mandarin remains dominant in all domains despite the implementation of local language and English language policies. The use rates of English increased mainly in the school/government domain, but not overall. This implies that the implementation of both local and English language policies has been moderately successful, but also that the government does not seem to value the learning of local languages as highly as that of English. While the English language policy has had a positive influence on people’s English proficiency and English use frequency in some domains, the level of proficiency in English remains weak in general and the overall English use frequency remains very low. This result shows that patterns of language use have only changed at a very moderate rate. The implication is that the pattern of change in language use and language proficiency has inevitably been influenced by Taiwan’s social, political, and linguistic environments although the language policy was formulated within the larger global context. The results also suggest that Taiwan’s society is not sufficiently internationalized to have opportunities to engage in frequent global contacts in daily life, which, in turn, may hinder the development of the population’s English proficiency. This corresponds with the people’s attitude that Mandarin rather than English should be used as a tool of learning. This will limit the extent to which the internationalization of higher education can be achieved. Therefore, in order to attain a higher level of internationalization, other aspects, social, political, and economic, need to be integrated within a globalization framework.

References Ang Ui-jin 洪惟仁. 1992. Táiwān yǔyán wéijī 台灣語言危機 [Taiwan’s language crisis]. Taipei: Qiánwèi chūbǎnshè. Bamgbose, Ayo. 1991. Language and the Nation: The Language Question in SubSaharan Africa. Edinburgh: Edinburgh University Press. doi:10.1075/lplp.19.1.10bre Chan Hui-Chen 詹惠珍, and Chen Su-Chiao 陳淑嬌. 2012. “Táiwān dìqū guómín zhōng-xiǎoxué běntǔ yǔyán jiàoxué xiànkuàng zhī zhěnghéxíng yánjiū, zǐ jì huà yī, èr: Táiwān Mǐnnányǔ kèchéng shíshī xiànkuàng yánjiū qīmò bàogào” 臺灣地區 國民中小學本土語言教學現況之整合型研究, 子計畫一、二：臺灣閩南語課程 實施現況研究期末報告 [Final report of the study on the local language teaching situation in nine-year integrated education project, subprojects 1 and 2: A study of the Taiwan Southern Mǐn teaching situation]. New Taipei: National Academy for Educational Research. Accessed February 11, 2020 via https://www.naer.edu.tw. Chang, Mau-Kuei. 1994. “Toward an Understanding of the Sheng-chi Wen-ti in Taiwan: Focusing on Changes after Political Liberalization.” In Ethnicity in Taiwan: Social, Historical, and Cultural Perspectives, edited by Chen Chung-Min, Chuang Ying-Chang, and Shu Min Huang, 93–150. Taipei: Institute of Ethnology, Academia Sinica. Chen, Chinfen. 2012. “Planning and Implementation of Elementary School English Education in Taiwan.” In Primary School English-Language Education in Asia, edited by Bernand Spolsky and Young-in Moon, 129–43. New York: Routledge.

Language Policy and Practice in Taiwan  139 Chen, Ping. 2001. “Policy on the Selection and Implementation of a Standard Language as a Source of Conflict in Taiwan.” In Language Planning and Language Policy: East Asian Perspectives, edited by Nanette Gottlieb and Ping Chen, 75–94. Richmond, Surrey: Curzon Press. ———. 2010. “EGP or ESP: Rethinking College English Instruction.” NCUE Studies in Language, Literature, Translation, and Interpretation 6: 119–40. Chen Su-Chiao 陳淑嬌. 2001. “Xīn Táiwān yǔyán zhèngcè yǔ yǔyán bù píngděng zhī yánjiū” 新台灣語言政策與語言不平等之研究 [A study of new language policies and linguistic inequality in Taiwan]. In Bāshíjiǔ nián guókēhuì yǔyán xuémén “yībān yǔyánxué” yánjiū chéngguǒ fābiǎohuì lùnwénjí, edited by National Science Council, 401–29. Taipei: National Taiwan University. ———. 2003. The Spread of English in Taiwan: Changing Uses and Shifting Attitudes. Taipei: Crane. doi:10.1075/lplp.31.1.09hai ———. 2006. “Simultaneous Promotion of Indigenisation and Internationalisation: New Language-in-Education Policy in Taiwan.” Language and Education 20, no. 4: 322–37. doi:10.2167/le632.0 ———. 2007. “Táiwān yǔyán huólì yánjiū” 台灣語言活力研究 [Linguistic vitality in Taiwan]. In Multicultural Thinking of Language Policy, edited by Cheng ChinChuan et al., 19–39. Taipei: Academia Sinica, Institute of Linguistics. ———. 2010. “Multilingualism in Taiwan.” International Journal of the Sociology of Language 205: 79–104. doi:10.1515/ijsl.2010.040 ———. 2018. “Patterns of Language Use in Taiwan in the Age of Globalization.” Modern Linguistics 6, no. 1: 93–105. doi:10.12677/ml.2018.61011 ———, and Ya-Ching Tsai. 2012. “Research on English Teaching and Learning: Taiwan (2004–2009).” Language Teaching 45, no. 2: 180–201. doi:10.1017/ s0261444811000577 Cheng, Robert L. 1979. “Language Unification in Taiwan: Present and Future.” In Language and Society: Anthropological Issues, edited by William C. McCormack and Stephen A. Wurm, 541–78. The Hague: Mouton. doi:10.1515/9783110806489.541 Chou, Cathy. 2003. “Schools Give Early English Lessons.” China Post. Accessed July 28, 2015 via http://www.chinapost.com.tw. Coleman, James A. 2006. “English-Medium Teaching in European Higher Education.” Language Teaching 39, no. 1: 1–14. Council for Cultural Affairs. 2017. “Announcement of National Language Development Law 國家語言發展法公告.” Taipei: Legislative Yuan. Accessed February 11, 2020 at https://www.moc.gov.tw/information_253_65745.html. Dupre, Jean-François. 2016. “Legislating Language in Taiwan: From Equality to Development to Status Quo.” Language Policy 15, no. 4: 415–32. doi:10.1007/ s10993-015-9376-9 Haugen, Einar. 1983. “The Implementation of Corpus Planning.” In Progress in Language Planning, edited by Juan Cobarrubias and Joshua A. Fishman, 269–89. Berlin: Mouton Publishers. Hsiau, A-Chin. 1997. “Language Ideology in Taiwan: The KMT’s Language Policy, the Tai-Yu Language Movement, and Ethnic Politics.” Journal of Multilingual and Multicultural Development 18, no. 4: 302–15. doi:10.1080/01434639708666322 Huang Shuanfan 黃宣範. 1993. Yǔyán, shèhuì yǔ zúqún yìshí: Táiwān yǔyán shèhuìxué de yánjiū 語言、社會與族群意識 ：台灣語言社會學的研究 [Language, society, and ethnic identity: A sociolinguistic study of Taiwan]. Taipei: Crane.

140  Su-Chiao Chen ———. 2000. “Language, Identity, and Conflict: A Taiwanese Study.” International Journal of the Sociology of Language 143: 139–50. doi:10.1515/ijsl.2000.143.139 Huang, Yi-Ping 2011. “English-Medium Instruction (EMI) Content-Area Teachers’ (CATs’) Pedagogical Content Knowledge of Scaffoldings: A Vygotskian Perspective.” Taiwan Journal of TESOL 8, no. 1: 35–66. ———. 2014. “Teaching Content via English: A Qualitative Case Study of Taiwanese University Instructors’ Instruction.” Foreign Language Studies 20: 27–62. Kaplan, Robert B., and Richard B. Baldauf Jr. 2003. Language and Language-inEducation Planning in the Pacific Basin. Dordrecht: Kluwer Academic Publishers. doi:10.1007/978-94-017-0145-7 Klöter, Henning. 2004. “Language Policy in the KMT and DPP Eras.” China Perspectives 56: 1–12. doi:10.4000/chinaperspectives.442 Kuo, Yu-Ying. 2016. “Taiwan Universities: Where to Go?” Humanities 5, no. 12: 1–13. doi:10.3390/h5010012 Law, Wing-Wah. 2004. “Translating Globalization and Democratization into Local Policy: Educational Reform in Hong Kong and Taiwan.” International Review of Education 5: 497–524. doi:10.1007/s11159-004-4637-4 McGroarty, Mary. 1996. “Language Attitudes, Motivation and Standards.” In Sociolinguistics and Language Teaching, edited by Sandra Lee McKay and Nancy H. Hornberger, 3–46. Cambridge: Cambridge University Press. doi:10.1017/ cbo9780511551185.004 MOE (Ministry of Education, Republic of China). 1971. Gāojí zhōngxué kèchéng biāozhǔn 高級中學課程標準 [Senior middle school curriculum standards]. Taipei: Cheng Chung Bookstore. ———. 1972. Guómín zhōngxué kèchéng biāozhǔn 國民中學課程標準 [Junior middle school curriculum standards]. Taipei: Cheng Chung Bookstore. ———. 2012. Higher Education in Taiwan 2012–2013. Taipei: Ministry of Education. Accessed February 11, 2020 via https://ws.moe.edu.tw. MOI (Ministry of the Interior, Republic of China). 2016. Nèizhèngbù tǒngjì niánbào 內政部統計年報 [Statistical yearbook of the ministry of the interior]. Accessed on April 8, 2019 via www.moi.gov.tw. Mok, Ka-Ho. 2003. “Globalisation and Higher Education Restructuring in Hong Kong, Taiwan, and Mainland China.” Higher Education Research & Development 22, no. 2: 117–29. doi:10.4324/9780203882221-21 NAER (National Academy for Educational Research), ed. 2013. Zhōnghuá Mínguó jiàoyù niánbào 102 中華民國教育年報102 [The republic of China education yearbook 2013]. Taipei. Accessed February 11, 2020 via https://www.naer.edu.tw. NDC (National Development Council), ed. 2005. Tiǎozhàn 2008: Guójiā fāzhǎn zhòngdiǎn jìhuà 挑戰 2008：國家發展重點計畫 [Challenge 2008 national development plan]. Taipei. Accessed February 11, 2020 via www.ndc.gov.tw. Scott, Mandy, and Tiun Hak-khiam. 2007. “Mandarin-Only to Mandarin-Plus: Taiwan.” Language Policy 6: 53–72. doi:10.1007/s10993-006-9040-5 Shohamy, Elana. 2006. Language Policy: Hidden Agendas and New Approaches. London: Routledge. doi:10.4324/9780203387962 Spolsky, Bernard. 2004. Language Policy. Cambridge: Cambridge University Press. doi:10.1017/S004740450808007X ———, and Elana Shohamy. 1999. The Languages of Israel: Policy, Ideology and Practice. Clevedon: Multilingual Matters. doi:10.1017/S004740450124305X

Language Policy and Practice in Taiwan  141 ———, and Young-In Moon, eds. 2012. Primary School English-Language Education in Asia: From Policy to Practice. London: Routledge. Tollefson, James W. 1981. “Centralized and Decentralized Language Planning.” Language Problems and Language Planning 5, no. 2: 175–88. doi:10.1075/ lplp.5.2.04tol Tsao, Feng-Fu. 2008a. “The Language Planning Situation in Taiwan.” In Language Planning and Policy in Asia, Vol. 1: Japan, Nepal and Taiwan and Chinese Characters, edited by Robert B. Kaplan and Richard B. Baldauf Jr., 237–84. Bristol: Multilingual Matters. doi:10.21832/9781847690968-008 ———. 2008b. “The Language Planning Situation in Taiwan: An Update.” In Language Planning and Policy in Asia, Vol. 1: Japan, Nepal and Taiwan and Chinese Characters, edited by Robert B. Kaplan and Richard B. Baldauf Jr., 285–300. Bristol: Multilingual Matters. doi:10.21832/9781847690968-008 ——— 曹逢甫. 2011. “Táiwān dìqū guómín zhōng-xiǎoxué běntǔ yǔyán jiàoxué xiànkuàng” 臺灣地區國民中小學本土語言教學現況 [The current situation of local language education at secondary schools in Taiwan]. Taipei: National Academy for Educational Research. Accessed February 11, 2020 via https://www.naer. edu.tw. Tse, John Kwock-Ping. 1987. Language Planning and English as a Foreign Language in Middle School Education in the Republic of China. Taipei: Crane. Weinstein, Brian. 1980. “Language Planning in Francophone Africa.” Language Problems and Language Planning 4, no. 1: 55–77. doi:10.1075/lplp.4.1.04wei Worldometers.info. 2019. Taiwan Population. Dover (Delaware, U.S.A.). Accessed April 23, 2019 via www.worldometers.info/. Yang Chen-yi 楊真宜. 2018. “Táiwān guózhōngshēng duì guóyǔ, běntǔ yǔyán yǔ yīngyǔ de yǔyán nénglì, shǐyòng yǔ tàidù zhī tànjiù” 台灣國中生對國語、本土 語言與英語的語言能力、使用與態度之探究 [A study of Taiwan junior high students’ language proficiency, use, and attitude: Mandarin Chinese, indigenous languages, and English]. PhD diss., National Tsing Hua University (Taiwan). Yeh, Hsi-Nan, and Shih Yu-Hwei. 2000. “A New Era for Foreign Language Education in Taiwan: English Teaching in the Elementary Schools.” In Proceedings of Conference on ELT Curriculum for Young Learners in East Asia, 85–97. Taipei: National Taiwan Normal University.

7

A Tale of Two Special Administrative Regions The State of Multilingualism in Hong Kong and Macao David C.S. Li and Choi-Lan Tong Abstract This chapter gives an overview of language diversity in Hong Kong and Macao. Both places enjoy a high level of sociopolitical autonomy, including the continued use of the former colonial languages—English and Portuguese, respectively—as co-official languages alongside Chinese, which is understood to refer to spoken Cantonese and Mandarinbased Standard Written Chinese (SWC). The language policies in both places aim at achieving biliteracy and trilingualism: the ability to read and write Chinese and English, and to speak and understand Cantonese, English, and Mandarin. Unlike elsewhere in the sinophone world, Cantonese as a regional Sinitic variety continues to be used as the medium of instruction (MoI) from kindergarten to secondary-level schools. Language contact phenomena, such as lexical borrowing from English and code-switching, are commonplace. With cross-border visits to and from mainland China becoming more and more frequent, the use of Mandarin is increasingly frequent in both communities.

Introduction Under the unprecedented “one country, two systems” postcolonial arrangement, Hong Kong and Macao1 were renationalized as Special Administrative Regions (SARs) of the People’s Republic of China in the late 1990s. The two SARs are rather different in terms of their principal economic activities and primary sources of revenue, which, in turn, explain their priorities in the development of human resources, including their respective language-in-education policies to meet local needs for specialist skills. With a population of over 7.4 million inhabiting a land area of about 1,100 square kilometers, Hong Kong ranks among the most densely populated cities in the world. With a much smaller land mass of barely 30.8 square kilometers, Macao is comparatively tiny, and yet in 2019 its population stood at 676,100, including 193,470 or 28.6 percent being non-resident workers (DSEC 2019). In the 1960s and 1970s, the former British colony literally made a name worldwide through its manufactured products that invariably carried the etiquette “Made in Hong Kong.” From the mid-1980s onward, the manufacturing sector was gradually re-located north of the border with Mainland

Tale of Two Special Administrative Regions  143 China to different parts of the Pearl River Delta. Since then, the lifeline of Hong Kong has shifted to a few other sectors which are more characteristic of a knowledge-based economy, the most vibrant of which are banking, investment and finance, imports/exports, telecommunications, transport and logistics, tourism, hotels, restaurants, insurance, wholesale/retail trade, and real estate services. Changing manpower needs and growth areas are actively monitored by the Hong Kong Government. Publicly funded universities, eight at present, are tasked by the University Grants Committee (UGC) to churn out employable graduates to meet these needs (e.g. healthcare workers). Despite being the most international of Chinese metropolises, there is some indication that the self-styled Asia’s World City is slowly losing its edge as the economic prowess it has enjoyed for decades is slowly being undermined by sister cities in the region. According to the Blue Book on Urban Competitiveness released by the Beijing-based Chinese Academy of Social Sciences in mid-May 2015, for the first time since the index was created in 2005, Hong Kong has lost its top spot to Shenzhen, a hi-tech hub which is ranked as the most innovative city nation-wide thanks to six emerging industries: biotechnology, the internet, new energy, new materials, information technology, and cultural and creative industries (the third- to fifth-ranked cities being Shanghai, Taipei and Guangzhou) (People’s Daily Online 2015). The measures of competitiveness are based on multiple performance indicators covering business environment, municipal harmony, efficiency, suitability for living and sustainability, among others. Macao, on the other hand, ranked ninth, rising one position compared to 2014 (He 2015; Lai and Nip 2015). Macao has traditionally relied on its robust gaming and tourism industries, which have undergone considerable expansion since the 1990s. With 35 casinos, thousands of table games and slot machines, Macao’s reputation as a world-class gaming capital may be gauged by the titles it has earned from international travelers and tourists, from the more old-fashioned “Monte Carlo of the Orient” to the more contemporary “Las Vegas of the Far East.” Annual revenues from gambling taxes amounted to over 80 percent at around US$45 billion—seven times the amount generated by the casinos on Las Vegas Strip (Fensom 2015). Since 2013, however, big drops in casino revenues caused a great deal of societal concern and prompted the former Chief Executive of Macao, Fernando Chui 崔世安, to call for the pace of economic diversification to be accelerated (Fensom 2015). According to Macao’s Gaming Inspection and Coordination Bureau, gross gaming revenue growth in 2018 reportedly rebounded to 300 billion patacas (ca. US$ 37 billion). Such an upward trend was reportedly matched by optimism in the performance of casino stocks of “the gambling hub of not just China, but all of Asia Pacific” (Wood 2018). Whether that rebound was linked to increasing numbers of mainland visitors after the opening of the Hong KongZhuhai-Macao Bridge in October 2018 remains to be seen. What is clear is that for economic growth to be sustainable, the Macau SAR government’s priority for future development is a more broadly based economic structure founded upon industrial diversification.

144  David C.S. Li and Choi-Lan Tong

Historical Language Contact in the Two SARs Contact between European languages and Chinese language varieties in Macao is closely related to the history of maritime trading activities and commerce since the arrival of Portuguese traders along the south China coast in the sixteenth century (van Dyke 2005; 2011). In 1557, the Ming emperor granted the Portuguese request to rent a settlement area in Macao in part to reward their active role in keeping pirates off the south China coast. Since then, commercial activities thrived, and Macao gradually developed into a Portuguese outpost for their trading activities in the Far East, especially with China and Japan, at a time when direct trading between these two countries was banned. The local agents—mainly traders and service personnel who spoke Cantonese or a Yuè dialect of Guangdong province as their vernacular—gradually picked up some “broken Portuguese” to do business with and meet the practical needs of Portuguese merchants and sailors. Over time, their broken Portuguese evolved into Macao Portuguese Pidgin (MPP). For about a 100 years until the 1830s, MPP served as the lingua franca within the Chinese-Portuguese trading community, including in Guangzhou (Canton): Macau had a significant impact on the environment in Canton, because much of the trade there was a direct extension of the market upriver. When the Portuguese ships arrived in Macau, Chinese merchants from Canton came downriver to buy their goods. (van Dyke 2005: 143) Based on European travelers’ anecdotal accounts and linguistic evidence such as Chinese-Portuguese glossaries, the earliest being Àomén jìlüè (澳門記略; “Abridged record of Macao,” 1751) and the Compendium of Assorted Phrases in Macau Pidgin (printed in the late nineteenth century, Anonymous, n.d.), Li and Matthews (2016: 143) postulate “a continuum of varieties from Portuguese via Macanese spoken natively to pidgin Portuguese spoken by Chinese traders,” as in Figure 7.1. In general, the creolized variety spoken by the Macanese is also known as Macao Creole Portuguese (MCP); in popular parlance their speakers would call it Patuá, or “Macanese” in English:

Portuguese Macanese Patuá (Pidgin) Macanese < -------------------------------------------------------------------------------------------------- - > as spoken by Portugueseas spoken by the Macau as spoken by the Chinese born or population of mixed internally with Portuguese Portuguese-educated ancestry speakers residents

Figure 7.1 Varieties of Portuguese in Macao. Source: Li and Matthews (2016: 143).

Tale of Two Special Administrative Regions  145 By the mid-nineteenth century, the trading monopoly of the Portuguese was increasingly challenged by the British at sea. After the signing of the Treaty of Nanking at the end of the first Anglo-Chinese War (also known as the First Opium War) in 1842, the cessation of the island of Hong Kong to the British Empire and the forced opening of four treaty ports greatly facilitated trading activities in south China for ships flying the Union Jack (Zhang 2009). Gradually an English-based pidgin—Chinese Pidgin English (CPE, also known as China Coast Pidgin, CCP)—arose under similar circumstances to MPP. With maritime trade in the region gradually gravitating away from Macao toward Hong Kong and elsewhere along the coast in south China, CPE proved to have greater vitality and currency among the agents actively involved in trading activities as well as services provided to English-speaking merchants and sailors. This led to the gradual decline of MPP as a preferred regional lingua franca between non-Chinese business partners and Chinese traders and service personnel. As observed by Li and Matthews (2016: 149), “The diminishing role of Portuguese and pidgin Portuguese in China trade gives rise to CPE as a trade pidgin used from the eighteenth century onwards” (cf. Van Dyke 2005: 77).

Governance in Colonial Hong Kong and Macao Compared with other former colonies like India and elsewhere in the Indian subcontinent, Southeast Asia, and Africa, the colonial history of both Hong Kong and Macao is untypical in many ways. Where plundering, enslavement, and slaughter were the order of the day in many former colonies, such brutalities were uncommon in Hong Kong and Macao which were not as rich in natural resources and which from the outset were intended to be trading outposts for merchants from the West. As Bolton has noted: “In spite of the virtual colonisation of treaty ports by western powers, led by Britain and America, there was no direct equivalent to the ‘Anglicist’ policy promoted by Macaulay’s (1835) Minute of Indian Education” (Bolton 2003: 192). This notwithstanding, in Hong Kong the British colonial government could not afford not to cultivate a class of elite bilinguals to serve as middlemen to facilitate governance, hoping that some of these would somehow take care of the education of the masses conducted in their local vernacular. In colonial Hong Kong, as in other British colonies, the grooming and presence of a small English-educated elite in the colonial government and civil service was instrumental in ensuring that the will of the colonizers was accurately and effectively conveyed to the colonized. The language situation in colonial Macao was very different (Hao 2011; Wú and Chan 2000). The beginning of colonial rule may be traced back to 1846, when Governor Ferreira do Amaral displaced Qing government officials and formally asserted Portuguese jurisdiction in Macao. But it was not until 1887, after the signing of the Sino-Portuguese Treaty of Peking, that the colonial government started exercising Portuguese sovereignty there. For nearly a century,

146  David C.S. Li and Choi-Lan Tong however, Macao was a colony more in name than in practice, because successive governments in Portugal were preoccupied with political instability at home, from the monarchy giving way to the First Republic in 1910 to the struggle against dictatorship in the 1930s through the two World Wars until the Carnation Revolution in 1974. It was only in 1976, after the Organic Statute of Macao was passed and implemented, that legally binding governance of the colony of Macao and the identity, rights, and obligations of its colonial subjects were formally established. For well over a century since 1846, therefore, the Chinese and Portuguese communities in Macao, including the ruling class, lived more or less in harmony, each tending to their own business, including education. Most of the schools were operated by the Catholic churches and Chinese groups, with Chinese being the main target language and MoI, supplemented by the teaching of some subjects in English. These schools were attended by children from more affluent families. Portuguese children and children from less well-off families, on the other hand, would attend government-operated schools. Thanks to its regional lingua franca status, English was widely perceived as being more useful and important than Portuguese. From the mid1970s onward, however, to facilitate colonial governance, a good knowledge of Portuguese was upheld as a requirement for joining the civil service. Beneficiaries of this preferential language policy included native speakers of Portuguese as well as bilingual Macanese, the latter being looked upon as the bridge or nexus between the non-Chinese-speaking rulers from Portugal and the local Chinese population. Unlike Hong Kong, therefore, for historical reasons there was less pressure in Macao to cultivate a group of Portuguese-speaking local elite, not until the mid-1970s, when governance in colonial Macao was increasingly characterized by bilingualism in Chinese and Portuguese. The term Macanese refers to a small ethnolinguistic group with distinctive identity. Born to parents of mixed marriage with Portuguese lineage, they grew up trilingual in Portuguese, Cantonese, and Macanese, the latter being a Portuguese-based creole enriched with vocabulary from Cantonese and substratum influence from other languages brought to Macao such as Malay.

Pidginization and Creolization: Linguistic Features In terms of linguistic features, both MPP and CPE are characterized by radical simplification of the respective European languages and considerable substrate influence from Cantonese. While Portuguese and English were the main lexifier languages, respectively, lexical sources from other languages such as Malay and Indian languages are also identifiable and attested. From the point of view of historical data, the research community of pidgins and creoles is fortunate in that valuable documentation can be found, in English and Chinese, among other European languages (van Dyke 2005). Most of the MPP and CPE data contained in English and Portuguese sources are

Tale of Two Special Administrative Regions  147 mainly anecdotes extracted from the personal narratives of merchants or from travelogues written by adventurers. By contrast, there is a sizable body of data in Chinese consisting of phrase books which were written with the explicit purpose of providing instruction to Cantonese-speaking readers on how to imitate and make sense of MPP or CPE. Just as the linguistic ecology leading to pidginization and the emergence of MPP and CPE was very similar, so the main factors leading to depidginization and its gradual demise were by and large the same. Of these, the key factor was related to the introduction of standard European languages in the education system—English in Hong Kong and Portuguese in Macao. As Bolton (2003) observes regarding the gradual disappearance of CPE: From the late nineteenth century onwards, in Hong Kong, as well as in the other parts of China, the key factor in shifting the acquisition and use of English from a pidgin form towards a more “standard” variety of English was access to instruction, and access to schools where English was taught. With access to the English of the classroom, the process of depidginization could then occur. (Bolton 2003: 191–92) In terms of language contact phenomena, lexical borrowing and translanguaging (i.e. code-switching and code-mixing) are very common among Cantonese-English bilinguals in Hong Kong and Macao (Matthews 2013). In Hong Kong, lexical borrowing from English to Cantonese has been documented since the early 1980s (e.g. kaang1 taa3 ‘counter’ and te1 laa3 ‘teller’ in the banking context). Conversely, quite a number of colloquial Cantonese expressions have also found their way into “Hong Kong English,” mostly nouns such as char siew (variant cha siu: 叉燒 caa1 siu1 ‘barbecue pork’), dai pai dong (大排檔 daai6 pai4 dong3 ‘street cafe’) (Bolton 2003: 212–13; Cummings and Wolf 2011). A similar trend has also been found in Macao, where vernacular Cantonese is characterized by the presence of many loanwords from Portuguese (Sun 2015; Tong 2015). This is especially evident in common nouns denoting food and beverage items, for example, aa3 dung1 jyu2 阿東魚 (from Portuguese átum ‘tuna fish’), guk1 gu2 唂咕 (cacao ‘cocoa’), so1 ba2 梳巴 (sopa ‘soup’), daai6 ma1 di4 大孖弟 (tomate, ‘tomato’), gaa3 fe1 咖啡 (café, ‘coffee’), maa5 gaai3 jau1 馬介休 (bacalhau, ‘Portuguese style salted fish’), and loan blends like but1 zau2 砵酒 (Porto, literally ‘Port wine’). Other borrowings include cultural loans from Portuguese such as laa1 daa2 喇打 (lata, a kind of food container), laang1 唥 (lã, ‘laine wool’), faat3 do1 法多 ( fado, a melancholic Portuguese singing style or genre), saa1 baa4 dou4 沙巴度 (sapato ‘leather shoes’), gam1 baa1 laa1 金巴喇 (câmara ‘city hall’), si1 sa4 司沙 (sisa ‘property transfer tax’), and adeus (‘goodbye’). Different from the tendency of English verbs and adjectives borrowed into Hong Kong Cantonese, more everyday Portuguese verbs and adjectives are commonly mixed into Macao Cantonese, resulting in translanguaging.

148  David C.S. Li and Choi-Lan Tong For example, bom ‘good’, mau ‘bad’, moderno ‘modern’, falar ‘speak’, falta ‘absent’, não tem ‘not have’, pouco ‘a little’, and tudo ‘total’. Whereas English words in Roman script are commonly mixed into informal sections of Hong Kong Chinese print and digital media, Portuguese words rarely appear in the Macao Chinese equivalents, which tend to be more conservative and adhere to Pǔtōnghuà-based SWC, occasionally mixed with vernacular Cantonese elements. Given that individual Portuguese words— in Roman script or Sinicized—tend to be mixed into spoken Cantonese but rarely written, plus the fact that since the handover in December 1999, written norms in Macao Chinese media generally adhere to Pǔtōnghuà-based standards, the visibility of Portuguese loanwords in writing is low compared with its relative vitality in spoken Cantonese of Macao (on loanwords resulting from contact between English and Mandarin, see Wasserfall, this volume).

Demographics, Functions, and Status of Local Languages Speakers of Cantonese make up the absolute majority in Hong Kong and Macao. In both SARs the numbers and percentages of bilingual speakers of Pǔtōnghuà (Mandarin) and English have been gradually on the rise. This is partly due to the steady immigration set for “family reunion” quotas. In Macao, this trend is further accentuated by the daily presence of a large contingent of non-resident, Pǔtōnghuà-dominant cross-border laborers. Concerning English, during the first decade of the new millennium until 2011, both SARs have witnessed a considerable increase in the number of speakers self-reporting an ability to use this international language: an increase of 3.1 percentage points (from 43.0 to 46.1 percent) in Hong Kong and 7.6 (from 13.5 to 21.1 percent) in Macao. As for other Chinese varieties, their speakers in both SARs have been declining in numbers except for Hokkien (i.e. the regional language of Fujian, including Southern Mǐn spoken in Taiwan), which appears to have remained more or less stable during that period. The principal languages of multilingual Hong Kong are neatly captured by the language-in-education policy goal of the HKSAR government, which came to be known as “biliteracy and trilingualism” (兩文三語): the ability to read and write Chinese and English, and to speak and understand Cantonese, English, and Pǔtōnghuà (Li 2009, 2017). The same languages are also widely used in Macau SAR, except that Portuguese and the Portuguese-creole-based Macanese may also be encountered in society. Macanese (or Patuá) has no place in the local curricula, however. Since the colonial era, students in Macao are expected to learn standard Portuguese, a curricular choice that the MSAR government has inherited. According to Young (2009), in the education system of postcolonial Macao, three written languages (Chinese, Portuguese, and English) and four spoken languages (Cantonese, Pǔtōnghuà, Portuguese, and English) are represented.

Tale of Two Special Administrative Regions  149 From the 1960s onward, Cantonese-L1 speakers make up the majority of the population in Hong Kong and Macao, which is why Cantonese has been used as the regional lingua franca since then. Speakers of other Chinese “dialects” have always found it useful to learn at least some Cantonese, which is the most prestigious of Yuè “dialects” in Guangdong province. Other Yuè dialects are widely spoken in Sìyì 四邑, Zhōngshān 中山, “three districts” (comprising Nánhǎi 南海, Pānyú 番禺, and Shùndé 順德), and Dōngguǎn 東莞. Among Yuè “dialect” speakers, the variety spoken in the provincial capital Guangzhou (Guǎngzhōuhuà) is generally held to be the standard, as Anne Yue explains:2 The modern Yue language is spoken in central and western Guangdong as well as eastern Guangxi, concentrating especially in the Pearl River Delta region and along the West River and the North River valleys. (…) Linguistically speaking, “Cantonese” designates the dialect spoken in the city of Guangzhou or Canton while “Yue” refers to a Han language composed of many varieties with Cantonese as their prestige dialect. There are also a sizeable number of Yue speakers in Southeast Asia and North America. (Yue 2016: 174) Language shift patterned along the classic three-generation span has been reported (e.g. from Hakka to Cantonese: Lee 2008). Today, Cantonese is not only widely used in the home and on the street in Hong Kong and Macao but also in school (as the MoI) and broadcast media. In social interaction, Cantonese is the unmarked or default language of meetings among Chinese government officials and debates among Legislative Councilors, provided no non-Cantonese speakers are present (in which case English would be used). This is nicely illustrated by Dr. Sales Marques, the Macanese mayor of Macao before the handover, who was quoted as saying that “[w]ithin the [legislative] council, most of the business is done in Cantonese now, and I speak it pretty well. It’s too late for me to learn how to read and write Chinese properly” (McGivering 1999: 33). In both SARs, the Basic Law stipulates that the language of the former colonizers will continue to function as a co-official language alongside Chinese: English in Hong Kong, Portuguese in Macao. The term “Chinese” is vague, in that no mention is made which Chinese variety it refers to. It is generally understood that whereas in speech, Chinese refers to the vernacular Cantonese in Hong Kong SAR and Portuguese in Macau SAR, in writing it refers in both SARs to Pǔtōnghuà-based SWC. As for orthography, unlike mainland China and Singapore where the simplified script is used for writing Chinese, both SARs continue to employ the traditional Chinese character script. Although not widely used in society, the former colonial language in Macau SAR, Portuguese, continues to occupy an important position in the domains of government and law. After the handover, however,

150  David C.S. Li and Choi-Lan Tong the number of Portuguese-speaking civil servants has declined, with the majority working in the bureaus and offices under the Secretariat for Administration and Justice such as the Legal Affairs Bureau, International Law Office, and Judicial Reform Office. Pǔtōnghuà, the national language of China, has been promoted in both SARs for over two decades, since the handover mainly through education from primary school onward, either as a subject (e.g. several 35- to 40-minute lessons per week) or teaching Chinese in Pǔtōnghuà (Chan 2016; Li 2017). In postcolonial Hong Kong as in Macao, the function of Pǔtōnghuà in society remains largely symbolic. For example, it is used after Cantonese but before English during the trilingual flag-raising ceremonies of the SAR Establishment Day (July 1 in HKSAR, December 20 in MSAR) and National Day (October 1) (Li 2017; Xi and Moody 2010; on the status of local vernaculars vis-à-vis the standard language[s] elsewhere in the sinophone world, see Chen; Goh and Fong; Klöter; Meierkord; Siemund and Li; Snow; Wasserfall, all in this volume). In both SARs, communication between ethnolinguistic groups has not always been smooth and, in some cases, not even feasible. In Hong Kong, students of South Asian descent reportedly have difficulty learning SWC and, to a lesser extent, spoken Cantonese (Li 2017; Li and Chuk 2015). According to Hong Kong population 2016 by-census (Census and Statistics Department 2018), within a decade (non-Chinese) ethnic minorities (EMs) increased 70 percent to 584,383. Of these, people of South Asian descent (excluding Filipinos and Indonesians, who were employed mainly as domestic helpers) accounted for about 14.5 percent, of which the biggest groups were Indian (6.2 percent), Nepalese (4.4 percent), and Pakistani (3.1 percent). For young South Asians growing up in Hong Kong, integration into mainstream society is expected of them but difficult; the language barrier is the most often cited problem—indeed a perennial impediment—to social mobility. With little knowledge of the community language, Cantonese, learning written Chinese side by side with their same-age Cantonese-dominant peers is simply out of the question. Apart from educational opportunities, inequality is also felt in terms of access to healthcare and different social assistance schemes, for which basic knowledge in the vernacular and literacy in Standard Chinese is needed (Erni and Leung 2014). Their linguistic plight and the discrimination they face have attracted a lot of media attention thanks to dedicated NGOs like Unison. In their sociological study of the plight of Hong Kong South Asian EMs through the lens of “critical multiculturalism,” Erni and Leung give the following summary after reviewing nearly two dozen NGO reports: Perhaps the most central sociological idea developed by these NGO reports is that there always exists a dominant-minority relation that is shaped by a pervasive social neglect and discrimination. To understand the evolution of Hong Kong’s EM cultures is to understand the history

Tale of Two Special Administrative Regions  151 of an assumed value of superiority among the Chinese majority (…). The values, beliefs, and attitudes of racial and economic superiority have left their imprint on a social structure that has been built to be generally exclusivist and neglectful of minorities’ needs, particularly in areas of education, employment, and health care. (Erni and Leung 2014: 198) Some of the inequalities identified by the Equal Opportunities Commission have been addressed. In two recent surveys (Bacon-Shone and Bolton 2015; Chan 2019), roughly half of the South Asian youths are able to communicate in Cantonese with different degrees of fluency. From 2008 onward, the threshold Chinese language requirement for admission into Hong Kong tertiary institutions has been relaxed, in that alternative qualifications like General Certificate of Education (GCE) and General Certificate of Secondary Education (GCSE) are also accepted. That said, many EM students still struggle when learning written Chinese, which is required in many job-related settings. Similar communication problems were also found in Macao. For example, for a long time, Portuguese-L1 speakers had difficulty communicating with Cantonese speakers, partly because they were reluctant to learn the local vernacular. Until the handover in 1999, therefore, there was a huge language barrier between the two groups, as Sir Roger Lobo, a “successful son of the enclave” (McGivering 1999: 71) remarked: “there was always a divide between the Chinese and the non-Chinese…. It was almost like two separate worlds” (ibid., 73). Communication between them was mediated by the Macanese, which is at the same time a label for inhabitants of mixed Chinese and Portuguese descent in Macao, as well as a creolized variety based on Portuguese but heavily influenced by different Chinese “dialects” and Malay. Thanks to their plurilingual repertoire in Portuguese, Macanese, English, and Cantonese, the Macanese were in an ideal and privileged position to serve as middle persons or go-betweens that linked the two communities.

Vernacular Literacy, Medium-of-Learning Effect, and Translanguaging Unlike elsewhere in the sinophone world, in both SARs vernacular Cantonese is used as the MoI in Chinese Medium-of-Instruction (CMI) classes, whereas literacy education comprises SWC written in traditional characters. These practices bring with them two pedagogical problems. First, Cantonese-L1 students do not write the way they speak, for SWC is based on Pǔtōnghuà or Mandarin. Second, the traditional script generally contains more strokes and therefore takes longer to write, and, probably for that reason, tends to be easier to forget (compare e.g. 听 and 聽, tīng ‘to hear/ listen’; 龙 and 龍, lóng ‘dragon’).3 Although Cantonese is not part of school literacy, in literacy-focused activities, Cantonese elements tend to crop up

152  David C.S. Li and Choi-Lan Tong due to the natural tendency to write the way one speaks. One of the goals of Chinese literacy training is to weed out students’ “dialect” elements in writing and to replace them with Pǔtōnghuà-based equivalents. This being a laborious process, primary school education (P1–P6) is looked upon as the curriculum space to consolidate students’ school literacy in Chinese by age 11–12. Beyond the education domain, in terms of literacy practices in society, it is very common for vernacular elements, including Cantonese-English code-mixing, to surface in “soft” genres or informal sections of the media, print or electronic. Colloquial written Cantonese may not have a standard orthography, but this does little to deter eager writers from writing in Cantonese (Snow 2004). Where no known Chinese characters are found to represent the Cantonese morpho-syllables, Roman-based Cantonese words are sometimes improvised, for example, hea (he3 ‘laid-back’ or ‘tardy’), chok (cok3 ‘suffocating’), and chur (coe2 ‘hard pressed for time’) (Li et al. 2016; on vernacular literacy elsewhere in the sinophone world, see Chen; Snow; and Su, this volume). In Hong Kong, the popularity of written Cantonese elements in soft genres of various public and social media, print or digital, suggests that there is a strong market for them. This, in turn, explains why they are picked up almost effortlessly by Cantonese speakers despite their “non-school literacy” status. By contrast, in Macao such informal written Cantonese elements are less visible, partly because print media tend to adhere to more conservative written Chinese standards.

Language Attitudes toward English, Portuguese, and Pǔtōnghuà In Hong Kong, language attitude research conducted in the early 1980s shows that Cantonese-dominant students in Hong Kong tended to be reluctant to use English for fear of undermining their Chinese identity (Fu 1975; Pierson, Fu and Lee 1980). Similar language attitude surveys were conducted around the time of the handover; the findings suggest that the earlier trend was reversed, in that students in the 1990s were no longer so wary of appearing less Chinese when using English (e.g. Hyland 1997; Lin and Detaramani 1998). These findings suggest that English is widely seen by children born during the post-1980s and post-1990s as a form of linguistic capital, which is instrumental in facilitating upward and outward mobility. This is succinctly captured by Lai’s (2009) language attitudes study, namely, “I love Cantonese but I want English” (on attitudes toward English in Taiwan, see Chen; Su; Wasserfall, this volume). Lai (2009) collected quantitative and qualitative data in 2003 from the first cohort of students to come under the mother-tongue education policy in secondary education since September 1998 (on what “mother tongue” means to Cantonese speakers in China, see Liang, this volume). Statistical analyses of the 1,048 valid questionnaires showed that the student respondents

Tale of Two Special Administrative Regions  153 loved Cantonese most, valued English the most highly, and gave the lowest rating to Pǔtōnghuà (Lai 2009: 80). The findings from the matched-guise test were very similar: on traits that index competence such as “intelligent,” “competent,” “industrious,” and “educated,” and personal attractiveness like “wealthy” and “trendy,” the English voice was rated the highest. By contrast, on traits of solidarity (e.g. “friendly,” “sincere,” “considerate”), the Cantonese voice was rated the highest (ibid., 81). Again, the Pǔtōnghuà voice was given the lowest rating. The quantitative findings were supplemented with a qualitative analysis of ten group interviews collected from 40 students selected through purposive sampling. Lai found strong evidence of integrative orientation toward Cantonese, which was perceived as the language of the home for expressing personal affections. Regarding English and Pǔtōnghuà, many interviewees reported a stronger liking for English: whereas Pǔtōnghuà did not evoke any special feeling despite their awareness of its national language status, English as an international language was seen as more prestigious and useful, more powerful than Cantonese, and so most students found English more desirable to use but regrettably more difficult to learn. More recently, Leung (2017) adduced findings from two language use surveys in 2009 and 2014, each based on over 1,000 valid questionnaires completed by a majority of Cantonese-L1 respondents (97.4 and 95.8 percent, respectively). Using a 5-point scale (5 being “most frequently used,” 1 being “least frequently used”), Leung found that Cantonese continued to be rated as the most often used language in both “workplace” and “beyond workplace” contexts, followed by English and Pǔtōnghuà (cf. Leung and Li 2020). According to the 2014 data, the most widely reported use of Pǔtōnghuà occurred in four contexts: talking to customers (1.73), talking to clients (1.28), participating in corporate or cultural activities (1.27), and watching TV or listening to radio broadcasts (1.29), with level of education being the most salient factor or correlate: the higher the respondent’s educational attainment, the more likely Pǔtōnghuà would be used, especially in the workplace (Leung 2017: 86). Leung attributes this trend to the increasing presence of native speakers of Pǔtōnghuà, thanks to the SAR government’s Admission Scheme for Mainland Talents and Professionals (ASMTP, 輸入內地人才計畫, 83,685 successful cases by 2016), the “Individual Visit Scheme” (個人遊計劃), as well as thousands of mainland students admitted into Hong Kong university programs (Leung 2017: 86; cf. Chan 2016: 197). In Macao, during the colonial era, the motivation to learn Portuguese as an additional language was very low. Instead, English was looked upon as a useful asset, and so the people’s attitude toward English has always been more positive compared with Portuguese. A gradual shift in attitude was observable in January 1992, when it was made known that by law Chinese and Portuguese would function as co-official languages with immediate effect. Consequently, speakers of Portuguese working in the civil service, the Macanese included, would need to learn Pǔtōnghuà and SWC (Lei 2001),

154  David C.S. Li and Choi-Lan Tong while existing and aspiring Chinese civil servants would need to learn Portuguese, a de facto language requirement for joining the civil service. That is the background against which the learning of Portuguese became more and more popular, as shown in evening classes attended by many young people (Gary Ngai, cited in McGivering 1999: 158). Two later developments separated by a decade gave Portuguese-learning a still greater boost: Beijing’s decision to set up “The Forum for Economic and Trade Cooperation between China and Portuguese-speaking Countries (Macau)” in 2003;4 and China’s Belt and Road initiative formally rolled out in late 2013, whereby MSAR is positioned as the platform connecting China and the rest of the Lusophone countries in the world. These developments have been instrumental in turning Portuguese into a form of language capital. All this explains why, from both the points of view of language policy and practice, Macao has gradually evolved into a triliterate (Chinese, English, and Portuguese) and quadrilingual (Cantonese, English, Pǔtōnghuà, and Portuguese) society since the 1990s (compare language planning and language policy implementation in Taiwan: Chen, this volume). As regards attitudes toward Pǔtōnghuà and China, in Macao, it is in general less of a problem than in Hong Kong, probably because nearly half (45–50 percent) of the population was born on the mainland (about 40 percent Macao-born). Furthermore, from the 1980s onward, when study abroad became a popular option for secondary school-leavers, quite a few chose to pursue higher education on the mainland or in Taiwan. One consequence of this study-abroad experience is that most if not all of the returnees would have become fluent speakers of Mandarin/Pǔtōnghuà. This fact notwithstanding, there are signs of popular concern about the future of Cantonese as a result of the MSAR government’s recent plan to quicken the pace of improving the quality of teaching and learning of Pǔtōnghuà in the school sector.

Language Planning and Language-in-Education Policy in the Two SARs Poon (2010) makes the point that while the colonial Hong Kong government had a language-in-education policy that may be traced back to the late 1980s, strictly speaking, it did not have any language planning. The language-in-education policy was triggered by societal concern about the high school failure rate due to the fact that most Cantonese-L1 students lacked the ability to learn through the medium of English. While English is looked upon by various stakeholders as an indispensable form of linguistic capital for upward and outward mobility, it is not easy to learn (Li 2010). The “mother tongue education” policy, implemented in September 1998, was largely a compromise and, by design, an attempt to get the best of two possible worlds: those who have demonstrated an ability to study through English-medium instruction (EMI) are assigned to EMI schools (about 30

Tale of Two Special Administrative Regions  155 percent of more than 400 schools, Li 2017), while those who do not have this ability will go on to secondary schools where Chinese (i.e. spoken Cantonese and SWC) is used as the MoI (CMI). As warned by many critics in education circles, one major problem engendered by this policy is an unwanted but unavoidable labeling effect, with CMI students being widely perceived as failures or second best. Those who found a place in an EMI school are pedagogically not necessarily better off, for research shows that many S1 (Grade 7) students are not ready to learn through the medium of English, suggesting that the quality of learning is compromised compared with learning through one’s L1. In some cases, EMI students may have to repeat a year to keep up with the EMI curriculum. Amidst strong criticism, in 2009, the Education Bureau of the SAR government agreed to fine-tune the mother-tongue education policy by allowing for more flexibility within CMI schools, such that principals may set up English-medium classes, by level or subject, for students who have met certain threshold conditions for learning through EMI. Regarding language policy and language planning in Macao, Bray and Koo (2004) noted that successive colonial administrations in Macao appeared to have a laissez-faire or indifferent attitude toward language policy. There is little evidence of the colonial government promoting the teaching and learning of the Portuguese language, which is why few local people had any knowledge of Portuguese. This is partly evidenced by the childhood memories of contributors to the collection of 26 first-person narratives by Jill McGivering in her book Macao Remembers (Liáng and Li 2011). For instance, Gary Ngai, an experienced Chinese promoter of local culture, was quoted as saying: “When I first came here [from mainland China] twenty years ago, almost no Chinese, beyond a handful of civil servants, spoke Portuguese…. The attitude used to be: ‘Why should I bother to learn Portuguese? It’s useless’” (McGivering 1999: 158). As a second language, English was much more popular, largely due to the influence of Hong Kong. As Bray and Koo (2004) observed, from an economic point of view, Macao was less of a colony than Hong Kong, for the latter’s economic success “contributed to a stronger role for Hong Kong’s colonial language (English) than Macao’s colonial language (Portuguese)” (Bray and Koo 2004: 233; cf. Xi and Moody 2010: 314). Xi and Moody (2010) reviewed a number of studies on the language-ineducation policy of Macao during the colonial era, and found that school principals were given a free hand to choose the MoI. This practice remained unchanged after the handover. According to Sou (2000), primary and secondary schools may be divided into three categories depending on their MoI, which may be Chinese, Portuguese, or English. “Official schools” may choose one of the official languages—Chinese or Portuguese—as the principal MoI, but the other language must be taught as a second language. Private schools, on the other hand, have the option of choosing Chinese, Portuguese, or English as the principal MoI, but a second language, to be

156  David C.S. Li and Choi-Lan Tong taught as a subject, must be selected from the other two languages in the curriculum. Such a policy is enshrined in law as follows: Under Law No. 9/2006, the latest law which outlines the non-tertiary education system in post-colonial Macao, all government educational institutions may adopt one official language (Portuguese or Chinese) as medium of instruction while private educational institutions may choose the medium of instruction according to the needs of their students. In addition, both government and private educational institutions are suggested to provide students with an opportunity to learn at least one official language that is not used as the medium of instruction. (Young 2009: 416) The most recent statistics in the school sector are based on the 2017/18 school year (Education and Youth Affairs Bureau 2017, 2018, 2020). Of the 74 schools, 10 are public and 64 are private. The student population in public schools accounts for only 3.5 percent (n = 2,707). Only four schools— one public and three private—have adopted Portuguese as the MoI (1.3 percent, 956 students). A total of 56 of the schools are Chinese-medium (84.1  percent, 64,236 students), and the remaining 14 are English-medium (14.6 percent, 11,154 students). Portuguese has been actively promoted by the Education and Youth Affairs Bureau (Direcção dos Serviços de Educação e Juventude in Portuguese), supporting the teaching of Portuguese in public and private schools alike, but also organizing extra-curricular activities for students (n = 6,548). While by law schools are given the freedom to use Cantonese or Pǔtōnghuà as the medium for teaching the Chinese subject, about a quarter (24 percent) of the CMI schools and the majority of the EMI and PMI schools have adopted Pǔtōnghuà as the MOI (i.e. teaching Chinese in Pǔtōnghuà, 普教中). Additional measures to boost MSAR’s Pǔtōnghuà level include providing subsidies for Chinese language teachers to attend courses to improve their Pǔtōnghuà, and inviting leading Pǔtōnghuà teacher trainers from the mainland to visit Macao. Finally, to enhance the standard of the target languages, i.e. Pǔtōnghuà, English, and Portuguese, from 2013/14 onward, the MSAR government started subsidizing secondary language teachers and students to sit for internationally recognized language tests. Elsewhere in the sinophone world, the choice of MOI in the education sector is no less a contentious issue (see Goh and Fong; Klöter; Meierkord; Siemund and Li, this volume).

The Future of Cantonese Just before Hong Kong was renationalized as part of China, several predictions and recommendations were made regarding the language situation in the former British colony. Snow (1991) pointed to the widespread use of written Cantonese as an indicator of Hongkongers’ cultural autonomy.

Tale of Two Special Administrative Regions  157 Pierson (1998) likewise emphasized the symbolic value of Cantonese for Hongkongers, and suggested that Cantonese could well be seen as a symbol of freedom, democracy, and independence. Yau (1992) also sounded a warning that Cantonese could eventually yield its place to Pǔtōnghuà as the principal vernacular in post-1997 Hong Kong should nothing be done to disambiguate the term “Chinese” (中文 zung55 man 21, Zhōngwén) in the Basic Law. Bauer (1995: 290) echoed Yau’s concern and believed that the absence of any mention of “the potentially contentious issue of the relationship between Putonghua and Cantonese” in the Basic Law might be a foreboding that Pǔtōnghuà might be imposed on Hong Kong, probably beginning with the domains of government, education, and mass media. In this case, the mother-tongue education policy implemented from September 1998 and the promotion of Pǔtōnghuà in Hong Kong (and Macao) have also spawned speculation that the Beijing government was adamant in its agenda to replace Cantonese with Pǔtōnghuà over time. Critics who subscribed to this theory would point to the fate of Cantonese in Guangzhou and elsewhere in Guangdong province (Li 2000: 226). By contrast, Erbaugh (1995: 82) was more optimistic in regard to the future of Cantonese in Hong Kong. Her judgment was based on the PRC government’s long-standing policy of “dialect bilingualism” which, in her view, would most likely be extended to Hong Kong (and Macao) given that “language tolerance offers a low-cost, low-risk token of goodwill.” As for the critics’ concern about the ambiguity of the exact referent(s) of the term “Chinese” in the Basic Law, Erbaugh (1995) interpreted that differently, contending that Cantonese would probably be tacitly allowed as a regional vernacular. Based on these observations, Erbaugh (1995) believed that the status quo of Hong Kong’s language situation would more likely remain unchanged (cf. Bradley 1992). Until today, two decades after the handover, none of the earlier predictions has come true. Quite the contrary, Cantonese elements in the Hong Kong Chinese press are alive and well, suggesting that Erbaugh’s (1995) prediction is more in line with post-handover sociopolitical development. Thus, the burden of proof seems to rest with those who see a threat to the sustained vitality of Cantonese in Hong Kong SAR. Crucial to this up-beat prognosis is community-wide critical awareness of Cantonese being endangered (compare Liang, this volume). This is corroborated by news events in the Pearl River Delta during the past decade. In July and August 2010, several months before the China-hosted Asian Games in November, a proposal by a municipal government official in Guangzhou, apparently to make visitors feel welcome, to switch the language of some programs on local television from Cantonese to Pǔtōnghuà sparked social protest leading to the arrest of about two dozen demonstrators, including some Hong Kong journalists. Such street rallies in the provincial capital were echoed by like-minded protesters in Hong Kong, who objected strongly to Cantonese being relegated to a dialect of the home and the street similar to what their brethren across the

158  David C.S. Li and Choi-Lan Tong border have to put up with (see e.g. Ramzy 2010). As one would expect, much more heated discussions and commentaries could be found online (see e.g. the Chinese version of a hashtag like “Cantonese is not a dialect”: Gao 2012). Identity concerns, much more pervasive in Hong Kong than in Macao, are also evidenced by Cantonese-dominant university students’ (passive) resistance to using Pǔtōnghuà for teaching Chinese language courses (e.g. the HKBU “occupy Language Center” incident in early 2018; see Cheung 2018). Such student sentiments are indicative of an ethnolinguistic fault line between “them,” Pǔtōnghuà-speaking Mainlanders from across the border, and “us” Cantonese Hongkongers, which is deep in the psyche of the new millennial generation in HKSAR. This “us vs. them” mindset has been intricately embedded in various social campaigns and movements, big and small. From mass protests such as the “umbrella movement” in the 79-day “Occupy Central” saga in 2014 and the worst riot in decades on the eve of the second Lunar New Year day in February 2016 to wild-cat flash mobs chanting kau1 wu1 (written Cantonese 鳩嗚, punning on Pǔtōnghuà 购物 gòuwù ‘shopping’) in mockery of Chief Executive C.Y. Leung’s appeal for shopping and doing business in riot-hit Mongkok, that Cantonese-Pǔtōnghuà fault line has increasingly morphed into and coincided with a “friend or foe” divide. Such a worrying trend is further evidenced in the spontaneous outrage expressed by young people in the street against the “onslaught” of Pǔtōnghuà speakers, be they disruptive parallel traders of baby formula, shopaholic mainlanders loaded with bags and boxes braving their way through narrow streets, or “dancing aunties” (middle-aged mainland Chinese women commonly referred to as大媽 dàmā) singing to amplified karaoke music in crowded public spaces or tourist spots like Mongkok and Tsim Sha Tsui (Dapiran 2017). In Macao, by contrast, the future of Cantonese was not a matter of concern across society until recently thanks to the relatively stable political situation after the handover in December 1999.

Coda Hong Kong is often cited as a quintessential example of an economic miracle. From a “barren rock” to a financial center rivaling New York, London, and Tokyo, Hong Kong has come a long way. Its economic well-being and prosperity today could hardly be imagined without several generations of hard-working people, mostly migrants-turned settlers, contributing their talents and physical labor each in their own way. The presence of English— the language of the colonial masters—has played a facilitating if not instrumental role, in that it has enabled part of the local workforce to benefit from and participate in many of the processes of globalization that turned metropolitan Hong Kong into a glittering “Pearl of the Orient” (東方之珠) some 40 years ago (etiquette earned based on an aerial picture of Hong Kong taken by night), and a self-styled “Asia’s World City” (亞洲國際都會) more recently. Macao, tiny though it is, made a mark on history first through its role as a Portuguese colonial enclave, before emerging as the world’s third-largest

Tale of Two Special Administrative Regions  159 gaming capital rivaling Monte Carlo and Las Vegas, even though the widely  acclaimed label it was once famous for—The Monte Carlo of the East (東方蒙特卡羅)—has faded into oblivion. It should be remembered that Macao was where the earliest European traders and adventurers settled in the mid-sixteenth century, and the site where the first university built after a western model was established in south China. Founded by the Jesuits, the mission of University College of St. Paul was “to cultivate missionaries for the region” (Young 2009: 412). For centuries, however, Macao’s Cantonesedominant residents continued to look to English rather than Portuguese as a springboard for gaining access to lucrative work opportunities and a better life, possibly in Hong Kong. Thus is a tale of China’s two SARs. To better understand their past is arguably a key to understanding the present. One question concerning their respective state of multilingualism is: What is the connection between the plurilinguality of their people and the vicissitudes of their language and education policies since colonial times? This question awaits fine-grained research, which will be the subject of another tale.

Notes

References Anonymous. (n.d.). Ou 3 mun 2 faan1 jyu 5 zaap 6 zi 6 cyun 4 bun 2 澳門畨語雜字全本 [Compendium of assorted phrases in Macau Pidgin]. Guangzhou: Wǔ guì táng. Bacon-Shone, John, and Kingsley Bolton. 2015. “A Study on the Use of Language in the Workplace of Hong Kong.” Hong Kong: Social Sciences Research Centre, The University of Hong Kong. Accessed May 28, 2019 via http://www.nzu.edu.sg. Bauer, Robert S. 1995. “Syllable and Word in Cantonese.” Journal of Asian Pacific Communication 6, no. 4: 245–306.

160  David C.S. Li and Choi-Lan Tong Bolton, Kingsley. 2003. Chinese Englishes: A Sociolinguistic History. Cambridge: Cambridge University Press. Bradley, David. 1992. “Chinese as a Pluricentric Language.” In Pluricentric Languages: Differing Norms in Different Nations, edited by Michael Clyne, 305–24. Berlin: Mouton de Gruyter. Bray, Mark, and Ramsey Koo. 2004. “Postcolonial Patterns and Paradoxes: Language and Education in Hong Kong and Macao.” Comparative Education 40, no. 2: 215–39. Census and Statistics Department (Hong Kong SAR). “2016 Population By-Census.” Accessed May 28, 2019 via https://www.bycensus2016.gov.hk. Chan, Shui-Duen. 2016. “Putonghua and Cantonese in the Chinese Territories.” In Communicating with Asia: The Future of English as a Global Language, edited by Gerhard Leitner, Azirah Hashim, and Hans-Georg Wolf, 188–204. Cambridge: Cambridge University Press. ——— 陳瑞端. 2019. “Xiānggǎng Nányà zúqún de yǔyán shíyòng zhuàngkuàng” 香港南亚族群的语言使用状况 [A survey of language use among South Asian Hongkongers]. In Zhōngguó yǔyán shēnghuó zhuàngkuàng bàogào 2019, edited by the State Language Commission, 255–62. Beijing: Shāngwù yìnshūguǎn. Cheung, Elizabeth. 2018. “Hong Kong Baptist University Students Lose Appeal against Punishment for Role in Rowdy Mandarin Protests.” South China Morning Post, May 18. Cummings, Patrick J., and Hans-Georg Wolf. 2011. A Dictionary of Hong Kong English: Words from the Fragrant Harbor. Hong Kong: Hong Kong University Press. Dapiran, Anthony. 2017. City of Protest: A Recent History of Dissent in Hong Kong. Hawthorn: Penguin Books Australia. DSEC (Direcção dos Serviços de Estatística e Censos, Macau SAR). 2019. “Demographic Statistics: 3rd Quarter 2019.” Accessed March 4, 2020 via https://www. dsec.gov.mo. Education and Youth Affairs Bureau (Macau SAR). 2017. “2017 Annual Report of Education and Youth Affairs Bureau.” Accessed May 28, 2019 via https://portal. dsej.gov.mo. ———. 2018. “Overview of Education in Macao: Review 2017.” Accessed May 28, 2019 via https://portal.dsej.gov.mo. ———. 2020. “Education Institution: Statistics.” Accessed February 25, 2020 via https://portal.dsej.gov.mo. Erbaugh, Mary S. 1995. “Southern Chinese Dialects as a Medium for Reconciliation within Greater China.” Language in Society 24: 79–94. Erni, John Nguyet, and Lisa Yuk-ming Leung 2014. Understanding South Asian Minorities in Hong Kong. Hong Kong: Hong Kong University Press. Fensom, Anthony. 2015. “Macau Gambles on More Casinos amid Downturn.” The Diplomat, May 6. Fu, Gail S. 1975. “A Hong Kong Perspective: English Language Learning and the Chinese Student.” PhD diss., University of Michigan. Gao, Xuesong. 2012. “‘Cantonese is Not a Dialect’: Chinese Netizens’ Defence of Cantonese as a Regional Lingua Franca.” Journal of Multilingual and Multicultural Development 33, no. 5: 449–64. Hao, Zhidong. 2011. Macau: History and Society. Hong Kong: Hong Kong University Press. He, Huifeng. 2015. “Shenzhen Fails to Meet Economic Growth Targets.” South China Morning Post, May 31.

Tale of Two Special Administrative Regions  161 Hyland, Ken. 1997. “Language Attitudes at the Handover: Communication and Identity in 1997 Hong Kong.” English World-Wide 18, no. 2: 191–210. Kwok, Bit-Chee, Andy C. Chin, and Benjamin K. Tsou. 2016. “Grammatical Diversity across the Yue Dialects.” Journal of Chinese Linguistics 44, no. 1: 109–52. Lai, Mee-Ling. 2009. “‘I love Cantonese but I Want English’—A Qualitative Account of Hong Kong Students’ Language Attitudes.” The Asia-Pacific Education Researcher 18, no. 1: 79–92. Lai, Ying-kit, and Amy Nip. 2015. “Shenzhen Knocks Down Hong Kong to Take Top Spot in China’s Ranking of Most Competitive Cities.” South China Morning Post, May 15. Lee, Sherman. 2008. “A Study of Language Choice and Language Shift among the Hakka-Speaking Population in Hong Kong, with a Primary Focus on Sha Tau Kok.” PhD diss., City University of Hong Kong. Lei, Sio Lin. 2001. “The Application of the Linguistic Relativity Thesis to the Situation in Macau: The Reflection of Chinese Religious Culture in Macanese Lexical Items.” Master’s thesis, University of Hong Kong. Leung Wai Mun 梁慧敏. 2017. “Xiānggǎng Pǔtōnghuà shǐyòng de shízhèng yánjiū” 香港普通話使用的實證研究 [Empirical research on the use of Pǔtōnghuà in Hong Kong]. Yǔyán wénzì yìngyòng no. 3: 79–90. ———, and David C.S. Li 李楚成. 2020. Liǎng wén sān yǔ: Xiānggǎng yǔwén jiàoyù zhèngcè yánjiū 兩文三語：香港語文教育政策研究 [Biliteracy and trilingualism: Language education policy research in Hong Kong]. Hong Kong: City University of Hong Kong Press. Li, David C.S. 2000. “Phonetic Borrowing: Key to the Vitality of Written Cantonese in Hong Kong.” Written Language and Literacy 3, no. 2: 199–233. ———. 2009. “Towards ‘Biliteracy and Trilingualism’ in Hong Kong (SAR): Problems, Dilemmas and Stakeholders’ Views.” AILA Review 22: 72–84. ———. 2010. “Improving the Standards and Promoting the Use of English in Hong Kong: Issues, Problems and Prospects.” In English Language Use and Education across Greater China, edited by Anwei Feng, 95–113. Clevedon: Multilingual Matters. ——— 2017. Multilingual Hong Kong: Languages, Literacies and Identities. Cham: Springer. ———, and Joanne Y. P. Chuk. 2015. “South Asian Students’ Needs for Cantonese and Written Chinese in Hong Kong: A Linguistic Study.” International Journal of Multilingualism 12, no. 2: 210–24. ——— et al. 2016. “Facilitation of Transference: The Case of Monosyllabic Salience in Hong Kong Cantonese.” Linguistics 54, no. 1: 1–58. Li, Michelle, and Stephen Matthews. 2016. “An Outline of Macau Pidgin Portuguese.” Journal of Pidgin and Creole Languages 31, no. 1: 141–83. Liáng Huìmǐn 梁慧敏, and David C.S. Li 李楚成. 2011. “Àomén yǔyán xiànzhuàng zàitàn” 澳門語言現狀再探 [The state of multilingualism in Macau SAR: An update]. Shénzhōu jiāoliú 8, no. 1: 82–96. Lin, Angel M.Y., and Champa Detaramani. 1998. “By Carrot and by Rod: Extrinsic Motivation and English Attainment of Tertiary Students in Hong Kong.” In Language in Hong Kong at Century’s End, edited by Martha C. Pennington, 285–301. Hong Kong: Hong Kong University Press. Macaulay, Thomas B. 1835. “Minute on Education.” Accessed May 19, 2019 via http://www.columbia.edu.

162  David C.S. Li and Choi-Lan Tong Matthews, Stephen. 2013. “Language Contact and Chinese.” In The Handbook of Language Contact, edited by Raymond Hickey, 757–69. Malden, MA: Wiley-Blackwell. McGivering, Jill. 1999. Macao Remembers. Hong Kong: Oxford University Press. Moody, Andrew. 2008. “Macau English: Status, Functions and Forms.” English Today 24, no. 3: 3–15. People’s Daily Online. 2015. “Shenzhen Ranks Top in China City Competitiveness.” Accessed May 28, 2019 via http://en.people.cn. Pierson, Herbert D. 1998. “Social Accommodation to English and Putonghua in Cantonese-Speaking Hong Kong.” In Language in Hong Kong at Century’s End, edited by Martha C. Pennington, 91–111. Hong Kong: Hong Kong University Press. ———, Gail S. Fu, and Sik-yum Lee. 1980. “An Analysis of the Relationship between Language Attitudes and English Attainment of Secondary School Students in Hong Kong.” Language Learning 30, no. 2: 289–316. Poon, Anita Y. K. 2010. “Language Use, and Language Policy and Planning in Hong Kong.” Current Issues in Language Planning 11, no. 1: 1–66. Ramzy, Austin. 2010. “Protesters Stand Firm on Cantonese Rights.” Time, August 2. Snow, Donald B. 1991. “Written Cantonese and the Culture of Hong Kong: The Growth of a Dialect Literature.” PhD diss., Indiana University. ———. 2004. Cantonese as Written Language: The Growth of a Written Chinese Vernacular. Hong Kong: Hong Kong University Press. Sou, Chio Fai. 2000. “Establishment of an Education System in Macau.” In Reform of Educational Systems and Teacher Education: The Asian Experience, edited by Philip K. F. Hui and Peijia Chan, 205–9. Hong Kong: The Hong Kong Institute of Education. Sun Seak-Leong 孙锡亮. 2015. “Àomén Yuèyǔ zhōng de Púyǔ jiècí yánjiū” 澳门粤 语中的葡语借词研究 [On loanwords from Portuguese in Macao Yuè dialect]. Zhōngguó fāngyán xuébào, no. 1: 130–43. Tong Choi-Lan 湯翠蘭. 2015. “Àomén Yuèyǔ zhōng de Púyǔ jiècí de xiāotuì xiànxiàng fēnxī: Yǐ shíwù lèi cíhuì wéi lì” 澳門粵語中葡語借詞的消退現象分析—— 以食物類詞彙為例 [An analysis of the declining phenomenon of Portuguese loan words in Macao Cantonese: The food glossary as a case study]. In Àomén yǔyán wénhuà yánjiū 2014, edited by Lei Heong-Iok, 210–22. Macao: Macao Polytechnic Institute. Van Dyke, Paul A. 2005. The Canton Trade: Life and Enterprise on the China Coast, 1700–1845. Hong Kong: Hong Kong University Press. ———. 2011. Merchants of Canton and Macao: Politics and Strategies in EighteenthCentury Chinese Trade. Hong Kong: Hong Kong University Press. Wàn Yèxīn 万业馨. 2012. Yìngyòng Hànzìxué gàiyào 应用汉字学概要 [Essentials of applied Chinese character studies]. Beijing: Shāngwù yìnshūguǎn. Wei, Yiyang, and Timmy Shen. 2019. “Macau’s Gambling Sector Reports Biggest Revenue in Four Years.” Caixin, January 3. Accessed May 28, 2019 via https:// www.caixinglobal.com. Wood, Laura. 2018. “Analyzing the Casinos & Gambling Industry in China 2018.” Business Wire, October 18. Accessed May 28, 2019 via https://businesswire.com. Wú Zhìliáng 吳志良, and Chan Ian Ian 陳欣欣. 2000. Àomén zhèngzhì shèhuì yánjiū 澳門政治社會研究 [Politics and society of Macao]. Macao: Associação de Educação de Adultos de Macau.

Tale of Two Special Administrative Regions  163 Xi, Yan, and Andrew Moody. 2010. “Language and Society in Macao: A Review of Sociolinguistic Studies on Macao in the Past Three Decades.” Chinese Language and Discourse 1, no. 2: 293–324. Yau, Shun Chiu. 1992. “Language Policies in Post-1997 Hong Kong.” In Into the Twenty-First Century: Issues of Language in Education in Hong Kong, edited by K.K. Luke, 15–29. Hong Kong: Linguistic Society of Hong Kong. Young, Ming Yee Carissa. 2009. “Multilingual Education in Macao.” International Journal of Multilingualism 6, no. 4: 412–25. Yue, Anne O. 2016. “The Yue dialect.” In Oxford Handbook of Chinese Linguistics, edited by William S-Y. Wang and Chao Fen Sun, 173–88. Oxford: Oxford University Press. Zhang, Zhenjiang. 2009. Language and Society in Early Hong Kong (1841–1884). Guangzhou: Sun Yat-sen University Press.

8

One People, One Nation, One Singapore Language Policy and Shifting Identities among Chinese Singaporeans Yeng Seng Goh and Yeow Wah Fong Abstract This chapter provides an overview of language planning in Singapore since the country’s independence in 1965. It examines official responses toward societal multilingualism and the ideologies underlying language planning, notably the promotion of English and Mandarin. Before independence, Mandarin was hardly ever used within the Chinese population. It was only after 1965 that it became an official language—alongside English, Malay, and Tamil—and gradually gained dominance over the other Sinitic varieties through education and intensive campaigning. The chapter argues that the evolution of bilingualism within the Chinese community, which accompanied a shift in identity from overseas Chinese to Chinese Singaporeans, resulted in an increasing percentage of Chinese Singaporeans using English predominantly as a home language. Meanwhile, in an era of global Mandarin, Singapore is gradually shifting toward a language situation where Chinese is increasingly learned merely as a foreign language and is not being used in the home context.

Introduction: The Changing Landscape Singapore is a multi-ethnic and multilingual society where language policy and planning play a critical role in national development. One of the most significant sociolinguistic developments over the past few decades has been the rapid rise of English. This internal shift toward English as the dominant language, with a strong tendency among the Chinese population to abandon Mandarin as home language in favor of English, is paralleled by a contradictory trend evidenced in the (external) re-emergence of China as a global political and economic power in the world. With the passing of Singapore’s first Prime Minister Lee Kuan Yew 李光耀 (1923–2015), who is widely acknowledged as the chief architect of the country’s language policy, language policy and planning in Singapore crossed a watershed. How will it advance in the post-Lee era? This chapter draws extensively upon the views and visions of Lee on language policy and planning in Singapore, as set out in his memoirs and books, with a particular focus on the Chinese language.

One People, One Nation, One Singapore  165 The Internal Factor: Rapid Rise of English On August 9, 1965, in his first press conference after Singapore became independent, then Prime Minister Lee Kuan Yew declared: We are going to have a multiracial nation in Singapore. We will set the example. This is not a Malay nation, this is not a Chinese nation, this is not an Indian nation. Everybody will have his place: equal language, culture, religion. (Lee 1965) Ever since then, students in Singapore assemble at the start of each school day to sing the Malay national anthem and recite the pledge “We, the citizens of Singapore, pledge ourselves as one united people, regardless of race, language or religion, to build a democratic society, based on justice and equality so as to achieve happiness, prosperity and progress for our nation.” After over 50 years of nation-building efforts, Singapore has constructed a distinctive national identity. Of all the nation-building initiatives, bilingual policy is one of the most significant. Lee Kuan Yew, who played a pivotal role in shaping Singapore’s language policy, showed a keen personal interest in the nation’s language issues,1 resulting in the socio-political nature of language planning in Singapore. In an account of his struggle to transform the language ecosystem in Singapore, he pointed out the importance of language policy in nation-building: Language policy is a vital instrument for achieving national interest objectives and meeting the needs of governance. Rightly conceived, it can help unite a population that is racially and linguistically diverse, as well as build a platform for communication with the outside world. (Lee 2012: 224) With the implementation of the bilingual policy, the last five decades have seen a drastic shift from multilingualism to the dominance of English in the sociolinguistic landscape. For Singapore-Chinese families, English has increasingly become the dominant home language. It is now the most frequently spoken language in the home for more than 50 percent of younger Chinese aged 5 to 24 years old (Department of Statistics 2016). The External Factor: Re-emergence of China and the Global Spread of Mandarin China’s re-emergence as a global political and economic power is reflected in its “One Belt One Road” (OBOR; 一带一路) initiative, and in the creation of the Asian Infrastructure and Investment Bank (AIIB). Both initiatives

166  Yeng Seng Goh and Yeow Wah Fong are an important zǒu chūqù (走出去 ‘venture out’) attempt by China to build a new global order and to generate goodwill and exert influence beyond its own borders. Support of both initiatives has spread beyond China’s immediate neighbors to more than 60 countries, demonstrating a further increase in China’s clout and economic influence. In 2014, after proposing the OBOR initiative, China’s President Xi Jinping 习近平 (b. 1953) explicitly appealed to all overseas Chinese, whether they be Chinese citizens (华侨) or foreign citizens of Chinese origin (华人), to come together to work for the common good implicit in the “Chinese Dream” (中国梦). Subsequent policies such as the introduction of “Huáyìkǎ” (华裔卡, Green Card for descendants of Chinese) are evidence of China’s adoption of a transnational policy. In January 2018, the Chinese government announced that foreigners of Chinese origin would be able to apply for multiple-entry visas for a period of five years from February 1, 2018. Former Chinese citizens who have acquired foreign citizenship and ethnic Chinese whose parent, grandparent, or ancestor is or was a Chinese citizen are also eligible to apply for the new visa (Cheong 2018). The policy change is an attempt to attract overseas Chinese to live and do business in China. China’s immense influence politically and economically in the international arena has enhanced the status of Mandarin. Mandarin Chinese, which has been the dominant working language of administration (in both the public and private sectors), education, law, mass communication, science, technology, commerce, and so on in mainland China and Taiwan, and a lingua franca among the Chinese diaspora in the sinophone world, is widely believed to have the potential to achieve global status alongside English. With the substantial flows of Chinese citizens, whether for tourism, business, study, or migration, the use of Mandarin is becoming increasingly prevalent in the world. The establishment of Confucius Institutes around the world has also promoted the spread of Mandarin.

The Ethnic Composition and Sociolinguistic Landscape of Singapore The population of Singapore comprises three major ethnic groups of Chinese, Malay, and Indian. Table 8.1 shows the ethnic composition in the period from 1824 to 2010. Since the beginning of the twentieth century, Chinese have constituted a majority in Singapore with a stable population of about 70 percent. Up until 1965, Singapore was a truly multilingual society. According to the 1957 census, 33 languages or dialects were spoken, 20 of which by at least a thousand people. The Chinese, comprising 75.4 percent of the population, was the most linguistically diverse group, with Hokkien being the predominantly spoken dialect (30 percent), followed by Teochew (17 p ercent), Cantonese (15 percent), Hainanese (5 percent), Hakka (5 percent), and other Chinese dialects. Mandarin was only spoken by 0.1 percent of the

One People, One Nation, One Singapore  167 Table 8.1 The ethnic composition in Singapore in percent (1824–2010) Year

Chinese

Malays

Indians

Others

Total

1824 1871 1901 1931 1947 1957 1970 2000 2010

31.0 57.6 72.1 75.1 77.8 75.4 76.2 76.8 74.1

60.2 27.6 15.8 11.7 12.1 13.6 15.0 13.9 13.4

7.1 10.9 7.8 9.1 7.7 9.0 7.0 7.9 9.2

1.7 4.0 4.3 4.2 2.4 2.0 1.8 1.4 3.3

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Sources: Saw (2007), Singapore Department of Statistics (2011).

population (Chua 1964). In brief, upon its founding in 1965, Singapore can be considered a textbook example of sinophone polyphonicity. It was only after the independence of Singapore in 1965 that, as a result of official language planning, Mandarin gained dominance over the other Sinitic varieties. As a consequence, Singapore’s sociolinguistic landscape has undergone a major shift. According to the Census of Population 2010, Mandarin had become the most frequently spoken language at home (35.6 percent), which was a jump of more than 30 percent as compared to the 1957 census. English 2 was found to be the second most widely spoken language with a percentage of 32.3, whereas the use of Hokkien, Teochew, and Cantonese had dropped drastically to 7.0, 2.8, and 3.6 percent, respectively (Singapore Department of Statistics 2011).

The Bilingual Policy of Singapore during the Lee Kuan Yew Era: An Overview Language Planning and Policy Implementation at the National Level During the colonial era, the British government basically adopted a laissezfaire language policy whereby each community took care of its own language needs. Under British rule, English, as the language of administration, was the prestige or “high” language. Other community languages, such as Chinese, Malay, and Tamil, were relegated to “low” status. However, English did not serve as a common language for inter-ethnic communication. Indeed, the British saw no need for a common language, for it was easier to rule and govern the colony if the various communities remained divided and separate. This laissez-faire attitude effectively led to the flowering of multiple monolingual education spheres within the various communities, with each community naturally adopting a monolingual policy that supported its own heritage language.

168  Yeng Seng Goh and Yeow Wah Fong In his memoirs, Lee Kuan Yew described the language policy status of pre-independence Singapore as follows: Singapore never had one common language. It was a polyglot community under colonial rule. The British left the people to decide how to educate their children. The government provided a limited number of English-language schools to train people to be clerks, storekeepers, draughtsmen and such subordinate workers, and Malay-language primary schools for Malays. The Indians ran their own Tamil and other Indian-language schools or classes. The Chinese set up schools financed by successful members of their community, to teach in Chinese. Because the different races were taught their own languages, their emotional attachment to their mother tongue was deep. (Lee 2000: 169–70) When Singapore became independent in 1965, the government restated that all four major languages (Chinese, Malay, Tamil, and English) were to be recognized as official languages at the national level. In reality, English occupies a dominant position. Not only does it serve as a language of communication among the different ethnic groups, it is also the language used in the fields of politics, business, law, education, technology, and administration. This results in diglossia whereby English and the other three major languages have different functions in the society. English: The High Variety It is particularly unusual that even though Singapore has had a majority Chinese population since the beginning of the twentieth century, it nevertheless adopted English, which was then a colonial language spoken by a minority, as the dominant language under its bilingual policy.3 Lee Kuan Yew reflects upon the rationale behind his decision as follows: “Unlike countries with ethnically and linguistically homogeneous populations where the language of administration follows that of the population, it was not obvious to us at the start which language we should choose as our official or working language” (Lee 2012: 225). He furthermore argues that: “But making Chinese the official language of Singapore was out of the question; the 25 per cent of the population who were not Chinese would revolt. In addition, Singapore is located in Southeast Asia, which is predominantly Malay-speaking” (Lee 2012: 59). The rationale for the dominance of English in Singapore is thus twofold. First, English has a political role as a neutral language to promote inter-ethnic communication and racial harmony. It is thus an ideal choice to serve as a common language among different ethnic groups: “The nub of the problem was that in our multiracial and multilingual society, English was the only acceptable neutral language, besides being the language that

One People, One Nation, One Singapore  169 would make us relevant to the world” (Lee 2000: 173). Meanwhile, English also fulfills an economic function since it remains the dominant global language for international diplomacy, business, trade, science, and technology: “We realized English had to be the language of the workplace and common language. As an international trading community, we would not make a living if we used Malay, Chinese or Tamil” (Lee 2000: 170). Recognizing the constraints that Singapore has as a nation, Lee Kuan Yew envisaged foreign investment as a viable solution to the country’s economic survival. In Lee’s view, equipping his people with English language skills would create a world-ready workforce and attract foreign investments: How would Singapore make a living? With barely 700 sq km of land, agriculture was out of the question. Trade and industry were our only hope. But to attract investors here to set up their manufacturing plants, our people had to speak a language they could understand. That language had to be English—since World War II ended, the English language had spread. It was the language of international diplomacy, the language of science and technology, and the language of international finance and commerce. Singapore would have increased opportunities if they had a strong mastery of English. (Lee 2012: 59) The mastery of English among Singaporeans has not only attracted foreign investors but also enabled Singaporeans to secure employment which, in turn, has increased domestic stability. According to Lee, “knowing English has also enabled our people to secure jobs much faster; the multinational companies which set up their factories, research laboratories and regional headquarters in Singapore demand workers who are fluent in spoken and written English” (Lee 2012: 226). “Mother Tongue Languages”: The Low Variety Before independence, the languages used within the three ethnic groups in Singapore, namely Chinese, Malays, and Indians, were extremely diverse. Malay, Javanese, and Boyanese were spoken among the Malays, whereas in the Indian community, Tamil, Malayalam, Punjabi, Bengali, Gujarati, and Hindustani were used. The Chinese community was the most linguistically diverse, with southern Chinese dialects such as Hokkien, Teochew, Cantonese, Hainanese, and Hakka existing alongside Mandarin.4 As a new nation, it appeared critical to identify a common language within each ethnic group to act as a unifier. Apart from facilitating intra-ethnic communication, the identified common language is also vital for the maintenance of cultural identity. For the Malays and Indians, Malay and Tamil, the two most frequently spoken languages in the community, were chosen to

170  Yeng Seng Goh and Yeow Wah Fong represent the respective group. For the Chinese, by contrast, Mandarin was selected although it was the least widely spoken language among the Chinese during the early days of independence. The choice of Mandarin over other Chinese dialects is attributable to the fact that Mandarin was widely used as a common language among the Chinese diaspora5 (Goh 2009; Kuo 1980). Moreover, Mandarin is a “neutral” language, in which no dialect group has any advantage, thus avoiding disgruntlement among the dialect groups. As compared to English, the functions of these languages are largely confined within the ethnic group. They serve to maintain ethnic distinctiveness and to transmit cultural heritage. Lee believed that the adoption of a monolingual policy in English would have caused Singapore to lose its cultural identity and cultural heritage: “Becoming monolingual in English would have been a setback. We would have lost our cultural identity, that quiet confidence about ourselves and our place in the world” (Lee 2000: 181). Bilingualism: Language Planning and Policy Implementation at the School Level Since 1966, the second year into independence, Singapore has put in place a bilingual policy of studying two languages—English and one’s ethnic “mother tongue language” (i.e. Chinese, Malay, or Tamil). The term “mother tongue(s)” or “mother tongue languages” as used in the context of Singapore does not refer to a first language acquired at home (see Liang, this volume). Instead, it refers to the language associated with a student’s ethnic group, regardless of whether the student had an opportunity to acquire it at home. Hence, if a Chinese student’s first language acquired at home is English, his or her mother tongue language would still be Chinese. Singapore’s language policy is a truly bilingual one as there is no other bilingual education policy in the world that puts together languages of different language families with three orthographic systems. English is an Indo-European language, Chinese is Sino-Tibetan, Malay is Austronesian, and Tamil is Dravidian. They are largely non-cognate languages with virtually no common linguistics features. In terms of orthographic system, English and Malay both use the Roman alphabet, Chinese uses logographic characters (汉字), and written Tamil is based on a syllabic Brahmi system. The bilingual policy was implemented gradually in Singapore. In 1983, it was announced that all schools would use English as the main language of instruction from 1987 onward. This resulted in a shift in the allocation of curriculum time for both English language and the ethnic “mother tongue languages.” English in Schools English occupies a predominant and unshakable position in Singapore’s education system. It is not only the first language taught in school to all

One People, One Nation, One Singapore  171 Singapore students but it is also the main medium of instruction in both humanities and science subjects. English is also the dominant language for formal activities in school. Lee Kuan Yew described the varying proportion of curriculum time for English in Singapore schools as follows: After kindergarten and primary school, English will become their master language. Primary schools have 75 per cent of curriculum time in English. … In secondary schools, 85 per cent of curriculum time is in English. … For tertiary education in Singapore polytechnics and universities, English is the sole medium of instruction. (Lee 2012:19) “Mother Tongue Languages” in Schools Under the current unified national education system in Singapore, all students must learn their ethnic “mother tongue language” as their second language and as a stand-alone subject— Chinese language for Chinese, Malay language for Malays, and Tamil for Indians. Lee Kuan Yew saw a cultural basis for the transmission of the “Mother tongue languages”: Language is more than a tool of communication; it transmits values too. That is why we have insisted that all school-going children learn their mother tongue, whether Chinese, Malay or Tamil, as their second language. This way, they will have the means to tap on the rich heritage that their respective culture has to offer. Folk stories, for instance, are replete with tales of courage, generosity, hard work, honesty and filial piety. These are all good traditional values that are worth teaching and transmitting. By getting children to read these stories in their formative years, we are helping them to imbibe values that will shape their — character. Reading these stories in their mother tongue will also help the children to develop a realization that they are descended from civilizations that are thousands of years old. This will help them develop a sense of identity as to who they are and where they have come from. (Lee 2012: 227) According to Lee, the Chinese language fulfills a cultural function in respect of the Chinese community in Singapore: Why did I insist on Singaporeans all needing to learn their mother tongue? For the Chinese, why did I insist that they learn Chinese? Because I believe language transmits values. Learning the Chinese language means imbibing the core items of Chinese history, tradition and culture. The Confucianist values of loyalty, honor, discipline, filial piety, emphasis on family; respect for authority—all vital for nation-building

172  Yeng Seng Goh and Yeow Wah Fong and for cultivated citizens with honorable personal attributes. These values will provide cultural ballast to our people as we adjust to a fast-changing world. (Lee 2012: 70) The bilingual policy evolved and was developed further after independence in 1965. It has exerted a huge influence on the transformation of the Chinese6 language policy in Singapore, reshaping the learner profile, the teaching pedagogy, the learning environment, and the relevant standards to be set.

The Rise and Dominance of English Over the years, the dominance of English as a working language and common language of communication between all ethnic groups and as a main medium of instruction under the bilingual educational policy has had a direct and huge impact on the pattern and choice of language use within the Chinese community in Singapore. A survey by the Singapore Ministry of Education (MOE) on the most frequently used home languages of Primary One Chinese students7 clearly demonstrated that the home language of Chinese families is shifting from Mandarin to English. The results are shown in Figure 8.1. The percentage of students from Chinese-speaking families whose home language was Mandarin increased from 25.9 percent in 1980 to reach a

Figure 8.1 Most frequently used home languages of Primary One Chinese students. Source: Ministry of Education (2004, 2011).

One People, One Nation, One Singapore  173 peak of 67.9 percent in 1990. The percentage then gradually declined from 55.4 percent in 2000 to 47.3 percent in 2004. At the same time, the percentage of Primary One Chinese students speaking English at home rose from 10.3 percent in 1980 to 26.3 percent in 1990 and to 42.3 percent in 2000. In 2004, for the first time, this percentage exceeded that of students speaking Mandarin at home. In 2009, the percentage increased sharply to 60 percent, and then to 61 percent the following year (Heng 2011). This means that nowadays Singapore primary school Chinese students increasingly come from English-speaking homes. In the recent 2015 General Household survey, English was found to be the most frequently spoken language at home among the younger Chinese population between 5 and 14 and 15 and 24 years of age, an increase from 51.9 percent in 2010 to 61.3 percent in 2015 and from 40.7 percent in 2010 to 50.2 percent in 2015, respectively (Department of Statistics 2016: 20; see also Siemund and Li, this volume).

The Heterogeneity of Ethnic Chinese in Singapore The Chinese population in Singapore is not homogeneous but highly diverse. The flow of Chinese migration into Singapore has taken place over the centuries and is thus multi-layered. Depending on the time of migration, the rate of heritage language attrition, and degree of acculturation, the ethnic Chinese in Singapore can be broadly sub-divided into three groups, namely the Straits Chinese, the Sinkeh Chinese, and the New Chinese immigrants. Straits Chinese As early as the fourteenth and fifteenth century, merchants in Fujian and Guangdong provinces in China had moved to the Malay Peninsula. Some of them intermarried with local Malay women and created a unique identity through the fusion of Malay and Chinese cultures. The males are known as Baba and the women as Nonya. This group of early Chinese immigrants is known as Straits Chinese or Peranakans (Skinner 1996). When Singapore was founded by the British in 1819 as a trading port, many Malacca and Penang-born Straits Chinese emigrated to Singapore. As most of them were fluent in English, they were engaged to work in colonial government offices and foreign firms. Enjoying a higher social status during the colonial era, the Straits Chinese identified with the British culture and switched from Chinese dialects or Baba Malay to English as their main language (Gupta 1994). Following the wave of decolonization after World War II, there was an awakening of political awareness among the Straits Chinese. Whether rich or poor, they claimed to be children of the soil. They actively played a leading role in Singapore’s independence. Lee Kuan Yew is one of the best-known leaders among the Straits Chinese. The affiliation of the Straits Chinese to their ancestral homeland in China has been relatively weak, using mainly lineage, traditional customs and beliefs as identity markers. At present, only

174  Yeng Seng Goh and Yeow Wah Fong a small number of Straits Chinese still speak a dialect or Baba Malay at home (Gupta 1994; Lee 2014). Most use English as their home and working language. They cannot usually converse well in Mandarin, let alone read and write in Chinese, and often have to understand Chinese cultural history through English. “Sinkeh” Chinese During the early colonial period there was an increased labor demand which resulted in large-scale immigration, especially from the southern provinces of China (Yen 1986). The Straits Chinese called these Chinese immigrants “Sinkeh Chinese” (新客, lit. ‘new guests’, i.e. newcomers). They spoke mostly the Chinese dialects of their homeland, including Hokkien, Teochew, Hainanese, Foochow, Cantonese, Hakka, and so on. The Sinkeh Chinese received little education and were mostly illiterate. They were interested only in making a living in Singapore either through trade or by offering their labor. The influx of Sinkeh Chinese was so massive and rapid that within a short period of time, they outnumbered the Straits Chinese. According to the demographics in 1891, Straits Chinese accounted for about 50,000, while there were as many as 175,000 Sinkeh Chinese. According to the Singapore Population Census, the proportion of Chinese born in 1921 and 1931 outside Singapore (mainly China) was as high as 76.4 and 64.2 percent, respectively (Saw 2007). At this time, a large number of Chinese women from China also poured into Singapore, married Sinkeh Chinese and had children. This inevitably led to an ever-increasing number of Chinese in Singapore. Since then, the Sinkeh Chinese and their descendants constitute the majority of the Chinese population in Singapore. At the turn of the twentieth century, the establishment of the Republic of China led to a rise in Chinese nationalism. The new government adopted a nationality law based on the principle of ancestral lineage. All who were of Chinese descent, regardless of place of birth, were considered Chinese nationals. This directly affected the identity of the Sinkeh Chinese and motivated them to re-establish their identification with their homeland (Kuhn 2008). Chinese community leaders and clan associations committed themselves to building Chinese schools and supporting Chinese education. In these schools, students learned Mandarin and Chinese culture and values. They also acquired practical skills and were taught crafts in order to make a living. Through Chinese education, it was hoped that Chinese cultural identity could be fostered and Mandarin would become an essential identity marker of Chinese ethnic identity. It was through Chinese education that Chinese ethnicity was re-defined based on language and culture, that is, speaking Mandarin, writing Chinese, reading Chinese books, and having an understanding of Chinese culture and history (Tan 1998).

One People, One Nation, One Singapore  175 Before Singapore gained independence in 1965, the Chinese-educated community still hoped to maintain a link with their homeland, even though they had identified with the local culture. However, at the peak of Cold War tensions in the region, the political allegiance of the Chinese-educated community—with their self-professed stronger “Chineseness”—was naturally mistrusted by the Straits Chinese and other ethnic groups (Goh 2017). In the early days of Singapore’s independence, many Chinese-educated parents, not wanting their children to be marginalized in a society where English was dominant, began to send their children to English-medium schools. New Chinese Immigrants Since the mid-1990s, with the global trend of attracting talent, people from mainland China and Hong Kong have been encouraged to migrate to Singapore. The recent influx of Chinese from mainland China and Hong Kong formed a local community of New Chinese immigrants. Unlike the early immigrants before independence, who were from the southern provinces of China, most of these came from other parts of China and spoke mainly Mandarin. Although most Chinese in Singapore traced their ancestry to China, several decades of divergence had created different community identities, value systems, and varieties of language among the local and the New Chinese immigrants. It is due to this difference in socio-cultural background that the New Chinese immigrants made an impact on the Chinese community’s sense of “Chineseness.” They often regarded their own culture as more orthodox or legitimate than that of the local Singapore culture and local Mandarin pronunciations. They were also the strongest advocates of promoting standardized Mandarin (Goh 2010). The dissatisfaction expressed by these New Chinese immigrants is a reflection of their complex emotions: physically in Singapore, yet their hearts in their homeland, China. Change in Self-identity: From Overseas Chinese to Chinese Singaporean Over the years, as generations of ethnic Chinese settled and integrated into Singapore mainstream society, there was a significant change in self-identity: from “seeking one’s roots” (落地归根 ‘falling leaf returns to the root’) to “putting down new roots” (落地生根 ‘fall to earth and grow roots’). The first generation of Chinese immigrants often identify themselves as overseas Chinese and have strong sentimental attachments to their ancestral homeland in China. However, the second and third generations no longer share the same emotional bonds to the homeland of their ancestors. They view Singapore as their home. They no longer have the emotional ties to their heritage languages that their ancestors had.

176  Yeng Seng Goh and Yeow Wah Fong Being born, raised, and educated in Singapore, these second and third generations have undergone a common socialization process. They have close contacts with the Malays, Indians, and Eurasians in schools, in their neighborhood, at the workplace, and in public spaces such as community clubs, parks, and semi-public spaces like food centers, movie theaters, and restaurants. All young adult male Singaporeans have even done their National Service together. In addition to learning English as a first language in school, second- and third-generation Chinese receive much greater exposure to English than to Mandarin in the broader social context. The English language is prevalently used in most—if not all—aspects of life within mainstream Singapore society. All these experiences have shaped a distinctive identity different from their ancestors. Second- and third-generation Chinese have constructed their nationality and emotional and cultural belonging to Singapore.

The Generational Evolution of Bilingualism within the Chinese Community The long-term transformation of the language ecosystem of the Chinese community in Singapore has induced an evolution of bilingualism across different generations. It can be broken down into the following stages (Figure 8.2).

Generation First generation Second generation

Third generation

Fourth generation

Childhood 1

a Chinese only a / ab Chinese only or ChineseEnglish bilingualism ab / ba ChineseEnglish bilingualism or EnglishChinese bilingualism ba / b EnglishChinese bilingualism or English only

Changes Emigrate as an adult Having completed their education in Singapore

Adulthood a / Ab Chinese only; or bilingual with Chinese dominant Ab / aB Bilingual, with Chinese as the dominant language or English as the preferred language

Having completed their education in Singapore

Ba / aB Bilingual, with English as the dominant language or preferred language

Having completed their education in Singapore

Ba / B Bilingual, with English dominant; or English only

Figure 8.2 The generational evolution of bilingualism within the Chinese community.

One People, One Nation, One Singapore  177 The first-generation Singaporeans would be monolingual, speaking either Mandarin or one of the Chinese dialects as home language. This first generation would include subsequent adult Chinese immigrants to Singapore who had received their Chinese-stream education in their former homelands (i.e. mainland China and Taiwan) and local-born Singaporeans who had received a Chinese-stream education often prior to the full implementation of the bilingual policy in 1987. The term “Chinese-stream education” refers to an educational policy whereby the Chinese language was taught in schools as a first language and was also used as a medium of instruction for all school subjects. English (or other languages) as a second or foreign language might also have been taught as a stand-alone subject. As a result of a predominantly Chinese educational background with minimal or no exposure to the English language, first-generation Chinese primarily spoke a Chinese dialect and occasionally Mandarin. Under the influence of a three-decade long Speak Mandarin Campaign,8 they began to speak mainly Mandarin at home in place of Chinese dialects in order to give their children a head start in school. By contrast, their offspring, the second generation, were educated under the English-dominant bilingual policy. Being proficient in both Mandarin and English, they speak mainly Mandarin at home to accommodate the language habits of their parents and grandparents. Nevertheless, as products of the English-dominant bilingual education system, they are more comfortable with English and tend to speak the language with their siblings, peers, and subsequently their children, the third generation. This has resulted in a further shift to English in the home language of the current cohorts of Chinese school-age students (Aman et al. 2009). This shift will become increasingly pronounced with the fourth generation. Most fourth-generation descendants will have become strictly monolingual in English, with only a small minority retaining a degree of proficiency in Mandarin or Chinese dialects. But even for them, English will be the preferred language over Mandarin or Chinese dialects and will occupy an absolutely dominant position.

Chinese Language Policy and Planning in the Post-Lee Kuan Yew Era The Repositioning of Chinese Language Planning and Policy To date, the Chinese language has mainly served the function of maintaining cultural identity so as to keep Chinese Singaporeans in touch with their cultural roots and traditional values. People proficient in the Chinese language work mainly as Chinese language teachers and journalists or government departments. To elevate the social and economic status of the Chinese language and make it into a valuable asset for jobs and careers, it is thus urgent to extend “Duìnèi Huáyǔ” (Huáyǔ 华语, ‘the language of the Hua’, i.e. the language of the local Chinese community) for affective

178  Yeng Seng Goh and Yeow Wah Fong or integrative purposes within the Chinese community in Singapore to “Duìwài Hànyǔ” (Hànyǔ 汉语, ‘the language of the Han’, i.e. the language of the broader Chinese community beyond Singapore) for utilitarian or instrumental purposes with the Chinese-speaking world. In other words, the Chinese language in Singapore needs to expand its function from a heritage language (Huáyǔ), which is inextricably linked to the formation and maintenance of Chinese ethnic identity to a trade language (Hànyǔ), for business, trade, and professional knowledge, which, in turn, enhances one’s employability in the global job market. Lee Kuan Yew urged the bilingual Chinese Singaporeans to seize the golden opportunities inherent in China’s rise: “China’s rise presents many opportunities for Singaporeans who are bilingual. Whether working in China or doing business with Chinese companies, those who are comfortable in the Chinese language have an advantage over those who know only English” (Lee 2012: 204). However, as it turns out, the promotion of the Chinese language in school has only a limited effect. It is impossible to achieve the goals of Chinese language teaching without the cooperation of family, society, and the state. The Singapore government should connect Chinese language with more practical issues like job opportunities and promotion to check the decline in the use of Chinese. Teaching Mandarin as an International Language: A Global Perspective The generational evolution of bilingualism in the Chinese community, with a shift in identity from overseas Chinese to Chinese Singaporeans, resulted in a higher percentage of Chinese Singaporeans using English predominantly as a home language. This continued rise of English in Singapore has led some Singapore academics such as Pakir (2001) to argue that Singapore increasingly belongs to the Inner Circle of global English speakers.9 In contrast, in the realm of global Mandarin, Singapore is gradually shifting from the Outer Circle to the Expanding Circle of Mandarin users,10 where Chinese is increasingly learned as a foreign language and is not used in the home (Goh 2017). The daunting challenge now and in the years to come is to find ways to reverse the current decline in the use of the Chinese language in Singapore.

Concluding Remarks: A Never-Ending Journey The nation-building processes in the past 50 years and more have indeed fostered a Singaporean national identity. The introduction of English as a common language across different ethnic groups has united the population to become “One people, One nation, One Singapore.” However, Lee Kuan Yew did not anticipate that English would become so dominant and prevalent in mainstream Singapore society. This is an unintended outcome of

One People, One Nation, One Singapore  179 a language policy that had originally aimed at striking a balance between differentiated functions of English and Chinese. The contradictory trend between the internal dominance of English in Singapore on the one hand and the external re-emergence of China with the growing importance of mastering Mandarin on the other hand means that it is becoming increasingly pressing to tackle the various challenges associated with Chinese language teaching in Singapore. As Lee Kuan Yew pointed out, the development of language policy and planning is “a never- ending journey”: “Singapore’s bilingual language policy has undergone many refinements and will continue to do so in keeping with developments in the wider world” (Lee 2012: 231). For a nation with no natural resources and whose people are its key resource, it is timely for Singapore language policy makers to think ahead in the post-Lee era to ensure Singapore’s continuous relevance to the changing global environment.

Notes 1 There is no specific government agency that oversees language planning in Singapore. Most, if not all, of the language policies were rationalized by Lee Kuan Yew based on his understanding of Singapore’s national interest. 2 According to the 1957 census, English was spoken by 1.8 percent of the population (Chua 1964). 3 Singapore is also the only nation in Southeast Asia that has adopted the colonial language as dominant language. 4 Though not a home language for most Chinese, Mandarin was learned in the Chinese-medium schools (see Siemund and Li, this volume). 5 This is largely due to the promotion in China of Guóyǔ before 1949 and Pǔtōnghuà after 1949, whose pronunciation is based on the Beijing dialect of Mandarin. 6 In the context of this chapter, no distinction is made between “Mandarin” and “Chinese.” 7 This is a profiling survey conducted every year for all first-year students (aged seven years old) enrolled in mainstream schools in Singapore. 8 The Speak Mandarin Campaign was first launched in 1979. Its initial objective was to encourage Singaporean Chinese who could only converse comfortably in their respective Chinese dialects to adopt Mandarin as the common language within the Chinese community. See Goh (2010) and Siemund and Li (this volume) for a discussion of the Speak Mandarin Campaign. 9 The Inner Circle of English refers to the traditional base of native speakers which include Britain, the United States, Canada, Australia, and New Zealand (see Kachru 1989; Meierkord, this volume). 10 Goh (2010, 2017) modified Kachru’s model of the three concentric circles of global English, namely, the Inner Circle, the Outer Circle, and the Expanding Circle, to represent the spread of global Mandarin. The Inner Circle of global Mandarin refers to the traditional base of native speakers, which include mainland China and Taiwan where Mandarin is the dominant language in society. The Outer Circle represents overseas Chinese communities around the world where Mandarin is not the dominant language of their countries of residence but is usually confined to the home domain or within the Chinese community. The Expanding Circle refers to those regions of non-native users, such as Japan, South Korea, North America, and Europe, where Mandarin is learned and spoken as a foreign language.

180  Yeng Seng Goh and Yeow Wah Fong

References Aman, Norhaida et al. 2009. The Sociolinguistic Survey of Singapore 2006. Singapore: National Institute of Education. Chen, Ping. 1999. Modern Chinese: History and Sociolinguistics. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139164375 Cheong, Danson. 2018. “China to Issue 5-Year Visas for Foreigners of Chinese Origin.” The Straits Times January 31, 2018. Accessed October 22, 2018 via https://www.straitstimes.com. Chua, Seng Chew. 1964. “State of Singapore: Report on the Census of Population, 1957.” Singapore: Department of Statistics, Government Printing Office. Goh, Yeng Seng. 2009. “Bilingual Education Policy in Singapore: Challenges and Opportunities.” In Language Teaching in a Multilingual World: Challenges and Opportunities, edited by Christopher Ward, 171–90. Singapore: SEAMEO Regional Language Centre. ——— 吴英成. 2010. Hànyǔ guójì chuánbō: Xīnjiāpō shìjiǎo 汉语国际传播：新加坡 视角 [The globalization of Chinese: A Singapore perspective]. Beijing: Shāngwù yìnshūguǎn. ———. 2017. Teaching Chinese as an International Language: A Singapore Perspective. Cambridge: Cambridge University Press. doi:10.1017/9781107280472 Gupta, Anthea Fraser. 1994. The Step-Tongue: Children’s English in Singapore. Clevedon: Multilingual Matters. Heng, Swee Keat. 2011. “Speech by Mr Heng Swee Keat, Minister for Education, at the Singapore Chinese Chamber of Commerce & Industry (SCCCI) Mid-Autumn Festival at the SCCCI auditorium on 9 September 2011.” Accessed on March 1, 2012 via http://www.moe.gov.sg Kachru, Braj Bihari. 1989. “Standards, Codification and Sociolinguistic Realism: The English Language in the Outer Circle.” In English in the World: Teaching and Learning the Language and Literature, edited by Randolph Quirk and Henry G. Widdowson, 11–30. Cambridge: Cambridge University Press. Kuhn, Philip A. 2008. Chinese among Others: Emigration in Modern Times. Singapore: NUS Press. Kuo, Eddie C.Y. 1980. “The Sociolinguistic Situation in Singapore: Unity in Diversity.” In Language and Society in Singapore, edited by Evangelos A. Afendras and Eddie E.C. Kuo, 39–62. Singapore: Singapore University Press. Lee, Kuan Yew. 1965. “Transcript of a Press Conference Given by the Prime Minister of Singapore, Mr. Lee Kuan Yew, at Broadcasting House, Singapore, at 1200 Hours on Monday 9th August, 1965.” Accessed May 17, 2018 via http://www.nas. gov.sg/ archivesonline. ———. 2000. From Third World to First: The Singapore Story, 1965–2000. Memoirs of Lee Kuan Yew. Singapore: The Straits Times Press. ———. 2012. My Lifelong Challenge: Singapore’s Bilingual Journey. Singapore: Straits Times Press. Lee, Nala Huiying. 2014. “A Grammar of Baba Malay with Sociophonetic Considerations.” PhD diss., University of Hawaiʻi. Ministry of Education. 2004. “Refinements to Mother Tongue Language Policy.” January 9. Accessed August 10, 2018 via https://www.nas.gov.sg/archivesonline. ———. 2011. “Nurturing Active Learners and Proficient Users. 2010 Mother Tongue Languages Review Committee: Executive Summary of Recommendations.” Accessed August 10, 2018 via http://www.moe.gov.sg.

One People, One Nation, One Singapore  181 Pakir, Anne. 2001. “The Voices of English-Knowing Bilinguals and the Emergence of New Epicentres.” In Evolving Identities: The English Language in Singapore and Malaysia, edited by Vincent B.Y. Ooi, 1–11. Singapore: Times Academic Press. Saw, Swee-Hock. 2007. The Population of Singapore. 2nd ed. Singapore: Institute of Southeast Asian Studies. doi:10.1355/9789812307392 Singapore Department of Statistics. 2011. “Singapore Census of Population 2010: Statistical Release 1: Demographic Characteristics, Education, Language and Religion.” Accessed October 20, 2018 via https://www.singstat.gov.sg. ———. 2016. “General Household Survey 2015.” Accessed October 20, 2018 via https://www.singstat.gov.sg. Skinner, G. William. 1996. “Creolized Chinese Societies in Southeast Asia.” In Sojourners and Settlers: Histories of Southeast Asia and the Chinese, edited by Anthony Reid, 50–93. Sydney: Allen and Unwin. Tan, Chee Beng. 1998. “People of Chinese Descent: Language, Nationality and Identity.” In The Chinese Diaspora: Selected Essays, edited by Ling-Chi Wang and Guangwu Wang, vol. 1, 29–48. Singapore: Times Academic Press. Yen, Ching-hwang. 1986. A Social History of the Chinese in Singapore and Malaya: 1800–1911. Singapore: Oxford University Press.

Part III

Multilingual Practices

9

Speakers of “Mother Tongues” in Multilingual China Complex Linguistic Repertoires and Identity Construction Sihua Liang Abstract This chapter examines language attitudes and identities of migrant students at two primary schools located in the southern Chinese city of Guangzhou. Guangzhou has historically been known for the high vitality and wide use of Cantonese. Due to the city’s great economic success in the past few decades, it has been a popular destination for businessmen and migrant workers from all over the country, which turned Guangzhou into a linguistically superdiverse city. Reflecting on notions of “mother tongue” and hometown language in narratives of second-generation migrant students in Guangzhou elicited through ethnographic fieldwork, the chapter identifies mismatches between the participants’ regional identities, shaped by a sense of belonging to an ancestral place, and “mother-tongue” competence. It is argued that the relationship between language and ethnolinguistic identity demands radical reconceptualization due to fundamental changes brought about by mobility and superdiversity in modern societies as well as, in the Chinese case, by the rigorous promotion of Pǔtōnghuà.

Introduction During a sociolinguistic conference in China, the discussion on a paper about language policy and language ideology in Hong Kong became a bit overheated.1 One attendee raised his voice and asked the presenter, “Forget about that. Can you give a clear definition of ‘mother tongue’ from a purely linguistic perspective?” The presenter, probably feeling intimidated, threw back the question instead of answering it, which further fueled the confrontation. The status of Chinese regional dialects and the relations between these dialects and Pǔtōnghuà (the standard variety) are delicate issues that can spark debate or even conflict at many levels, from the critical review of language education policies such as the case of Hong Kong (Liú 2017; Tián 2012; see also Li and Tong, this volume) to the daily construction of national and regional identities (such as the cases to be discussed in this chapter). Controversy is often stoked in the name of defining one’s mother tongue, which is the focus of this chapter.

186  Sihua Liang In retrospect, the question raised at the conference reminds us precisely why it is so problematic to arrive at a definition of “mother tongue.” On the one hand, a “purely linguistic” definition of “mother tongue” is impossible, because the meaning of mother tongue is socially and historically constructed and highly context-dependent. Some have argued that the concept was infused with political and ideological intentions, in the name of linguistics, from the very beginning of its making (Hutton 1999), a history that we should not forget. In turn, the construction and penetration of the term “mother tongue” into social life change our perceptions of language, nation, and identity in profound ways that might have been thus far understated. On the other hand, “mother tongue” is in all evidence an imported term popularized only in the past two decades or so, which is a phenomenon that has hardly been critically examined in the literature. What terms were used before the introduction of “mother tongue”? What terms are used currently in daily life and how do they compare to “mother tongue”? This chapter addresses the historical and theoretical underpinnings of the term mother tongue in general and the empirical and ideological problems of using the term in the context of contemporary China in particular. I draw on empirical data derived from an ethnographic project on language attitudes and language use from 2009 to 2010 in Southern China. By combining the theoretical, historical, and empirical perspectives in a critical discussion of the concept mother tongue in the context of modern China, I hope to gain a better understanding of the problems surrounding the concept per se and its application in the sinophone world as an ethnolinguistic identity marker.

The Conceptualization of Mother Tongue Mother Tongue as an Umbrella Term Rampton (1990) observed that terms like native speaker and mother tongue seemed to be particularly resilient despite widespread dissatisfaction and criticism. He proposed alternatives such as language expertise, language inheritance, and affiliation in the context of education. These alternatives distinguish three most commonly (mis)conceived dimensions of “mother tongue”: language competence, ethnic or ancestral “root,” identity and sense of belonging. Rampton argues that by replacing “native speaker” and “mother tongue” with more appropriate alternatives, we may be able to think about complex sociolinguistic situations more clearly and critically. Rampton is certainly right in criticizing the muddle surrounding “mother tongue,” but decades have passed since and the use of “mother tongue” in the field of linguistics is still prevalent. More alternatives or related synonyms have been in use though, such as first language, home language, community language (Lytra and Martin 2010), and heritage language (Wiley 2009), sometimes interchangeably. With such proliferation, “mother tongue” seems to include even more dimensions such as the age, order and context

Speakers of “Mother Tongues”  187 of acquisition, ancestry and nationality, language competence, frequency of use, symbol of identity. In cosmopolises where mobility and linguistic superdiversity are the norm (Blommaert 2013a), the individual may find that no single language in his repertoire fulfills all the dimensions—he might have multiple identities, shifted group memberships, or simultaneously acquire several languages with varied and variable competence (Annamalai 2006). “Mother tongue” serves as an umbrella term for the series of terms mentioned earlier, and its strategic ambiguity partly accounts for its sustained popularity in the field. Mother Tongue as an Ideological Notion It has been argued that if a term is used frequently, it must have captured something “real,” so that not much is to be gained if it is discarded (Kandiah 1998). I would argue that the “real thing” sustaining the vitality of the notion of “mother tongue” in the field of sociolinguistics and beyond lies in its ideological value. The key element “mother,” which is “hardwired” into the term, endows “mother tongue” with “natural” symbolic power: “the language that a child learns … in processes of the basic and primary socialization … which … are symbolized by one’s mother” (Skelin Horvat and Muhvić-Dimanovski 2012: 494). In such light, the label seemingly conveys a “natural” and “fundamental” connection to life, the earth and therefore identity, which is socially and historically constructed but rarely acknowledged. Hutton (1999) scrutinizes the links between the rise of the term “mother tongue” and Nazi scholarship, unpacking many assumptions and politics underlying what seemed to be pure “linguistic” arguments. He argues that the notion of the “mother tongue” was key to the Nazi conceptualization of a strong, pure, unified German nation. Loyalty to mother tongue was deemed natural and necessary for unifying an otherwise endangered German diaspora. The underlying assumption was that “[o]nly if the mother-tongue was the same as the father-tongue, both literally (the biological father) and metaphorically (the language of the state), and if the borders were secure, could assimilation be avoided” (1999: 6). Hutton demonstrated convincingly that “Mother-tonguism” is a linguistic and political ideology, which must be seen in social and historical contexts. Linguistics as a discipline has been deeply involved in reconstructing the notion of mother tongue, projecting it back onto history and making it appear to be something natural and universal. If the history of Nazi scholarship sounds a bit far-fetched, the reader may recall that it is still not uncommon to find people’s mother tongue defined by patrilineal ancestry, especially if it is officially designated. This is why the “mother tongues” of Singapore are also sometimes called “father tongues,” which the “mother-tongue speakers” may not be able to speak before they go to school (Gupta 1997). While the meaning of these “father tongues” is not exactly the same as proposed by Hutton (1999), the idea of patriarchy and ideological engineering of linguistic differentiation looks familiar.

188  Sihua Liang In a sense, the current language conflict in China between the promotion of Pǔtōnghuà and the defense of regional dialects is a mirror image of the ideological conflicts between the “father-tongue” and “mother-tongues.” Lǐ Yǔmíng 李宇明 (b. 1955), a distinguished linguist and language planner in China, proposed a conceptual distinction between mǔyǔ (母语, lit. ‘mother language’, which is commonly considered the Chinese equivalent of ‘mother tongue’) and mǔyán (母言, lit. ‘mother dialect’) (Lǐ 2003). He argued that  “language” (the yǔ of mǔyǔ) refers to the common language of the nation (民族), while “dialect” is the regional variation of language. Concerning the motivation for making such distinction, Lǐ wrote with an extraordinary degree of ideological explicitness: A correct representation of mother tongue is socially significant especially in contexts where the regional dialects are relatively strong, and the speech community is relatively autonomous. Take the issue of a person’s right to use their mother tongue, for instance. Since “mother language” covers and stands for “mother dialect,” the right to one’s mother tongue, therefore, primarily means the right to use the common language of the nation, rather than the right to the “mother dialects.” Hence, the “mother tongue education” or “native language education” being advocated today across the world should be seen as the provision of education in the medium of the national common language, rather than in regional dialects (“mother dialect”). (Lǐ 2003: 52) Lǐ lays bare the political and ideological agendas inherent in the definition of mother tongue, which are usually more often subtle or hidden (Shohamy 2006). In other words, a linguistic variety cannot be defined as one’s mother tongue because the politics of nation building prohibits it. Since regional dialects are not mother tongues according to Lǐ’s definition, the right to mother tongue education excludes them. Pǔtōnghuà as the “father tongue,” the common language of the nation, is recast as the only legitimate mother tongue for sinophones in China, while mother tongue education becomes synonymous with the promotion of Pǔtōnghuà. When Lǐ puts forward his argument, he writes in a matter-of-fact tone: The mother tongue of children of Chinese descent is definitely Huáyǔ (Mandarin) rather than the various Chinese dialects. As for Taiwan … the medium of instruction should be the common language of the nation (i.e. the so-called Guóyǔ in Taiwan), which is mandated by people’s right to education in the mother tongue. (Lǐ 2003: 52) However, regarding the definition and recognition of mother tongue, nothing can be taken for granted. Where identity politics and the rhetoric of

Speakers of “Mother Tongues”  189 national identities vary, the definitions of mother tongue and mother tongue education are sometimes diametrically opposed in the sinophone world. The Mother Tongue Movement in Taiwan is inextricably linked with “deSinicization” or “Taiwanization,” and oppositional party politics (Curtin 2007; Klöter 2004). The aim of the movement was to promote all the major local languages to be recognized as national languages, which serve to construct a Taiwanese identity distinct from a Chinese identity (see Chen, this volume, for a detailed review of the history and ideological conflicts of mother-tongue-related language policy in Taiwan). Goh and Fong (this volume) track the changes of attitudes towards Mandarin and other Chinese varieties in the heterogeneous ethnic Chinese community in Singapore, in relation to Singaporean nation building, different waves of migration, and the rise of China as a significant global economic power (see also Siemund and Li, this volume). With the shift of identity from Overseas Chinese to Chinese Singaporeans, the majority of the younger generation barely speak Chinese and use English predominantly as a home language. Mandarin, the officially defined “mother tongue,” becomes more or less symbolic rather than the “real” linguistic resources for identity construction. Chinese (Mandarin) is increasingly learnt as a foreign language, which is an ideological conflict challenging to reconcile for Singaporean language planners. The Concept of Mother Tongue in the Chinese Context As can be seen from the earlier discussion, issues concerning mother tongue in the sinophone world are as complicated and politically sensitive as in other contexts. However, it was during my ethnographic fieldwork from 2009 to 2010 in Southern China when I realized that a problematic aspect of “mother tongue” in the Chinese context has thus far been overlooked in the research literature primarily written in English. Namely, the correspondence between the English term “mother tongue” and its Chinese equivalent mǔyǔ has been taken for granted. In fact, such a problem is not unique to “mother tongue.” The equivalence between “dialect” and fāngyán (方言) is no less problematic (Mair 1991). When interviewing participants during my fieldwork, I encountered a great number of “methodological inconveniences.” Lengthy discussions on places of origin and mother tongues often occurred, sometimes to the annoyance of the participants. As the fieldwork proceeded, it became clear that the methodological inconveniences were signs that traditional and “Western” conceptualizations of “mother tongue” did not match the local lived experiences. Although “mother tongue” has become a catchphrase in the discourse of language planners, language activists, and the mass media, ordinary people talk about places of origin and local languages or dialects in other terms. A search using mǔyǔ as the keyword in one of the most comprehensive archives of historical newspaper and journal articles2 revealed that there

190  Sihua Liang were only four entries (as of October 2019) containing the Chinese characters 母语 for mǔyǔ, but none of them conveyed the meaning of “mother tongue.” The content was about what someone’s mother (mǔ) said (yǔ) on certain occasions.3 The two characters have been used as one word for the equivalent of “mother tongue” since the 1980s. Interest in the topic of mother tongue has increased significantly since 2000. The first few articles were concerned with language transfer from the mother tongue (Chinese) in foreign language learning, while issues about language policy and language ecology (the situation of regional dialects) came up a couple of years later. Based on the search results, the introduction of mǔyǔ as the Chinese equivalent of “mother tongue” was a relatively recent development, and it has attracted increasing attention in the past two decades. This increase in attention coincided with a significant international event—the launch of the International Mother Language Day (IMLD) in 1999. Based on the Baidu search engine, the most frequently used search engine in China, the number of news items about the IMLD has risen steadily in the past two decades. It is increasingly associated with the language rights movement, rights to mother tongue education. Wherever strong regional dialects exist, the status of dialects is heatedly debated (Gao 2012; Liang 2014; Xiong 2018). As “mother tongue,” or rather mǔyǔ, is still not part of the ordinary, everyday repertoire, whenever someone uses it in conversation, it may be seen as a deliberate display of stances on mother-tongue-rights issues. This also accounts for the confrontation described at the beginning of this chapter. What was normally used in China before mǔyǔ to refer to concepts comparable to “mother tongue” in the English world? One possible candidate is xiāngyīn 乡音 ‘hometown pronunciation’. A well-known poem written in the Tang dynasty (618–907 CE) has a line about an old man returning home after many years, realizing that “the local accent [xiāngyīn] has not changed” (乡音无改). The morpheme xiāng (乡) is often used in compound words to express nostalgic emotions toward one’s hometown, kinship, and “nativeness” such as xiāngchóu 乡愁 ‘nostalgia’, yuánxiāng qíngjié 原乡情结 ‘cultural hometown complex’, and xiāngqīn 乡亲 ‘fellow villager’. On the other hand, people have generally referred to linguistic variations within the Chinese language as differences in yīn 音 (literally ‘sound’, but may be interpreted as ‘accent’), whether these were phonological, lexical, or syntactical differences. One well-known saying in this respect is shílǐ bùtóng yīn 十里不同音 ‘the pronunciations become totally different ten miles apart’. During the fieldwork, I noticed that the most frequently used term for concepts similar to “mother tongue” was jiāxiānghuà 家乡话 ‘hometown speech’, which seems to be the contemporary counterpart of xiāngyīn. In conversations about hometown speech, a strong sense of the entanglement of place and identity emerges. An investigation of the meaning of “mother tongue” in contemporary multilingual China is also an investigation of how they orient to place linguistically (Johnstone 2004), through languages or languaging (Li 2011).

Speakers of “Mother Tongues”  191 In the remainder of this chapter, I present the results of an empirical investigation of the notion of “mother tongue,” as it is perceived and negotiated by young multilingual Chinese speakers. The data are based on the aforementioned ethnographic fieldwork, which focused on language attitudes and identities in two primary schools in Guangzhou, a linguistically superdiverse city in Southern China (Liang 2013). I looked into how the speakers discursively negotiate their cultural connection with places, particularly in the cases of second-generation migrants. By analyzing the co-constructed narratives, I explored the diverse meanings and ways of being “mother-tongue” speakers in a multilingual city—the problems in construing what is mother tongue and what is their mother tongue, as well as the entitlements, constraints, and prices of various kinds of identity work (Blommaert 2013a).

Mother Tongue versus “Hometown Speech”: Language and Place Internal Migration and the Changing Language Ecology Before going into details of the empirical project, I will give a brief review of the historical and ideological contexts necessary for understanding the indexical meanings of the data. However, this is not intended to be a comprehensive summary of China’s sociolinguistic history. For that purpose, the readers may refer to Ramsey (1987), Chen (1999), and Liang (2014). Since the political and economic reform in the 1980s, urbanization and internal migration in China have been happening at a phenomenal speed. While significantly weakened, the household registry system known as hùkǒu (户口; Chan and Zhang 1999; Mackenzie 2002) still regulates internal migration by controlling migrants’ access to valuable resources such as social insurance and education. Suffice to say that it is still extremely difficult to move one’s hùkǒu to highly sought-after cities such as Beijing, Shanghai, and Guangzhou, and people who physically move despite such institutional constraints show a strong desire to move socially and economically upward. In the official census, people are officially defined as migrants (or non-hùkǒu population) if their hùkǒu is not registered in the region where they are resident. Hùkǒu status and jíguàn (籍贯, official ancestral place of origin) are institutionally salient census identities (Kiesling 2006) imposed on all Chinese citizens. As the capital city of Guangdong Province, the most economically advanced province in Southern China, Guangzhou is one of the top destinations for opportunity-seeking migrants from other parts of the country. From 1983 to 2011, the total population of Guangzhou more than doubled from 5.2673 million to 12.75 million, while the migrant population (as defined earlier) grew from less than 1 percent to well over one-third (Guangzhou Bureau of Statistics 2001). Early migrants arrived with diverse

192  Sihua Liang regional Chinese dialects and plunged headlong into the sea of Cantonese, a strong local dialect of Guangzhou and the regional lingua franca (Snow 2004). Many of those arriving before or during the 1990s quickly learned Cantonese to make a living. The unparalleled popularity of Cantonese pop culture produced in Hong Kong during that period (Gold 1993) greatly motivated and facilitated their language acquisition. Meanwhile, the central government had been forcefully and systematically promoting Pǔtōnghuà, the national “common speech” and a standard variety based on northern Mandarin (see the chapters of Simmons and Klöter in this volume). Guangdong was considered to have failed in the promotion of Pǔtōnghuà and authorities renewed their determination (GDCCP and GDGOV 1992). Into the new millennium, a whole raft of language laws and regulations (Liang 2014), including the Law of the People’s Republic of China on the National Commonly Used Language and Script (中华人民共和国国家通用语言文字法; CPG 2005), propelled the Pǔtōnghuà promotion campaign to new heights. Comparing two surveys on language use and attitudes in the Pearl River Delta region and Guangzhou in 1993 (Leung 1993) and 2005 (Táng 2006), respectively, Liang (2014) suggests that the language situation in Guangzhou has changed from Cantonese-dominant multidialectalism to a loosely diglossic situation (Ferguson 1959; Fishman 2002) in which Pǔtōnghuà serves as the high variety. Functional differentiation between Pǔtōnghuà and Cantonese (and other regional dialects) may be observed but is not clear-cut. P ǔtōnghuà advances into many private and everyday domains. Schools now teach almost exclusively through Pǔtōnghuà. The service industry uses Pǔtōnghuà as a golden standard while Cantonese service is a bonus rather than a must. Announcements on public transportation such as the MTR usually come in Pǔtōnghuà first, Cantonese second, and English third. Occasionally the Cantonese version went missing, but the company would bring it back if there were too many complaints. Compared to many other Chinese regional dialects (except Mandarin dialects, Shanghainese, and certain Mǐn dialects such as Taiwanese), Cantonese still plays a considerably more visible part in public domains, but it is no longer a required skill for living in Guangzhou. Pǔtōnghuà now provides access to almost all social resources. Data Collection and Analysis The cases referred to in this chapter are part of a large linguistic ethnographic study (Maybin and Tusting 2011) of language attitudes and identity constructions of young people in Guangzhou, which started in 2009. The first period of intensive fieldwork was between 2009 and 2010 when the participants were primary school students. I have since kept in touch with some of the key participants and re-interviewed them in 2015. Forms of data include field notes, interviews and focus group discussions, informal surveys, photographs, and social media interactions. The original study involved

Speakers of “Mother Tongues”  193 participants of diverse ethnolinguistic backgrounds and mobility experiences, while this chapter reports on the cases of two second-generation migrant students, born in the late 1990s or later. The participants were in year five, pupils of the GE School (a pseudonym created by the author) at the onset of the study. The school is situated in Tiānhé District 天河区, the new urban center, where the migrants account for over 45 percent of the total population according to census data (THGOV 2016) and is a well-known Pǔtōnghuà-speaking region. In previous language and identity research on China, such as Dong’s (2009) study of rural and urban identities in a Beijing school, the migrant students often belong to an underprivileged group, whose parents have low-paying jobs and are unable to obtain the local hùkǒu, a socially and economically valuable resource. In contrast, most “migrant students” in this school are not migrants according to their census identity, as their families manage to obtain a Guangzhou hùkǒu and their families are seemingly well off. This is a major difference between the current and previous studies on migrant identities in China. Case One: Connections and Conviction As mentioned earlier, one methodological inconvenience I encountered in the field was how to ask the participants about their mother tongue in Chinese. As mentioned before, mǔyǔ is not an everyday term. To work around the term “mother tongue,” I asked a range of other questions such as their parents’ ancestral homes, their birthplace, and their first dialect and home languages. It turned out that most interviewees used jiāxiānghuà ‘hometown speech’ to refer to the regional dialects of their parents’ hometown. By talking about “hometown speech,” language and the sense of place and belonging are inextricable. In other words, the participants made decisions about their regional identities as soon as they talk about “hometown speech.” In modern societies where mobility and superdiversity are the norm, however, sense of place is highly ambiguous. A logical follow-up question is the definition of hometown. The question “Where are you from?” was asked in Cantonese (Nei5 hai 6 bin1 dou 6 jan4?) and Pǔtōnghuà (Nǐ shi nǎli rén?), the two main media of interaction with the participants.4 A typical answer to the question takes the form of 我是x人 ‘I’m a x person’, and x is the place that connects with their identity. People from the Hakka and Teochew-speaking areas are likely to answer “I am a Hakka/ Teochew person.”5 Similar to Cantonese, Hakka and Teochew are labels for the ethnic languages, the ethnic groups, and the main geographical regions for the groups. The Hakkas and Teochews in Guangdong are known for their strong loyalty to their languages and for maintaining close in-group social networks even in diaspora (Yan 2006: 167). For example, C, a secondgeneration migrant boy in the GE Primary School, whose parents are from

194  Sihua Liang the Teochew region, did not hesitate to identify himself as a Teochew p erson, and he explained his rationale eloquently in his interview as follows:6 C: Some classmates say that people’s AUTHOR: What do you think? C: I don’t agree with them. AUTHOR: So, what’s your opinion? C: It is wherever the jiāzú is.

birthplaces are their hometowns.

AUTHOR: So, where do most members of your jiāzú live? C: Chaozhou. Two thirds of the population live in Chaozhou.

Err…three fourths, actually, because only my father, his brother and the sons of some relatives have left (Chaozhou) to work elsewhere. Very few of them live in Guangzhou. People from Chaoshan (the Teochew Swatow region) who are our relatives (and live in Guangzhou) … very few.

The term jiāzú (家族) fuses the small families into the larger clan, traditionally based on patrilineal lineages. Jiāzú denotes unbroken ancestry and gives a sense of rootedness to all its members. Those who have left the “original” location of their jiāzú often maintain a social network of Teochew people by doing business with or marrying Teochew people. Returning to one’s hometown regularly is an important ritual and the peak time is in spring, when thousands of Teochew people travel from afar to sweep their ancestors’ tombs, despite gridlocked roads and extreme difficulties in getting public transport tickets. For C, every summer vacation is spent in his Teochew hometown to brush up his Teochew dialect and to remind him of his identity. C’s mother put particular emphasis on his ability to speak the Teochew dialect during the home visit: Some children are from Teochew but they do not understand the Teochew dialect. I don’t think that is appropriate. We should still teach them…teach them how to speak it. I’m okay with other things, but the Teochew dialect must be taught. In Mrs. C’s and C’s narratives, place, specifically the patrilineal hometown, is constructed as the key of “hometown speech,” while competence in the hometown speech, ethnic identity, and ancestral heritage form a seamless core, to which everything else such as birthplace and new languages (Pǔtōnghuà, Cantonese, English) are regarded as peripheral. Wherever they go, the place they culturally identify with is always the Teochew region, and they manifest their cultural identity by speaking the Teochew dialect. Case Two: Ambivalent Mother Tongues and Ambivalent Identities Not everyone shares C’s conviction. Due to large-scale migration and increasingly common intermarriage, many participants faced choices between

Speakers of “Mother Tongues”  195 their  paternal hometown, maternal hometown, place of birth, hùkǒu, and other personal factors such as place of upbringing. Some of them switched ethnic identities in different situations (Skelin Horvat and MuhvićDimanovski 2012), and some found the question confusing, annoying, or intimidating. Q was a typical example. Her parents are originally from different provinces and speak different regional dialects as their first dialect. Q’s self-claimed first language was Pǔtōnghuà, which she spoke without recognizable regional accent. The following conversation took place during the first interview with her, which was carried out in Pǔtōnghuà. AUTHOR: Where is hometown? Q: You mean Dad’s or Mom’s? AUTHOR: It’s up to you. You can talk about both. Q: Mom is from Guiyang (Guizhou Province). Dad is from Teochew. AUTHOR: Then which do you think is your hometown? Q: I don’t know. AUTHOR: Then where do you think you are from? Q: Me? Probably Guiyang then, because I like Guiyang better.

The question of hometown seemed to be an open question for Q. She was aware of how others conventionally defined a person’s hometown and asked me for clarification. As I refrained from confirming the conventions, she expressed her uncertainty about her place of belonging. Her preference for Guiyang was vague and general but her reason for not liking Teochew, her father’s hometown, was specific.7 “The Teochew dialect is very unpleasant to the ear,” she said. She claimed that she could not understand a word in the Teochew dialect but could speak a little of her mother’s Guiyang dialect. It seemed as though she was inclined towards Guiyang as her hometown and indeed most of her classmates knew her as a Guiyang person. In a private conversation (in contrast to the recorded interview), she gave a pragmatic reason for being a “Guiyang person.” A few years ago, her parents changed her ethnicity from Han Chinese to Tǔjiā ethnic minority (土家族), since her mother belongs to a Tǔjiā ethnic minority, one of the national minority groups in Guizhou Province. National minority students have advantages in the national college entrance exam, which would give her advantage. Her behavior during a focus group discussion,8 however, displayed a different sense of place belonging. During the focus group discussion, one of the topics was how to decide others’ place of origin. It was a question that I had already asked in the individual interviews. I brought it up again to see how the group would approach the matter. After a while, they appeared increasingly impatient. Y, an outspoken girl, protested: “Why do you always ask people where they are from? Every question is about where somebody is from.” Q, who was usually timid, seemed encouraged and hammered on the table, shouting, “Where from! From here!” Q’s rare outburst underlined her desire to be identified as a Guangzhou person, a local. Reviewing the recording of the

196  Sihua Liang focus group session afterward, I found that all the girls had claimed to be a Guangzhou person in their “off-the-record” remarks. It dawned on me that during the first interview, by saying “you can talk about both,” I had already ruled out Guangzhou as an option. However, the lack of language competence in Cantonese has seemingly become a hindrance for Q to claim local identity. Her parents mentioned the same theme during the home visit, describing their daughter as a “failure” because she was not able to speak any regional dialects. Q said, “They (parents and relatives) said before that I am a Guangzhou person, but they also say, ‘why are you not able to speak Cantonese given that you are a Guangzhou person?’” On various occasions, Q brought up the issue of Cantonese competence as soon as some form of Guangzhou identity was mentioned, whether it was in the one-to-one interviews or during the focus group discussions (adapted from Liang 2014: 76): F: I can speak standard (Pǔtōnghuà), BUT: I am a Guangzhou Q: But I can’t speak it. AUTHOR: What is it that you are not able to speak? Q: I can’t speak Cantonese.

person.

In this episode, F, who spoke Pǔtōnghuà as her first language, was trying to overturn the stereotype that people in Guangzhou could not speak standard Pǔtōnghuà. It is remarkable how Q twisted the argument so that a person could be a Guangzhou person without speaking the local dialect, Cantonese. Putting together Q’s comments on her relationships to various “hometowns” and the related dialects, we find that her construction of her regional identities is multiple, fluid, ambivalent, and highly dependent on situation. On certain occasions, she recognized the social and economic advantage of belonging to a national minority and thereby a Guiyang person. When that advantage was temporarily irrelevant, she navigated among multiple identities. In particular, she attempted to claim Guangzhou identity on various occasions—when she argued for the validity of people’s self-claimed regional identity, claiming to be “from here,” and when she twisted F’s statement on Guangzhou residents’ Pǔtōnghuà proficiency into an argument against the mandatory connection between Cantonese proficiency and the “Guangzhou person” identity. She seemed to be constantly at issue with the notable mismatch between a sense of belonging to Guangzhou and Cantonese as the only “legitimate” mother tongue that gives such a sense of belonging. As a result, while she seemed to prefer the “Guangzhou person” identity, Q had strategically adopted an ambivalent attitude toward the construction of her regional identities. Case Three: Pǔtōnghuà as the “Step-Mother Tongue” In the eyes of many Cantonese language activists, Tiānhé District is “hopeless” because nearly half the population are migrants (THGOV 2016) whose

Speakers of “Mother Tongues”  197 mother tongue is not Cantonese, while the other half seem too willing to communicate in Pǔtōnghuà. Incidentally, 46 percent of the 39 students in the GE School Class claimed that Pǔtōnghuà was the first variety they learned at home. In previous studies on language attitudes and language use among students in Guangzhou, Táng (2006) reported that 19 percent of the participants spoke Pǔtōnghuà as their first language, while Wang and Ladegaard (2008) reported 15 percent. The contrasts between these percentages indicate how first dialect acquisition in Guangzhou is changing. Remember that the three girls mentioned in the last case study, Q, F, and Y, claimed to be Pǔtōnghuà L1 speakers. The focus of this section is how Pǔtōnghuà fares in the competition for being a mother tongue and a tool for constructing a local identity. For this purpose, I chose another second-generation migrant boy L for my case study. L’s parents come from different parts of Shaanxi Province and both speak their local varieties of the Shaanxi dialect (which belong to the Mandarin dialect group), Pǔtōnghuà, and some Cantonese. His elder sister was born in the early 1980s and worked in another city in Guangdong Province. She speaks Cantonese fluently according to him. His grandmother was staying with the family and spoke only the Shaanxi dialect. A Shaanxi dialect-speaking housemaid was hired to take care of her and the whole family. L himself is multilingual. He likes learning foreign languages and tries to learn Japanese in his spare time, partly because both his parents worked with Japanese companies and partly because he was keen on the Japanese ACG subculture (Animation, Comic, and Games). He speaks Pǔtōnghuà without a noticeable regional accent, understands Cantonese and occasionally uses it, mostly “dirty words.” According to L, he was much better at Cantonese when he was in kindergarten where he had more Cantonese-speaking friends to talk to. However, he claimed to know little or no Shaanxi dialect and that his parents switched to Pǔtōnghuà when talking to him at home. The following conversation took place during the first interview with L. The discussion preceding this was about the languages he liked. The conversation was in Pǔtōnghuà. Words in bold are my emphasis. It is not a matter of whether I like it or not. Pǔtōnghuà is the most commonly used. If there were ten people in China, eight and a half of them were sure to use Pǔtōnghuà. AUTHOR: When you say they use Pǔtōnghuà, do you mean they can speak Pǔtōnghuà or they … often use Pǔtōnghuà. Then their mother tongue (mǔyǔ)…do you know what is mǔyǔ? L: I know. Our own regional dialect (我们的方言). AUTHOR: Right. L: They probably know quite a lot of their mǔyǔ. AUTHOR: Do you think Pǔtōnghuà is your mǔyǔ or…? L: Not really mǔyǔ. Oh, Pǔtōnghuà can be (my) mǔyǔ. AUTHOR: What is Shǎnxīhuà then? L:

198  Sihua Liang L: Step-mother mǔyǔ (后妈母语). AUTHOR: What is “step-mother tongue” L:

(后母语)? Can you explain that? I won’t understand (when I review our conversation) afterwards. Step-mother tongue, that is, (obviously), the second mother.

As can be seen, we had an explicit discussion on mother tongue, using the term mǔyǔ. It began as a slip of the tongue on my part, and I decided to turn it into an impromptu metapragmatic discussion. This was the only instance when I used the term with the primary school participants and it turned out to be a rare opportunity to examine the participant’s perception of the term mǔyǔ, rather than circumscribing the concept of mother tongue. L immediately made the connection between mǔyǔ and regional dialects, and added “our” and “own” to strengthen the sense of belonging and ownership. His comments on others’ good command of their mother tongues seemed to covertly contrast with his own minimal command of the Shaanxi dialect, his “first-order” mother tongue. When I proposed Pǔtōnghuà as a possible candidate, L hesitated. According to the “common sense” that “mǔyǔ should be a regional dialect,” Pǔtōnghuà is out of the question, but it seems to fit other criteria. Pǔtōnghuà was his favorite variety and the variety he had learnt from birth; he “knew quite a lot” of it just like “them”—other people who have “their own regional dialect.” Therefore, if he gave up the premise that mǔyǔ must be a regional dialect, Pǔtōnghuà could be his mother tongue. In just a fraction of a second L made up his mind. A second “problem” is that by allowing Pǔtōnghuà to be his mother tongue, L would have two different mother tongues, which contradicted conventional knowledge. Therefore, he invented the notion of “step-mother tongue” (后妈母语) to accommodate his need. By ranking his mother tongues, L made it possible to move a single variety up and down the hierarchy, and thus the Shaanxi dialect moved down and Pǔtōnghuà took its place to become L’s first-order mother tongue. As mentioned earlier, the definition of mother tongue sometimes resorts to its symbolic connection to the mother (Skelin Horvat and MuhvićDimanovski 2012). L’s intriguing invention of the term “step-mother tongue” highlighted this symbolic aspect at the core of the term. By using “stepmother,” he characterized the relationship between the Shaanxi dialect and himself as legitimate but not affectionate. He has not completely abandoned the conventional expectation about being able to speak the regional dialect of one’s professed hometown. He has merely downgraded the importance of that expectation to make space for Pǔtōnghuà. L is one of the participants with whom I have kept in touch. He was a senior high school student in a school near the GE community in 2015 when I revisited some of the participants. After the face-to-face interview, I asked if he, having lived in Guangzhou for 16 years, had ever considered himself as a “Guangzhou person.” He said he always introduced himself as a “Shaanxi person,” because “I reckon I should not forget my roots (忘本).” Comparing

Speakers of “Mother Tongues”  199 this to his “step-mother tongue” concept, it seems that while Pǔtōnghuà can be a mother tongue, there are roles that Pǔtōnghuà cannot take on. It does not give the sense of “having roots,” in other words, the sense of (regional) place and belonging. Pǔtōnghuà cannot be “hometown speech” because as a constructed common speech it does not have any regional hometown. L went to Hong Kong as an undergraduate in 2018, where Cantonese proficiency became a survival skill and the key to participation in local life. Such experience forced him to reevaluate his linguistic repertoire, particularly his Cantonese proficiency, and the role of languages in his construction of regional identity. In his interview (Liang, forthcoming), he admitted that he wanted to become a “real” Guangdong person more than ever before and considered Cantonese proficiency a must. He did not construe Cantonese as his mother tongue. However, by re-identifying his place of belonging and regarding Guangzhou as his symbolic hometown, he made it possible for Cantonese to become his symbolic “hometown speech,” and vice versa.

Conclusion: Hometown, Mother Tongues, and Identities This chapter began by considering the concept of “mother tongue” from social, historical, and ideological perspectives. It is argued that in the Chinese context “mother tongue” is an etic notion which has a relatively short history of common use while inextricably wired with ideological agendas. “Hometown speech” is one of the most frequently used “unmarked” Chinese terms that partially correspond to the concept of mother tongue in the English-speaking world. Among other overlapping connotations such as nativeness, heritage, and identity, the sense of place is central and integral to “hometown speech,” while the search for a legitimate “hometown” often requires competence in the “hometown speech.” In this chapter, we have encountered multiple ways in which second-generation migrant students define their hometowns, mother tongues, and ethnolinguistic identities, which seem to belong to different historical times and continue to vary. The complexity and diversity across time and cases remind us that social and sociolinguistic changes take place on different scales at different paces. In Blommaert’s term, different historicities coexist and synchronize within a superdiverse city (Blommaert 2013b). On the one hand, some age-old maxims about language and ancestral traditions are still influencing family language socialization and identity construction. For instance, in C’s case it was the notion of “family and clan” and his parents’ belief in the essential link between regional identity and regional dialect. On the other hand, competence in the regional dialect is considered essential for claiming the local identity traditionally connected with that region. For Q, the identity of a Guangzhou person was desirable but untenable because Cantonese proficiency was deemed key to this exclusive membership. Although Q and her friends were able to discursively problematize the rigid boundaries of the local identity, their endeavor testified to the

200  Sihua Liang dilemma and challenges that they face on a daily basis. In the case of L, his first language Pǔtōnghuà and his parents’ hometown speech Shaanxi dialect were strategically ranked as “mother tongue” and “step-mother tongue.” In his case, “mother tongue” is a more appropriate label than “hometown speech,” because Pǔtōnghuà as a translocal linguistic variety is not connected to any particular place. It also demonstrates that while Pǔtōnghuà is an important part of his identity, it does not provide connection to places. In L’s early discourse, it sounds as if such connection is not necessary, but his new-found passion for becoming a Guangdong person after living and learning Cantonese in Hong Kong seems to show that it is desirable if accessible. The participants in this study are both second-generation migrants and Pǔtōnghuà L1 speakers. Mobility and superdiversity challenge what used to be familiar such as places of origin and languages of roots, intimacy, and identity. They grew up facing the double dilemmas of having ambiguous hometowns and an unconventional, translocal L1. According to traditional language ideology, the “ideal” mother tongue for the participants is often the father’s native regional dialect connected with the father’s hometown. However, this ideal mother tongue has become unviable for those on the move. Complex trajectories of moving and socialization present competing alternatives. A prerequisite for such flexibility is the dissolution of the bonds between hometown, hometown speech, and mother tongue. Q, for example, argued for a local Guangzhou identity even though she can speak little Cantonese, meanwhile claiming alternative regional identities on other occasions. In the case of L who redefined the positions of different language varieties in his repertoire, Pǔtōnghuà emerged as a language of intimacy and importance, while Shaanxi remains the symbolic hometown. A few years later, he expressed his intention to identify Guangzhou as his second hometown and become a Guangdong person by learning and using Cantonese. It is notable how diverse sociocultural and ethnolinguistic dynamics in different parts of the sinophone world interact with speakers’ identity construction and recognition of mother tongues. While the participants managed to construct alternative languageidentity possibilities, we can also see their struggles and the price they have to pay for not complying with conventional norms. The current identity discourses have not accepted Pǔtōnghuà as a pathway to regional identities. Language activists often argue that Pǔtōnghuà is a language without roots. This is similar to the dilemma observed among transnational immigrants (Walker 2011) who seek to build a hybrid identity but are also faced with the risk of belonging “neither here nor there.” However, as the number of Pǔtōnghuà L1 speakers continues to increase, new ways of utilizing Pǔtōnghuà as an identity-constructing device are bound to be created. Although Pǔtōnghuà started as a constructed language with no native speakers, the longer and wider it has spread, the more likely it will become localized and vernacularized so that it can index local history and experience such as Taiwan Mandarin, Singaporean Mandarin, and the cosmopolitan Mandarin used in some foreign businesses (Zhang  2006).

Speakers of “Mother Tongues”  201 It is a matter of time and recognition. Moreover, with such dissolution, a person does not feel obliged to speak any particular first language. People would not be judged for not speaking certain regional dialects as their first language or identifying those dialects as their mother tongue. They would be able to make legitimate claims to regional identity based on other criteria, so that “hometown speech” can do without the “speech.” Young people may pick up elements or conversational ability in different regional dialects through their social media interactions and mobility experience, with or without constructing regional identities. Such a dissociation between regional dialects and regional identities may presage a gloomier future for the survival of Chinese regional dialects.

Notes

References Annamalai, Elayaperumal. 2006. “Mother Tongue Education: Nonstandard Language.” In Encyclopedia of Language and Linguistics, edited by Keith Brown et al., 342–45. Amsterdam: Elsevier. doi:10.1016/b0-08-044854-2/00650-7 Blommaert, Jan. 2013a. “Citizenship, Language, and Superdiversity: Towards Complexity.” Journal of Language, Identity and Education 12, no. 3: 193–96. doi:10.108 0/15348458.2013.797276 ———. 2013b. Ethnography, Superdiversity and Linguistic Landscapes: Chronicles of Complexity. Bristol: Multilingual Matters. doi:10.21832/9781783090419 Chan, Kam Wing, and Li Zhang. 1999. “The Hukou System and Rural-Urban Migration in China: Processes and Changes.” The China Quarterly 160: 818–55. doi:10.1017/s0305741000001351

202  Sihua Liang Chen, Ping. 1999. Modern Chinese: History and Sociolinguistics. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139164375 CPG (Central People’s Government of the PRC). 2005. “Zhōnghuá Rénmín Gònghéguó guójiā tōngyòng yǔyán wénzì fǎ” 中华人民共和国国家通用语言文字法 [Law of the people’s republic of China on the national commonly used language and script]. Accessed March 08, 2020 via http://www.gov.cn. Curtin, Melissa L. 2007. “Language Ideologies on Display: Local, Regional, & (Trans)national Identities in Taipei’s Linguistic Landscape.” PhD diss., The University of New Mexico. Dong, Jie. 2009. “‘Isn’t It Enough to Be a Chinese Speaker’: Language Ideology and Migrant Identity Construction in a Public Primary School in Beijing.” Language & Communication 29, no. 2: 115–26. doi:10.1016/j.langcom.2009.01.002 Ferguson, Charles A. 1959. “Diglossia.” Word 15, no. 2: 325–40. doi:10.1080/004379 56.1959.11659702 Fishman, Joshua A. 2002. “Diglossia and Societal Multilingualism: Dimensions of Similarity and Difference.” International Journal of the Sociology of Language 157: 93–100. doi:10.1515/ijsl.2002.045 Gao, Xuesong. 2012. “‘Cantonese Is Not a Dialect’: Chinese Netizens’ Defence of Cantonese as a Regional Lingua Franca.” Journal of Multilingual and Multicultural Development 33, no. 5: 449–64. doi:10.1080/01434632.2012.680461 GDCCP (Guangdong Chinese Communist Party), and GDGOV (Guangdong Government). 1992. “Zhōnggòng Guǎngdōng shěngwěi, Guǎngdōngshěng rénmín zhèngfǔ guānyú dàlì tuīguǎng Pǔtōnghuà de juédìng” 中共广东省委、广东省人 民政府关于大力推广普通话的决定 [Decision on vigorously promoting Pǔtōnghuà by Guangdong CCP committee and the Guangdong provincial government]. Yǔwén jiànshè 4: 2–3. Gold, Thomas B. 1993. “Go with Your Feelings: Hong Kong and Taiwan Popular Culture in Greater China.” The China Quarterly 136: 907–25. doi:10.1017/ s0305741000032380 Guangzhou Bureau of Statistics. 2001. Guǎngzhōu tǒngjì niánjiàn 2001 广州统计 年鉴2001 [Guangzhou statistical yearbook 2001]. Guangzhou: Zhōngguó tǒngjì chūbǎnshè. Gupta, Anthea F. 1997. “When Mother-Tongue Education Is Not Preferred.” Journal of Multilingual and Multicultural Development 18, no. 6: 496–506. doi:10.1080/01434639708666337 Hutton, Christopher. 1999. Linguistics and the Third Reich: Mother-Tongue Fascism, Race and the Science of Language. London: Routledge. doi:10.4324/9780203021019 Johnstone, Barbara. 2004. “Place, Globalization, and Linguistic Variation.” In Sociolinguistic Variation: Critical Reflections, edited by Carmen Fought, 65–84. Oxford: Oxford University Press. Kandiah, Thiru. 1998. “Epiphanies of the Deathless Native User’s Manifold Avatars: A Post-Colonial Perspective on the Native Speaker.” In The Native Speaker: Multilingual Perspectives, edited by Rajendra Singh, 79–110. New Delhi, Thousand Oaks, CA: Sage Publications. Kiesling, Scott. 2006. “Identity in Sociocultural Anthropology and Language.” In Encyclopedia of Language and Linguistics, edited by Keith Brown, 495–502. Oxford: Elsevier. doi:10.1016/b0-08-044854-2/03027-3 Klöter, Henning. 2004. “Language Policy in the KMT and DPP Eras.” China Perspectives 56: 56–63. doi:10.4000/chinaperspectives.442

Speakers of “Mother Tongues”  203 Leung, Yin-bing. 1993. The Use of Dialect in the Pearl River Delta Region: A Sociolinguistic Study. Hong Kong: Contemporary China Research Centre, City Polytechnic of Hong Kong. Li, Wei. 2011. “Moment Analysis and Translanguaging Space: Discursive Construction of Identities by Multilingual Chinese Youth in Britain.” Journal of Pragmatics 43, no. 5: 1222–35. doi:10.1016/j.pragma.2010.07.035 Lǐ Yǔmíng 李宇明. 2003. “Lùn mǔyǔ” 论母语 [On mother tongue]. Shìjiè Hànyǔ jiàoxué no. 1: 48–58. Liang, Sihua. 2013. “Construction of Language Attitudes in Multilingual China: Linguistic Ethnographies of Two Primary Schools in Guangzhou.” PhD diss., University of Cambridge. ———. 2014. Language Attitudes and Identities in Multilingual China: A Linguistic Ethnography. Cham: Springer. doi:10.1007/978-3-319-12619-7 ———. (forthcoming). “From Passive to Active: A Longitudinal Study of Young Bilingual’s Language Attitude and Identity Changes.” In Linguistic Sustainability, edited by C. Chang and C. Yu. Guangzhou: Sun Yat-sen University Press. Liú Lín 刘林. 2017. “Guǎngdōnghuà vs Pǔtōnghuà: Dāng Xiānggǎng de yǔyán yùshàng zhèngzhì” 广东话vs普通话：当香港的语言遇上政治 [Cantonese vs. Pǔtōnghuà: When languages in Hong Kong meet politics]. Accessed March 03, 2020 via https://www.bbc.com. Lytra, Vallz, and Peter W. Martin, eds. 2010. Sites of Multilingualism: Complementary Schools in Britain Today. Stoke-on-Trent: Trentham Books. Mackenzie, Peter W. 2002. “Strangers in the City: The Hukou and Urban Citizenship in China.” Journal of International Affairs 56, no. 1: 305–19. Mair, Victor H. 1991. “What Is a Chinese ‘Dialect/Topolect’? Reflections on Some Key Sino-English Linguistic Terms.” Sino-Platonic Papers 29: 1–31. Maybin, Janet, and Karin Tusting. 2011. “Linguistic Ethnography.” In The Routledge Handbook of Applied Linguistics, edited by James Simpson, 515–28. London: Routledge. Rampton, M. Ben H. 1990. “Displacing the ‘Native Speaker’: Expertise, Affiliation, and Inheritance.” ELT Journal 44, no. 2: 97–101. doi:10.1093/eltj/44.2.97 Ramsey, S. Robert 1987. The Languages of China. Princeton, NJ: Princeton University Press. Shohamy, Elana G. 2006. Language Policy: Hidden Agendas and New Approaches. London: Routledge. doi:10.4324/9780203387962 Skelin Horvat, Anita, and Vesna Muhvić-Dimanovski. 2012. “My Mother Tongue… Croatian, Istrian, Local,… Depends Where I Am: The Perception of Mother Tongue in Multilingual Settings.” Jezikoslovlje 13, no. 2: 493–511. Snow, Don. 2004. Cantonese as Written Language: The Growth of a Written Chinese Vernacular. Hong Kong: Hong Kong University Press. Táng Yè 唐叶. 2006. “Guǎngzhōushì zhōngxuéshēng yǔyán tàidù yánjiū” 广州市中 学生语言态度研究 [Middle school students’ language attitudes in Guangzhou]. MA thesis, Jinan University (Guangzhou). THGOV (Tiānhé District Government 广州市天河区人民政府). 2016. “Tiānhéqū jīběn qíngkuàng 天河区基本情况 [The general situation of Tiānhé district]. Accessed March 10, 2017 via http://www.thnet.gov.cn. Tián Xiǎolín 田小琳 2012. Xiānggǎng zhōngwén jiàoyù miànlín de zhòngyào juézé 香 港中文教育面临的重要抉择 [A crucial choice challenging Chinese education in Hong Kong]. Yúnnán Shīfàn Dàxué xuébào 44, no. 2: 14–21.

204  Sihua Liang Walker, Ute. 2011. “Linguistic Diversity as a Bridge to Adjustment: Making the Case for Bi/Multilingualism as a Settlement Outcome in New Zealand.” In Language Policy for the Multilingual Classroom: Pedagogy of the Possible, edited by Christine Hélot and Muiris Ó. Laoire, 149–73. Bristol: Multilingual Matters. doi:10.21832/9781847693686-010 Wang, Limei, and Hans Ladegaard. 2008. “Language Attitudes and Gender in China: Perceptions and Reported Use of Putonghua and Cantonese in the Southern Province of Guangdong.” Language Awareness 17, no. 1: 57–77. doi:10.2167/la425.0 Wiley, Terrence G. 2009. “Chinese ‘Dialect’ Speakers as Heritage Language Learners: A Case Study.” In Heritage Language Education: A New Field Emerging, edited by Donna Brinton, Olga Kagan, and Susan Bauckus, 91–106. London: Routledge. doi:10.4324/9781315092997-7 Xiong, Tao. 2018. “Media Representation of Attitudes to Multilingual Education: Local Language and Culture in Mainland China.” In Multilingual Education Yearbook 2018, edited by Indika Liyanage, 37–54. Cham: Springer. doi:10.1007/978-3-319-77655-2_3 Yan, Margaret M. 2006. Introduction to Chinese Dialectology. München: Lincom. Zhang, Qing. 2006. “Cosmopolitan Mandarin: Linguistic Practice of Chinese Waiqi Professionals.” Journal of Asian Pacific Communication 16, no. 2: 215–35. doi:10.1075/japc.16.2.05zha

10 Multilingualism and Language Policy in Singapore Peter Siemund and Lijun Li

Abstract Over the past decades, the aim of language policies in Singapore has been to reduce language barriers and to promote better communication among the Chinese, Malay, and Indian communities as well as to prevent heritage erosion of these ethnic groups. This chapter examines the impact of these policies on the multilingual landscape of Singapore from the point of view of the individual, complemented by official statistical data. It argues that language policies had a deep impact on the linguistic texture of Singapore, with the city-state being in transition from a diverse multilingual society toward one characterized by individual bilingualism. Differences in home language use and observable language repertoires can be explained by factors such as ethnicity and social background. Colloquial Singapore English (CSE) (Singlish), a variety of English strongly characterized by influence from the local substrate languages, is gaining growing acceptance as a national solidarity code.

The Multilingual Texture of Singapore The indigenous people of Singapore are the Malays, even though they only account for a small proportion of today’s population. They have inhabited the area that is now Singapore since as early as the thirteenth century, which predates the arrival of Chinese and Indian migrants during British colonial rule (see Lim 2010: 23). Many of the indigenous Malays lived on the island under the Johor Sultanate prior to the arrival of Sir Stamford Raffles. These early immigrants were mostly from the Malay Archipelago, from Malacca, Sumatra, and the Riau Islands, and later also from Java and Bawean Island, as well as Sulawesi and other islands (Lim 2010: 23). Most of the Malays in Singapore today have their roots in Indonesia and Malaysia. They were the largest ethnic group (60 percent in 1824) in Singapore until the 1830s, but their numerical dominance swiftly gave way to that of the Chinese. According to the 2019 census, Malays constitute 15 percent of the resident population (Wong 2019), making them the second largest ethnic group in the city-state. The Malay language (also called Bahasa Melayu) has been recognized as Singapore’s national language since independence and is mostly used within

206  Peter Siemund and Lijun Li the Malay speech community. Virtually the entire Malay population speaks Malay and despite their being an ethnic minority in Singapore, the status of the language is not threatened, as Malay is not only one of the official languages in the city-state, but is also the national language of Malaysia and Indonesia (Deterding 2007: 4). Besides Malay, a Malay pidgin called Bazaar Malay has served as the lingua franca in the region for centuries. It was also widely spoken in the East Indian archipelago and was the basis of the colonial language used in Indonesia by the Dutch (see Bao and Aye 2010). The version of Bazaar Malay used in Chinese merchant communities is called Baba Malay (Ansaldo et al. 2007). In the early 1970s, Bazaar Malay was still one of the most understood languages, second only to Hokkien and widely used for interethnic communication, with all Indians and 45 percent of Chinese claiming to understand it (Lim 2010: 27). However, its significance as a lingua franca declined starting with the independent era of Singapore, though it is still spoken by the older generation and in lower social strata (Lim 2010: 28). Singapore has always been an attractive destination for people from mainland China. Chinese immigration to the small island off the southern tip of the Malay Peninsula has a long history. First traders and settlers can be supposed to have traveled to what is today Malaysia and Singapore as early as around 1000 CE, often staying and marrying local Malay women and later establishing the economically successful group of Peranakans. Chinese immigration surged when the British established Singapore as a trading post in 1819 and as a separate crown colony in 1867. Among the first settlers with Chinese roots were Peranakans from Penang and Malacca (on Peranakans—or Straits Chinese—see Goh and Fong, this volume). Later and with the increasing demand for labor, a continuous stream of mainland Chinese began to make their way to Singapore—first settling in what now is China Town and subsequently spreading across the entire island. In view of the geographical location, it comes as no surprise that the overwhelming majority of Chinese settlers in Singapore hailed from southern China, especially the provinces of Fujian, Guangdong, and Hainan. These provinces were and still are home to speakers of distinct Sinitic varieties, including Cantonese (Yuè), Hainanese, Hakka, Hokkien, and Teochew. Emigration from these provinces is by no means restricted to Malaysia and Singapore; large numbers of speakers can also be found in Taiwan, Indonesia, and the Philippines. The British also shipped foreign laborers from India to Singapore— mainly Tamils—who now represent more than 50 percent of the Indian population in Singapore. However, the overall size of the ethnic Indian group is rather small, amounting to 7.47 percent of the total population in 2019 (Wong 2019). The Singaporean government has been making substantial efforts to avoid ethnic segregation, inter alia by enforcing the national population ratio in practically every city district using housing allocation measures. In spite of such efforts, there remain important differences regarding income, educational achievement, and other parameters of social

Multilingualism and Language Policy  207 stratification in regard to the three major ethnic groups, some of which will also be addressed in this chapter. English, as a former colonial language and today’s global lingua franca, enjoys a special status in the city-state. As one of the official languages besides Mandarin, Malay, and Tamil, it is the main language of business, administration, and the public sphere. Moreover, it is used as the main medium of instruction in the education system, the so-called “mother tongues”—i.e. Mandarin, Malay, Tamil—being learned by and large as second languages. The government of Singapore views English as the gate opener to the world, while, at the same time, encouraging the maintenance and use of the mother tongues so as to embrace Asian cultural values and preserve local identities. Although one of the official languages, English is encountered in a broad spectrum of forms, ranging from standard Singapore English, which resembles standard British English in its major aspects, to CSE or Singlish. The latter form or variety of English is strongly characterized by influences from the local substrate languages, especially dialects of Chinese and Malay, and has been widely researched by scholars in World Englishes (see among many Bao 2015; Leimgruber 2013a, 2013b; Lim et al. 2010; Siemund and Li 2017; Ziegeler 2015). For instance, its syntactic structure parallels that of Chinese. The example in 1 finds a ready one-to-one translation in Chinese, as shown in 2. 1 Colloquial Singapore English Ask also not get. ‘Even if you ask, you won’t get it.’ (Siemund and Li 2017: 16) 2 Mandarin 问而不得 wèn ér bù dé ask and not get ‘Even if you ask, you won’t get it.’ (adapted from Siemund and Li 2017: 16) Some of the other important features of CSE include missing or nonstandard subject-verb-agreement (example 3) and the extensive use of discourse particles like lah, ah, hah, meh, and lor (example 4), borrowed from Chinese, especially Cantonese, and Malay (see Li, Lazarov, and Siemund 2020; for the borrowing of utterance final particles see also Lin, this volume). These apart, there is of course also considerable lexical influence from all the languages spoken in the area, i.e. Malay (kantang ‘potato’), Hokkien (ang moh ‘red hair’ meaning ‘Caucasians’), Teochew (nia gong ‘your grandfather’), Cantonese (ta pau ‘take away’). 3 a b

Now St Margaret’s school, ø also a mission school now. [Oral History Interviews-000213-EQ] My father do the church works. [Oral History Interviews-000213-EQ]

4 Go to Chinatown lah. [ICE-SG-S1A-007]

208  Peter Siemund and Lijun Li Phonological features also show influences from the substrate languages. These include relatively shorter vowels, though this depends on conditioning by the phonological environment: beat and bit are both /bit/, but bee, being /bi/ in Standard English, is realized as [bi:] (Leimgruber 2013b: 64), reduced consonant clusters and unreleased or glottalized final consonants (ibid., 66), and labiodentalized dental fricatives as in health realized as [hɛlf] (Bao 1998: 154). For more information on the phonological features of CSE, readers are referred to Bao (1998), Deterding (2007), and Leimgruber (2013b). Many speakers of Singapore English are fluent in the whole range from standard to colloquial and can use these varieties to signal different social orientations, for example, more Western or more local. The main thrust of our study concerns issues of language policy and language change in the city-state, to be explicated in the following two sections. To that end, we will approach these issues from the point of view of the individual, thus complementing the officially available statistical data. Above and beyond such matters of linguistic landscaping, our study also offers a discussion of language proficiency levels in English and the respective mother tongues, both self-assessed and objectively measured. Last but not least, we will also be concerned with the status of CSE, especially its increasing acceptance as a national solidarity code.

Language Policy Issues Multilingual Policy and Bilingual Education Language planning in Singapore is highly centralized, resulting in a topdown approach in implementation and decision-making (see Chua 1995; Goh and Fong, this volume; Ho and Alsagoff 1998: 202). The main foci of the Singaporean language policy investigated in previous research (see Bokhorst-Heng 1999; Lim et al. 2010; Tan 2007; Wee 2008) lay on its official embedding and educational bilingualism, both of which are very closely related. Singapore’s official language policy identifies a connection between ethnicity and language as well as economic development and language. The Republic of Singapore Independence Act of 1965 gave Malay, Mandarin, Tamil, and English official language status. Of the four official languages, three (Mandarin, Malay, and Tamil) are seen as “mother tongues” of the Chinese, Malay, and Indian communities, respectively. They were adopted by the government to give Singaporeans an anchor in their ethnic identities and traditional values so as to avoid excessive Westernization and to prevent deculturalization (Ng 2011: 4). As a matter of fact, the “mother tongue” policy is a reconceptualization and an oversimplification of the differences among heterogeneous communities, making each definable in terms of one single language. This means that regardless of what language may actually have been spoken by a child in early childhood, the ethnic group of the child’s father determines which language is officially assigned as their

Multilingualism and Language Policy  209 “mother tongue.” In other words, even if a child whose parents are ethnically Cantonese spoke Cantonese in childhood, Mandarin would still be regarded as the child’s official “mother tongue” in Singapore. While “mother tongue” generally refers to the first language (L1) (for a discussion of the concept of mother tongue, see Liang, this volume), it is used to denote the “ethnic language” or the second language (L2) in the Singaporean context. Prior to 2011, “mother tongue” was not defined as the home language or the first language acquired by the student but by the father’s “race”1 (Leimgruber 2013b:11). Here, the traditional patrilineal idea of ethnicity and descent was obviously at play. The mother tongue policy is closely linked to the ethnic policies of the government, where one’s community membership is determined by one’s father’s ethnicity (Leimgruber 2013b: 11). The patrilineal idea is still present and plays an important role in the today’s Singapore, as exemplified by the recent campaign to revitalize the use of dialects among youths, which is called the “Father Tongue” campaign (see Goh et al. 2016). English was chosen as an official language in Singapore for pragmatic reasons. Being essential to Singapore’s independence from the early years and crucial to science and technology, English was seen as the language of technology and economic development. The policy of multilingualism was designed to serve the goal of establishing equal status for all four official languages, but in reality, English is still the most important language in government administration, business, law, banking, and accountancy. In addition, it has first language status in education, while the others are relegated to second language status and are only taught in language and literature classes. Hence, English is the main medium of instruction. Across ethnic groups, it serves as a neutral language for communication and can bridge cultural differences. For the individual, English is also crucial for one’s career path (Bokhorst-Heng 1999). The importance and dominance of English in Singapore may be due partly to language planning and partly to the development of English into a lingua franca even after the demise of the British Empire, as a result of geopolitical realities far beyond Singapore. It is generally assumed that neglect of English would result in marginalization of the country and loss of access to the extensive resources available in English (Chew 1999). The shift toward English has even extended beyond the public to the private domain, which will be further discussed below. Bilingual education is an important element of Singaporean language policy. It was implemented and made compulsory in schools in 1966 and defined by the government as “proficiency in English and one other official language” (Ng 2011; Pakir 1994: 159). School bilingualism was implemented by a series of detailed guidelines involving exposure time, subject-language matching, examination and attainment requirements (Gopinathan 1998: 21). Students are taught English as the medium of instruction in schools and they have the choice to learn one of the other official languages as an L2, very often their “mother tongue.” For example, Chinese students in Singapore

210  Peter Siemund and Lijun Li are required to study English as a “First Language” and Mandarin as a “Second Language.” While English is the medium of instruction in most of the content courses (i.e. arts; physical education; sciences including computing, mathematics, and electronics; and humanities including geography, economics, history, and social studies), only Chinese language and Chinese literature are taught in Mandarin (MOE 2018).2 The purpose of this policy was explained by former Minister of Education, Tony Tan Keng Yam: “Children must learn English so that they will have a window to knowledge, technology and expertise of the modern world. They must know their mother tongue to enable them to understand what makes us what we are today” (cited in Pakir 2008: 193). However, as mentioned earlier, linguistic ownership in Singapore is defined in terms of “mother tongue” rather than “native language.” Although Mandarin is taught as the “mother tongue” of the Chinese, it is not the language first learned and usually spoken by all Chinese in Singapore. Thus, the policy was often criticized as “ignoring an individual’s linguistic experience in favor of a community’s historical association or heritage” (Wee 2003: 289). The association of a single language with an ethnic group in Singapore did not appear suddenly but has its roots in the precolonial period. First of all, efforts were made to assign each immigrant an area of settlement under the administration plan of Sir Thomas Stamford Raffles (1781–1826), an official of the British East India Company, in 1826 (LePoer 1991: 15). However, the British were not particularly concerned about what ethnic languages the Singaporeans used as long as English was nurtured as a common tongue between the highly diverse residents of the island (also discussed in Goh and Fong, this volume). As stated by the first Prime Minister Lee Kuan Yew 李光耀 (1923–2015): Singapore never had one common language. … The British left people to decide how to educate their children. The government provided a limited number of English language schools to train people to be clerks, storekeepers, draftsmen, and such subordinate workers, and Malaylanguage primary schools for Malays. Indians ran their own Tamil and other Indian-language schools or classes. The Chinese set up schools financed by successful members of their community, to teach in Chinese. Because the different races were taught their own languages, their emotional attachment to their mother tongue was deep. They were like the 5 million people in Quebec tenaciously holding on to French in a continent of 300 million English speakers. (Lee 2000: 146) The choice of Mandarin (or Guóyǔ ‘national language’) over other “dialects” has its historical background related to China (for the status of dialects, see Klöter, this volume). Although Mandarin was not the language of the immigrants, it was considered a unifying factor by the Chinese leadership

Multilingualism and Language Policy  211 factions of both Singapore and China (LePoer 1991: 33). Between the World Wars (1919–41), the ties between the Straits-born Chinese and their homeland in China were strong, not only politically and economically, but also in terms of education. Politically, there was an active Singapore branch of the Nationalist Party of China (KMT), boycott of Japanese goods, and mass support for Chinese nationalism (see LePoer 1991: 33). Economically, there was heavy investment in Chinese industry and education by Singapore’s China-born businessmen, encouraged by Sun Yat-sen in the early 1920s (see LePoer 1991: 33). In return, the KMT sent teachers and textbooks to Singapore and promoted the use of Mandarin in Singapore’s Chinese schools. Singapore’s first Chinese secondary school (established in 1919) and a growing number of Chinese primary schools taught in Mandarin. The KMT made efforts to increase the number of promising students brought from Singapore to China for university education and began to extend its control over Chinese schools in Nanyang3 by supervising the curriculum and requiring the use of Mandarin. Although the colonial power sought to discourage the use of Mandarin as required by the KMT, aware of the spread of left-wing politics in the Chinese schools in the late 1920s, Mandarin had nevertheless become the medium of instruction in all of Singapore’s Chinese schools (LePoer 1991: 33). After independence, Lee Kuan Yew reinforced the choice of Mandarin over other Chinese dialects, extending its importance not just to the educational domain but also to its use as a home language (for details on Lee’s language ideology and governmental language planning under Lee, see Goh and Fong, this volume). The Speak Mandarin Campaign and the Speak Good English Movement One of the assumptions of language planning is that a government should encourage the use of high-status languages (Dixon 2005: 632). This assumption seems especially true regarding language planning in Singapore. Over the past few decades, the Singaporean government has actively encouraged the use of Standard (British) English over a localized variety of English (CSE) as well as Mandarin over Chinese dialects, i.e. Cantonese, Hokkien, Teochow. The Speak Mandarin Campaign (SMC) and the Speak Good English Movement (SGEM) are two very good examples. The SMC is an initiative launched in 1979 by the government of Singapore to encourage the Singaporean Chinese population to speak Mandarin, one of the four official languages of Singapore, not only as an educational language but also as a home language. The aim of the campaign was to reduce the degree of sinophone polyphonicity, which was regarded by then Prime Minster Lee Kuan Yew as problematic, making it difficult to unify the Chinese community, as well as hampering the successful implementation of the bilingual education policy (Lee 2000: 154–55). The bilingual education policy introduced in 1966 required students with a Chinese ethnic background

212  Peter Siemund and Lijun Li in Singapore to learn both Mandarin and English at school, but it was reported in the 1970s that there was a preference to use dialects in the home domain (Sim 2016). Using dialects at home was regarded by the government as a burden for students who had to master English and Mandarin at school. As put by Lee Kuan Yew: “It would make it easier for students to master English and Mandarin in school if they spoke Mandarin at home and were not burdened by dialects” (Lee 2000: 154). The reason for emphasizing Mandarin, besides the historical background mentioned earlier, was that Mandarin was seen as becoming more economically advantageous with the opening up of China (Lee 2000: 155). The SMC became an annual event promoting the use of Mandarin. Measures have included discouraging the Chinese civil servants from using dialects during office hours, producing cassette tapes of Mandarin lessons (in 1979) and distributing them via telephone on a 24-hour basis to the public (between 1983 and 1987), publishing a series of comic books titled “Mr Kiasu learns Mandarin” as a light-hearted way for the public to learn Mandarin, organizing Mandarin Film Festivals, and launching smartphone software designed as a bilingual language learning tool with a database containing some 50,000 commonly used business and Singapore-related terms in both English and Mandarin (Sim 2016). The SMC was initially a failure, as there were protests by dialect speakers, especially the elderly, complaining that Mandarin did not correspond to the language profiles of local Chinese. Parents also refused to register their children’s name in Mandarin using the Hànyǔ Pīnyīn transcription system (see Gopinathan 1998). However, a study conducted by Lock (1989) showed that certain nonstandard features (for example, lack of initial retroflexion in syllables beginning sh, ch, and zh) seemed to move closer to the prescribed standard in the 1980s. Recent research shows an increase of Mandarin use in different domains and positive attitudes toward Mandarin (Ng 2014). The majority of respondents in Ng’s study (ibid., 58) stated that they liked speaking Mandarin and wanted their children to speak Mandarin. Furthermore, 87 percent of the respondents agreed that Mandarin was the mother tongue of Chinese Singaporeans. It is also reported that the growing positive attitudes toward Mandarin are due to pragmatic reasons and that the great majority of Singaporeans view Mandarin as an important language in trade and business with China. The first SGEM was launched two decades later in 2000, and received mainly positive reactions right from the start. Its pronounced aim is to increase the use of Standard English and discourage the use of CSE by primary school children and in the media so as to improve Singapore’s stance in the industrial and economic activities of the region (Rubdy 2007; Tsui and Tollefson 2006). The movement was inaugurated with an official speech by the then Prime Minister, Goh Chok Tong 吴作栋 (b. 1941), who considered CSE to be “a corrupted form of English” and expressed concern that if Singaporeans continued to speak it, they would not be understood by others and

Multilingualism and Language Policy  213 therefore lose a key competitive advantage (Leimgruber 2014). Many believe that CSE is a “less prestigious dialect associated with low social status” and an obstacle to the nation’s continued economic growth, which undermines Singapore’s place in the global marketplace (Rubdy 2007: 308). Promotions of SGEM events often put a negative connotation on CSE speakers as being uneducated or uncultured (Hoon 2003). Over the years, the government’s stance against CSE has remained strong and unambiguous. While from a diglossic view, CSE is regarded as a perfectly viable vernacular, existing alongside the standard norm, for the policy makers the co-existence of CSE and Standard English is not “an option” (Rappa and Wee 2006: 95). The first Prime Minister described CSE as a “handicap” which must not be advocated (Tan 2007: 88). This standpoint is also evident in a strongly worded opening statement in an open letter: “While Singlish might be a fascinating academic topic for linguists to write papers about, Singapore has no interest in becoming a curious zoo specimen to be dissected and described by scholars” (as cited in Alsagoff 2015: 127). It is difficult to make an assessment of the success of the SGEM. On the one hand, since its implementation, the literacy rate in English increased from 70.9 to 79.9 percent and again to 83.1 percent, and the proportion of bilingual speakers rose by 17.2 percent between 2000 and 2015 (Wong 2011; 2016). Moreover, a comparison of attitudes toward English and Mandarin shows that the majority of Singaporeans consider English more important than Mandarin (Ng 2014: 60). The policies also induced a sharp increase in the use of English as home language. On the other hand, the initial goal of the SGEM—the eradication of “Singlish”—has not been achieved. Leimgruber (2013a: 250) predicts that CSE is unlikely to disappear in the near future. While Standard Singapore English is the desired prescribed norm for the city-state, CSE is believed to be the language closer to their culture for the great majority of Singaporeans (Cavallaro and Ng 2009). The studies by Siemund et al. (2014) and Tan (2014) suggest a tendency among the younger generation of Singapore to embrace CSE as a national solidarity code, which will be further discussed below.

Language Shift The language policies pursued by the Singaporean government have had noticeable effects, especially regarding the distribution of English, Mandarin, and Chinese dialects. Since the government efforts essentially boil down to a bilingual language policy promoting the use of English in combination with one of the mother tongues (i.e. Mandarin, Malay, and Tamil), we can witness a strong increase in home language use of English and Mandarin to the detriment of Chinese dialects. Figure 10.1 displays trends in home language use based on official statistics released by the Singaporean government. The sharp increase in the use of English and Mandarin as home languages is clearly visible.

214  Peter Siemund and Lijun Li 80% 70% 60% 50% 40% 30% 20% 10% 0% 1957 English

1980 Mandarin

1990

2000

2010

Chinese Vernaculars

Malay

2015 Tamil

Figure 10.1 Changes in home language use over time in percent. Source: Cavallaro and Ng (2021 forthcoming), Wong (2011, 2016, 2019).

One needs to interpret Figure 10.1 keeping in mind that Chinese immigrants to Singapore did not speak Mandarin when they arrived, but one of the aforementioned Chinese dialects. Since these practically represent languages in their own right and not just “dialects,” ethnic Chinese Singaporeans were confronted with a double language shift toward Mandarin on the one hand and English on the other hand. What is noteworthy is that the use of Malay and Tamil has practically remained unchanged over the period depicted here, even though the home language use of English has increased in these two ethnic groups. There is, however, not the same tension between standard language and dialects as there is in the Chinese community (i.e. Mandarin versus Cantonese, Hainanese, Hakka, Hokkien, and Teochew). Notice that Figure 10.1 does not differentiate home language use of English according to ethnic background. Figure 10.2 plots the differences in home language use according to age, supporting the developments depicted in Figure 10.1. English home language use is strongest in the youngest age group and there is an increase in the use of English negatively correlating with age, as expected. Home language use of Chinese vernaculars declines across the oldest to the youngest age groups. The use of Mandarin as a home language shows the same pattern as English, though there is a slight drop in the younger age groups. Home language use of Malay and Tamil is practically stable across the age groups distinguished here. It would appear to be a plausible conjecture that the observable changes in home language use are related to the bilingual policies pursued by the

Multilingualism and Language Policy  215 350 300 250 200 150 100 50 0

English

Mandarin

Chinese Vernaculars

Malay

Tamil

Others

Figure 10.2 Differences in home language use according to age (in thousands). Source: Wong (2016).

Singaporean government. If we look at the endpoints of the trajectories in Figure 10.1 or the younger age groups in Figure 10.2, we can identify the mother tongues and English. To be sure, these figures portray averaged home language use and do not inform us about the linguistic repertoires of the individual speakers, which will be the topic of the next section.

Individual Multilingualism Since the official statistics provided by the Singaporean government only indicate home language use and do not tap into the individual language repertoires of the speakers, we need to consult additional sources to gain more detailed knowledge about the linguistic texture of the city-state. Unfortunately, although the Singaporean linguistic landscape is quite well researched, especially the grammatical properties of CSE (see Bao 1995, 2005; Leimgruber 2013a, 2013b; Lim 2010), very little is known about individual multilingualism or about the domains of language use. One of the few studies dedicated to exploring individual multilingualism is Siemund et al. (2014), in which two groups of university and polytechnic students—each comprising 150 individuals—were asked to fill in detailed questionnaires about individual language repertoires and use (see also Leimgruber et al. 2018).4 Figure 10.3 shows some of the results of this study, listing the most prominent language profiles found among the two student groups.

216  Peter Siemund and Lijun Li 80 70 Number of students

60 50 40 30 20 10 0

70

31

29

35 10

E MC

E H MC

16

C E MC

2

15

EM

8 5

2 7

3 5

3

2

E H MC TC C E H MC E H M MC C E H MC TC

2 3 E MC TC

Language combinations University students

Polytechnic students

Figure 10.3 Language combinations amongst Singaporean university and polytechnic students that occurred at least five times in the questionnaire data. Source: Siemund et al. (2014: 353). E = English, MC = Mandarin Chinese, C = Cantonese, M = Malay, TC = Teochew, H = Hokkien.

We can make two important observations here. On the one hand, even though there is great variation in the observable language repertoires, there is nevertheless a small number of dominant and recurring profiles to which most students of the sample can be assigned. The figures shown in the bars in Figure 10.3 refer to the absolute numbers of students in the sample, e.g. 70 university students and 31 polytechnic students share the language profile “English + Mandarin Chinese.” Siemund et al. (2014: 351) identify the following four language profiles as covering around 70 percent of the respondents (208 out of 300 students). 1 2 3 4

English + Mandarin Chinese English + Hokkien + Mandarin Chinese Cantonese + English + Mandarin Chinese English + Malay

On the other hand, we can witness important differences between university and polytechnic students. For example, university students are overrepresented in the bilingual language profile “English + Mandarin Chinese,” while the bilingual profile “English + Malay” is practically restricted to polytechnic students. The trilingual profiles “English + Hokkien + Mandarin” and “Cantonese + English + Mandarin” are more strongly represented in the group of polytechnic students. On the whole, student bilingualism is

Multilingualism and Language Policy  217 strongest in the university group, whereas polytechnic students more often have trilingual language profiles. These findings are summarized in Figure 10.4. Trilingualism is essentially composed of English and Mandarin in combination with one of the Chinese dialects spoken in Singapore.5 One of the few additional studies that try to go beyond the information contained in the official government census is the Sociolinguistic Survey of Singapore (SSS) 2006 (Vaish et al. 2009). Here, 716 students from the Primary 5 cohort of Singapore schools were sampled, stratified according to ethnic background and socio-economic status (high, medium, low). SSS 2006 explored the use of English and the mother tongue in school, amongst family and friends, for religious activities, in public spaces, and in interaction with the media, considering fluency, proficiency, and attitudes. The study generally reports a higher home language use of English in comparison to the census report. This is consistent across all ethnic groups. Conversely, the levels of home language use of the mother tongue are lower than in the census, except for Tamil. These results are not unexpected, since the census data average across all groups, whereas SSS 2006 focuses on a particularly young age group (see Figure 10.2). Even though it is true to say that there are hardly any interethnic studies investigating individual multilingualism, we do find empirical research dedicated to specific ethnic groups. For example, Cavallaro and Serwe (2010) researched the Malay community and found substantially less use of Malay and, consequently, more use of English among young speakers, typically highly educated Malay women. Additional studies include Li et al. (1997) for Teochew and Schiffman (2002)

80

Number of students

70 60 50 40

73

68

30

49

20 10 0

44 22 23 9

1

2 1

2

3

5

4 5 6 Number of languages University students Polytechnic students

3

1 7

Figure 10.4 Number of languages spoken by university and polytechnic students. Source: Siemund et al. (2014: 353).

218  Peter Siemund and Lijun Li and Vaish et al. (2009) for Tamil. All studies have found an increase in the use of English, though typically in restricted domains.

Proficiency in English and the Mother Tongue Given that English is one of the official languages in Singapore and the compulsory language of instruction in the education system, proficiency in it can be assumed to be generally high, even though it is often not entirely clear which lect in the continuum from Colloquial to Standard Singapore English serves as the basis for the relevant assessments. The Singaporean Ministry of Education (MOE) regularly publishes the results of such assessment tests, the data having been gathered using various production and comprehension tasks, both written and oral, and there can certainly be no doubt that Standard English was used for the written tasks. Figure 10.5 shows the proficiency levels in English according to the figures released in 2018 for the three major ethnic groups, measured in terms of the percentage of students passing the relevant O-level exams—here used as a proxy for proficiency. One can see that the Indians consistently outperform the other two ethnic groups, with Malay students manifesting the lowest pass rate in English. Chinese students range in between, and on the whole there do not seem to be important differences over time. Figure 10.6 provides the corresponding figures for the mother tongues where it is now the group of Malays who consistently outperform the other two ethnic groups. There is no substantial difference between Chinese and Indian students in regard to mother tongue proficiency. Again, the official

95 90 85 80 75 70 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Malay

Chinese

Indian

Figure 10.5 Percentage of students who passed the O-level English language exam. Source: MOE (2018).

Multilingualism and Language Policy  219 100 99 98 97 96 95 94 93 92 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Malay

Chinese

Indian

Figure 10.6 Percentage of students who passed the O-level mother tongue language exam. Source: MOE (2018).

government data only release proficiency levels in English and the mother tongues, and not in the various vernacular forms spoken in the city-state or the various foreign languages known by its population. The aforementioned study by Siemund et al. (2014) offers more detailed information, though it only samples two highly specific social groups, namely university and polytechnic students. Their results in Figure 10.7 show a monotonic decrease in proficiency for each additional language (one to four). This is not particularly surprising, as the students were asked to rank their languages according to proficiency, but what is noticeable is that—at least on average—no student claimed to be fully proficient even in their first language. One may perhaps put this down to Asian modesty—if there is such a thing—but perhaps a more plausible explanation may lie in the widespread multilingualism found in the student sample. Since the knowledge of several languages usually makes for a higher degree of language awareness, speakers become more conscious of their first language and the limitations they have therein. Positive effects of individual multilingualism may also be deduced from Figure 10.8, where proficiency levels in the first four languages are plotted depending on the number of languages known by the students. We can see here that individuals reporting knowledge of four languages consistently outperform trilingual and bilingual students. In the same way, trilinguals report higher proficiency levels than bilinguals. Proficiency levels seem to increase with the number of languages known. These findings can be viewed as support for models of multilingualism that postulate positive effects with each additional language learnt such as Flynn et al.’s (2004) Cumulative

220  Peter Siemund and Lijun Li

Proficiency ratings

very good 1.65 x

good

2.27 x 3.01 x

fair

3.19 x

not good poor

LProf1

LProf2

LProf3

LProf4

Figure 10.7 Overall mean proficiency for language 1, 2, 3, and 4. Source: Siemund et al. (2014: 354).

Proficiency ratings

very good good

4 or more languages 1.55

1.67 1.69 ❍

❍ 3 languages

2 languages

2.17 2.21 ❍ 2.37

fair

2.91 3.09 ❍

3.19

LProf3

LProf4

not good poor

LProf1

LProf2

Figure 10.8 Proficiency in language 1, 2, 3, and 4 by number of languages spoken. Source: Siemund et al. (2014: 355).

Enhancement Model as well as Herdina and Jessner’s (2002) Dynamic Model of Multilingualism. A more detailed analysis of the data reveals that self-assessments are consistent with official school test results for Chinese university and polytechnic students, while Malay students tend to overrate their abilities in English and perhaps somewhat underrate their mother tongue abilities. This suggests that the Chinese students are the best judges of their language skills, which, again, may be interpreted as a positive effect of their higher degree of multilingualism.

Multilingualism and Language Policy  221

On the Increasing Acceptance of CSE as a National Solidarity Code As mentioned earlier, despite the strong opposition of the Singaporean government to CSE, this variety serves as the language for daily informal interactions among Singaporeans and a language of identity. And contrary to the belief that CSE impedes the learning of Good English by policy makers, studies have found that as the level of education rises, increasing numbers of people are able to switch without difficulty between CSE and Singapore Standard English (Cavallaro and Ng 2009). This confirms the assumption of Woolard (1985) that nonstandard linguistic codes may be alternative rather than oppositional. While institutionalized social forces are working on language development in Singapore, there are alternative forces that support the existence of CSE, promoting its legitimacy as a form of unique Singaporean culture. Many studies have shown that CSE is a variety that fosters high solidarity and is perceived as a language of solidarity (see Gupta 1994; Leimgruber 2014; Simpson 2007; Tan 2014; Wong 2005). Siemund et al. (2014: 356) pose the attitudinal questions shown in Table 10.1 to two groups of university and polytechnic students. They were meant to elicit the students’ attitudes toward English, their mother tongue, and CSE. Here, we are only interested in statement 7, i.e. “I think speaking Singlish is a critical part of my self-definition.” Figure 10.9 shows that university students agree significantly more often with this statement than polytechnic students, even though university students represent the higher social group. The result can be interpreted as another indicator for the vitality of CSE, as, apparently, the socially higher groups are willing to adopt it as a local solidarity code. Similar findings are reported in Leimgruber (2014) and Tan (2014). Leimgruber (2014) conducted a survey among young educated Chinese students at a university in Singapore, in which students were required to define CSE Table 10.1 Statements eliciting language attitudes #

Statement

S1 S2

Proficiency in English is more important than mother tongue proficiency. The mother tongue should only be maintained if it is not done at the expense of English. It is normal for Singaporean speakers to speak only in English. You are still a Singaporean even if you don’t speak your mother tongue very often. You are still a Singaporean even if you can’t speak your mother tongue. I don’t think speaking my mother tongue is a critical part of my self-definition. I think speaking Singlish is a critical part of my self-definition.

S3 S4 S5 S6 S7

Source: Siemund et al. (2014: 356).

222  Peter Siemund and Lijun Li + University students

x Polytechnic students

strongly agree agree agree somewhat

2.61 2.70 + x

neutral

3.38 3.83 x 3.87 3.85 + x +

2.53 + 2.83 x

2.98 3.09 + x

disagree somewhat

4.03 4.34 x +

3.37 + 3.70 x

disagree strongly disagree

S1

S2

S3

S4

S5

S6

S7

Figure 10.9 Attitudes toward English, Colloquial Singapore English, and mother tongue. Source: Siemund et al. (2014: 356).

and give attitude ratings. The results show a strong disagreement (48 percent strong disagreement and 37 percent disagreement) with the government’s belief that Singlish is “bad English” and that “it would be better for Singapore if Singlish did not exist” (Leimgruber 2014: 53). The majority of respondents regard CSE as an identity marker that makes them Singaporeans (12 percent strongly agreed and 47 percent agreed) and as a language that unites different ethnicities of Singapore (28 percent strongly agreed and 69 percent agreed). Tan (2014) also explores the attitudes toward CSE according to age and ethnic background. There is consistent age grading, with the youngest groups reporting the most positive attitudes toward Singlish (Tan 2014: 336). Moreover, the differences between the three ethnic groups are not particularly striking. Apparently, all ethnic groups are inclined to embrace CSE as their code of national identity. Li (2020) offers an approximation of the diachronic development of CSE by counting the frequencies of the aspect marker already and the additive adverb also in this variety—plausible indicators of the nonstandard code (Siemund and Li 2017)—among speakers who were born between 1899 and 1958. The data come from the Oral History Interviews collected by the National Archives of Singapore, the interviews having been conducted between 1979 and 2008 (NAS 2020). In sociolinguistics, certain linguistic features often identify ethnicity. Although variation will exist in both the ethnic community and in the mainstream community, they differ with regard to the frequency of variants (see Tagliamonte 2013). The aspect marker already and the additive marker also serve as good indicators of Chinese influence on CSE. For example, in addition to the use of already as a phasal polarity expression (see van der

Multilingualism and Language Policy  223 Auwera 1993) to mean something happened earlier than expected in mainstream varieties of English, the CSE already functions as the perfect aspect marker, which is usually expressed by the past tense or the present perfect in the standard varieties, as shown in example 5. Reduplication of already is also found in CSE, which parallels the Chinese double le or yǐjīng + le expressions, as exemplified by 6 and 7. 5 6 7

I bought a place already. “I have bought a place.” (Bao and Hong 2006: 108) After all this news going on, after the screening, the people were already living in fear already. [NAS 2020: Oral History Interviews-000265-LTS] By the time they talked about globalization all of Singapore’s economy was already oriented toward a global marketplace already. [NAS 2020: Oral History Interviews- 002715-CCB]

As shown in Figure 10.10, the results suggest an increase in the use of also over the whole period, a slight decline of already between interviewees who were born in 1900–20 and 1920–40, and an increase over the period of 1920–60.6 Considering our frequency data and the attitudinal results, we can detect no marked decrease and devaluation of CSE. On the contrary, CSE seems to be gaining popularity among young educated Singaporeans, serving as a national solidarity code and an alternative language alongside Standard Singapore English to express their identity. There is no convincing distributional evidence that CSE is losing out. Notwithstanding, additional grammatical and phonological features of CSE would need to be investigated to provide a sharper picture.

Frequency per page

1.0

0.9

0.8

0.4

0.7

0.6

0.6

0.5 0.3

0.2

1900-1920

1920-1940

0.2 0.0

already

1940-1960

also

Figure 10.10 Frequencies of the aspect marker already and the additive adverb also in Colloquial Singapore English. Source: Li (2020).

224  Peter Siemund and Lijun Li

Summary and Conclusion The aim of this chapter was to explore the impact of a series of language policy measures enforced in Singapore over the past few decades on its multilingual landscape. It seems reasonable to argue that the top-down language policies of the government have exerted a strong influence on the language shift observable in Singapore. The city-state seems to be transiting from multilingualism toward the official bilingual model, i.e. English and one other official language. However, the increase in the use of English and Mandarin in Singapore may also be driven by world economic trends. The SMC has been a great success (from the point of view of its instigators), as evidenced in growing positive attitudes toward speaking Mandarin and a rise in the use of Mandarin as a home language. It has also successfully reduced the use of Chinese dialects as home languages. The SGEM, on the other hand, has been unable to achieve its goals. On the contrary, there is a growing tendency to embrace CSE as a national solidarity code. Curiously enough, attitudes toward CSE seem even more positive in the upper social strata. We may add that the official bilingual policy may well have a stabilizing effect on CSE, since it can be viewed as a blend of English words in combination with Chinese and Malay grammar. As the trend toward English and the other official languages continues, it remains to be seen what the long-term effects of these policies will be.

Notes 1 In the Singaporean context, the term “race” is used in everyday discourse to denote ethnicity without the negative connotation prevalent in the European context (see Leimgruber 2013b: 11). 2 In 2007, a new subject called “China Studies” was introduced at pre-university level and was taught in Mandarin only. It includes lessons about China’s modern history, political structures, economic development, and future challenges. Since 2016, this subject has also been taught in English, so that students can choose between English and Mandarin. 3 Nanyang 南洋 literally means ‘southern ocean’; it refers to Southeast Asia (see LePoer 1991: 305). 4 Questions concerned the languages in the individual repertoires, the order in which they were acquired, their self-assessed proficiency in different CEFR domains. 5 The prevalence of trilingualism among polytechnic students may also be interpreted as a social class effect (see Leimgruber et al. 2018: 299). 6 One speaker born in 1899 was placed in the group of 1900–20.

References Alsagoff, Lubna. 2015. “Creative Enactments of Language Teacher Education Policy: A Singapore Case Study.” In International Perspectives on English Language Teacher Education: Innovations from the Field, edited by Thomas S. C. Farrell, 125–41. International Perspectives on English Language Teaching. London: Palgrave Macmillan UK. doi:10.1057/9781137440068_8

Multilingualism and Language Policy  225 Ansaldo, Umberto, Lisa Lim, and Salikoko S. Mufwene. 2007. “The Sociolinguistic History of the Peranakans: What It Tells Us about ‘Creolization’.” In Deconstructing Creole, edited by Umberto Ansaldo, Stephen Matthews, and Lisa Lim, 203–26. Amsterdam: John Benjamins. doi:10.1075/tsl.73.11ans Bao, Zhiming. 1995. “Already in Singapore English.” World Englishes 14, no. 2: 181–88. ———. 1998. “The Sounds of Singapore English.” In English in New Cultural Contexts: Reflections from Singapore, edited by Joseph A. Foley et al., 152–74. Singapore: Oxford University Press. doi:10.1111/j.1467-971x.1995.tb00348.x ———. 2005. “The Aspectual System of Singapore English and the Systemic Substratist Explanation.” Journal of Linguistics 41, no. 2: 237–67. doi:10.1017/ s0022226705003269 ———. 2015. The Making of Vernacular Singapore English: System, Transfer, and Filter. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139135375 ———, and Huaqing Hong. 2006. “Diglossia and Register Variation in Singapore English.” World Englishes 25, no. 1: 106–14. doi:10.1111/j.0083-2919.2006.00449.x ———, and Khin Khin Aye. 2010. “Bazaar Malay Topics.” Journal of Pidgin and Creole Languages 25, no 1: 155–71. doi:10.1075/jpcl.25.1.06bao Bokhorst-Heng, Wendy. 1999. “Singapore’s ‘Speak Mandarin Campaign’: Language Ideological Debates in the Imagining of the Nation.” In Language Ideological Debates, edited by Jan Blommaert, 235–65. Berlin: De Gruyter. doi:10.1515/9783110808049.235 Cavallaro, Francesco, and Bee C. Ng. 2009. “Between Status and Solidarity in Singapore.” World Englishes 28, no. 2: 143–59. doi:10.1111/j.1467-971x.2009.01580.x ———. 2021 forthcoming. “Multilingualism and Multiculturalism in Singapore.” In Multilingual Global Cities: Singapore, Hong Kong, Dubai, edited by Peter Siemund and Jakob Leimgruber. Abingdon and New York: Routledge. Cavallaro, Francesco, and Stefan K. Serwe. 2010. “Language Use and Language Shift among the Malays in Singapore.” Applied Linguistics Review 1, no. 1: 129–70. doi:10.1515/9783110222654.129 Chew, Phyllis. 1999. “Linguistic Imperialism, Globalism and the English Language.” In English in a Changing World, edited by David Graddol and Ulrike H. Meinhof, 37–47. Milton Keynes: AILA. Chua, Beng-Huat. 1995. Communitarian Ideology and Democracy in Singapore. London: Routledge. doi:10.4324/9780203033722 Cohen, James, ed. 2005. ISB4: Proceedings of the 4th International Symposium on Bilingualism. Somerville, MA: Cascadilla Press. Deterding, David. 2007. Singapore English. Edinburgh: Edinburgh University Press. doi:10.3366/edinburgh/9780748625444.001.0001 Dixon, L. Quentin. 2005. “The Bilingual Education Policy in Singapore: Implications for Second Language Acquisition.” In ISB4: Proceedings of the 4th International Symposium on Bilingualism, edited by James Cohen et al., 625–35. Somerville, MA: Cascadilla Press. Flynn, Suzanne, Claire Foley, and Inna Vinnitskaya. 2004. “The Cumulative Enhancement Model for Language Acquisition: Comparing Adults’ and Children’s Patterns of Development in First, Second and Third Language Acquisition of Relative Clauses.” International Journal of Multilingualism 1, no. 1: 3–16. doi:10.1080/14790710408668175

226  Peter Siemund and Lijun Li Goh, Siew Huang Melissa, Cherie Ying Wei Lim, and Fiona Yiling Seah. 2016. “My Father Tongue: A Campaign to Revitalise the Use of Chinese Dialects among Youth in Singapore.” BA Final Year Project, Nanyang Technological University. Accessed July 4, 2017 at https://hdl.handle.net/10356/66800. Gopinathan, Saravanan. 1998. “Language Policy Changes 1979–1997: Politics and Pedagogy.” In Language, Society, and Education in Singapore: Issues and Trends, edited by Saravanan Gopinathan et al., 19–44. 2nd ed. Singapore: Times Academic Press. Gupta, Anthea F. 1994. The Step-Tongue: Children’s English in Singapore. Clevedon: Multilingual Matters. Herdina, Philip, and Ulrike Jessner. 2002. A Dynamic Model of Multilingualism: Perspectives of Change in Psycholinguistics. Clevedon: Multilingual Matters. doi:10.21832/9781853595547 Ho, Lick C., and Lubna Alsagoff. 1998. “English as the Common Language in Multicultural Singapore.” In English in New Cultural Contexts: Reflections from Singapore, edited by Joseph Foley, 201–17. Singapore: Singapore Institute of Management. Hoon, Chng H. 2003. “‘You See Me No Up’: Is Singlish a Problem?” Language Problems and Language Planning 27, no. 1: 45–62. doi:10.1075/lplp.27.1.04hoo Lee, Kuan Yew. 2000. From Third World to First: The Singapore Story, 1965–2000. New York: HarperCollins Publishers. Leimgruber, Jakob R.E. 2013a. “The Management of Multilingualism in a CityState.” In Multilingualism and Language Diversity in Urban Areas: Acquisition, Identity and Education, edited by Peter Siemund et al., 227–56. Amsterdam: John Benjamins. doi:10.1075/hsld.1.12lei ———. 2013b. Singapore English: Structure, Variation, and Usage. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139225755 ———. 2014. “Singlish as Defined by Young Educated Chinese Singaporeans.” International Journal of the Sociology of Language 230: 45–63. doi:10.1515/ijsl-2014-0026 ———, Peter Siemund, and Laura Terassa. 2018. “Singaporean Students’ Language Repertoires and Attitudes Revisited.” World Englishes 37, no. 2: 282–306. doi:10.1111/weng.12292 LePoer, Barbara Leitch. 1991. “Historical Setting.” In Singapore: A Country Study, edited by Barbara Leitch LePoer, 1–64. 2nd ed. Washington, DC: Federal Research Division. Li, Lijun. 2020. “Language Contact: A Historical Sociolinguistic Reconstruction of Colloquial Singapore English in Relation to its Chinese Substrates,” PhD diss., University of Hamburg. ———, Stefan Lazarov, and Peter Siemund. 2020. “Ages of Pragmatic Particles in Colloquial Singapore English: A Corpus-Based Study.” Poster presented at New Ways of Analyzing Variation, Asia Pacific 6, February 19–22, National University of Singapore. Li, Wei, Vanithamani Saravanan, and Julia L. H. Ng. 1997. “Language Shift in the Teochew Community in Singapore: A Family Domain Analysis.” Journal of Multilingual and Multicultural Development 18, no. 5: 364–84. doi:10.1080/01434 639708666326 Lim, Lisa. 2010. “Migrants and ‘Mother Tongues’: Extralinguistic Forces in the Ecology of English in Singapore.” In English in Singapore: Modernity and Management, edited by Lisa Lim, Anne Pakir, and Lionel Wee, 19–54. Hong Kong: Hong Kong University Press. doi:10.5790/hongkong/9789888028436.003.0002

Multilingualism and Language Policy  227 ———, Anne Pakir, and Lionel Wee. 2010. English in Singapore: Modernity and Management. Hong Kong: Hong Kong University Press. doi:10.5790/ hongkong/9789888028436.001.0001 Lock, Graham. 1989. “Aspects of Variation and Change in the Mandarin Chinese Spoken in Singapore.” Australian Journal of Linguistics 9, no. 2: 277–94. doi:10.1080/07268608908599423 MOE (Ministry of Education, Singapore). 2018. “General Information on Studying in Singapore.” Accessed October 4, 2018 via https://www.moe.gov.sg. NAS (National Archives of Singapore). 2020. “Oral History Interviews.” Accessed February 20, 2020 via https://www.nas.gov.sg/archivesonline. Ng, Patrick. 2011. “Language Planning in Action: Singapore’s Multilingual and Bilingual Policy.” Ritsumeikan Asia Pacific Journal 30: 1–12. ———. 2014. “A Study of Attitudes towards the Speak Mandarin Campaign in Singapore.” Intercultural Communication Studies 23, no. 3: 53–65. Pakir, Anne. 1994. “Education and Invisible Language Planning: The Case of English in Singapore.” In English and Language Planning: A Southeast Asian Contribution, edited by Thiru Kandiah and John Kwan-Terry, 158–79. Singapore: Centre for Advanced Studies, National University of Singapore. ———. 2008. “Bilingual Education in Singapore.” In Encyclopedia of Language and Education, edited by Nancy H. Hornberger, 191–203. Boston: Springer. doi:10.1007/978-0-387-30424-3_125 Rappa, Antonio L., and Lionel Wee. 2006. Language Policy and Modernity in Southeast Asia: Malaysia, the Philippines, Singapore, and Thailand. New York: Springer. doi:10.1007/0-387-32186-1 Rubdy, Rani. 2007. “Singlish in the School: An Impediment or a Resource?” Journal of Multilingual and Multicultural Development 28, no. 4: 308–24. doi:10.2167/ jmmd459.0 Schiffman, Harold F. 2002. “Tongue Tied in Singapore: A Language Policy of Tamil?” Journal of Language, Identity and Education 2, no. 2: 105–25. doi:10.1207/ s15327701jlie0202_2 Siemund, Peter, and Lijun Li. 2017. “Towards a Diachronic Reconstruction of Colloquial Singapore English.” In Negation and Contact: With Special Focus on Singapore English, edited by Debra Ziegeler and Zhiming Bao, 11–32. Amsterdam: John Benjamins. doi:10.1075/slcs.183.02sie ———, Monika E. Schulz, and Martin Schweinberger. 2014. “Studying the Linguistic Ecology of Singapore: A Comparison of College and University Students.” World Englishes 33, no. 3: 340–62. doi:10.1111/weng.12094 Sim, Cheryl. 2016. “Bilingual Policy.” Singapore Infopedia. Accessed October 5, 2018 via http://eresources.nlb.gov.sg/infopedia. Simpson, Andrew. 2007. “Singapore.” In Language and National Identity in Asia, edited by Andrew Simpson, 374–90. Oxford: Oxford University Press. Tagliamonte, Sali A. 2013. Variationist Sociolinguistics: Change, Observation, Interpretation. Malden, MA: Wiley-Blackwell. Tan, Eugene K. 2007. “The Multilingual State in Search of the Nation: The Language Policy and Discourse in Singapore’s Nation-Building.” In Language, Nation and Development in Southeast Asia, edited by Hock Guan Lee and Leo Suryadinata, 74–113. Singapore: ISEAS-Yusof Ishak Institute. Tan, Ying-Ying. 2014. “English as a ‘Mother Tongue’ in Singapore.” World Englishes 33, no. 3: 319–39. doi:10.1111/weng.12093

228  Peter Siemund and Lijun Li Tsui, Amy and James W. Tollefson. 2006. Language Policy, Culture and Identity in Asian Contexts. New York: Routledge. doi:10.4324/9781315092034 Vaish, Vinti, Aisha Jamaluddeen, and Mardiana Roslan. 2009. “The Sociolinguistic Survey of Singapore 2006: Findings and Policy Implications.” Research Brief 09-001. Accessed July 4, 2016 via http://www.nie.edu.sg. van der Auwera, Johan. 1993. “‘Already’ and ‘Still’: Beyond Duality.” Linguistics and Philosophy 16, no. 6: 613–53. doi:10.1007/bf00985436 Wee, Lian-Hee. 2008. “Phonological Patterns in the Englishes of Singapore and Hong Kong.” World Englishes 27, no. 3–4: 480–501. doi:10.1111/j.1467-971x.2008.00580.x Wee, Lionel. 2003. “Linguistic Instrumentalism in Singapore.” Journal of Multilingual and Multicultural Development 24, no. 3: 211–24. doi:10.1080/01434630308666499 Wong, Jock. 2005. “‘Why You So Singlish One?’: A Semantic and Cultural Interpretation of the Singapore English Particle One.” Language in Society 34, no. 2: 239–75. doi:10.1017/s0047404505050104 Wong, Wee K. 2011. Census of Population 2010. Statistical Release 1: Demographic Characteristics, Education, Language and Religion. Singapore: Department of Statistics. ———. 2016. General Household Survey 2015. Singapore: Department of Statistics, Ministry of Trade and Industry, Republic of Singapore. ———. 2019. Population Trends 2019. Singapore: Department of Statistics, Ministry of Trade and Industry, Republic of Singapore. Woolard, Kathryn A. 1985. “Language Variation and Cultural Hegemony: Toward an Integration of Sociolinguistic and Social Theory.” American Ethnologist 12, no. 4: 738–48. doi:10.1525/ae.1985.12.4.02a00090 Ziegeler, Debra. 2015. Converging Grammars: Constructions in Singapore English. Boston: De Gruyter Mouton.

11 The Discourses of lào yīngwén Resistance to and Subversion of the Normative Status of English in Taiwan Hsi-Yao Su Abstract Analyzing language attitudes toward English in Taiwan, this chapter explores the slang term lào yīngwén, roughly meaning ‘to speak English fluently (in a showy manner)’, often with a jocular and disapproving connotation. The analysis is based on data collected from well-known bloggers’ posts about lào yīngwén. These data are used to investigate the subversive reactions to the normative status of English as a global language and as a symbol of workplace competitiveness. In other words, these posts represent competing discourses on the use of English in a non-English, sinophone environment. The use of English becomes more widespread globally, and this chapter contributes to a general understanding of the many ways in which English is discursively constructed and ideologically represented as playing particular roles and indexing particular identities, personae, or images in various local contexts.

Introduction English changes a child’s future. (by U.S. KIDS Language School) I spotted this slogan outside of a children’s private language school right across from an elementary school in my neighborhood in Taipei. As a linguist, I found the message rather amusing: Every language changes a child’s future if we consider the idea of linguistic relativity (or the Sapir-Whorf hypothesis) which, put simply, argues that the structure of a language affects speakers’ worldviews (Whorf 1956). But clearly, the slogan does not relate to the Whorfian concept that any language necessarily affects the ways a speaker conceptualizes the world. Instead, it highlights the importance of English (as opposed to other local languages) for social mobility and career advancement. Slogans like this and the underlying enthusiasm for English are, in fact, very common in Taiwan. A glimpse at the government proposals for the promotion of English reveals this predilection. In 2002, then-President Chen Shui-bian 陳水扁 proposed that English be designated as an official language. Along the same line, in 2003, then-Premier Yu Shyi-kun 游錫堃

230  Hsi-Yao Su announced that English was to be adopted as a semi-official language in eight to ten years. The goal was not pursued by the Ma Ying-jeou 馬英九 government (2008–16) after Chen’s presidency, but in 2018, the Executive Yuan, the executive branch of the Taiwan government, announced Blueprint for Developing Taiwan into a Bilingual Nation by 2030, in which “bilingual” refers to Mandarin Chinese and English. In the educational realm, the starting grade of English classes in the standard curriculum has lowered significantly in the past two decades. In Taipei and some other major cities, pupils start English classes in grade 1. In a similar fashion, private language schools and bilingual or even all-English kindergartens are quite prevalent, especially in metropolitan areas. According to the Information and Management System of Short-term Tutorial Center (直轄市及各縣市短期補習 班資訊管理系統) established by the Ministry of Education, the number of foreign language tutorial centers for elementary school students increased from 2,199 in 2008 to 2,674 in 2017 nation-wide, and those targeting junior high school students rose from 1,309 in 2008 to 1,602 in 2017.1 These policies and statistics are a symbolic statement of the English “rush” in Taiwan. The emphasis placed by the government and the public on English and international competitiveness, coupled with the popular concept of “the critical period of language learning,” prompts many parents to see English as an important resource to be acquired as early as possible. (See also Chen, this volume, and Wasserfall, this volume, for discussions on the implementation of English language policy and actual practice of English in Taiwan.) The English “rush” in the educational realm, however, is sometimes paralleled by negative sentiments toward the use of English in other everyday life contexts. In this chapter, I explore the subversive discourse on the global and normative status of English. Specifically, I explore a slang term, lào yīngwén or làu ing-bûn—depending on whether it is pronounced and Romanized in Mandarin or in Taiwanese Southern Min (a.k.a. Taiwanese, the term I will use hereafter)—which roughly means “to speak English fluently (in a showy manner).” It is a spoken slang term originating in Taiwanese, but it was later borrowed into Mandarin. Recently, the slang term has acquired its written forms and is popularly written as either 烙英文，落英文, or 撂 英文. The three initial characters 烙 (lào, originally meaning ‘to brand’), 落 (lào, originally meaning ‘to drop’), and 撂 (liào, originally meaning ‘to put down’) refer to the same verb that is meant to be pronounced as làu in Taiwanese, which roughly means ‘to speak a (foreign/standard) language fluently (in a showy manner)’. The Taiwanese Southern Min Dictionary of Common Expressions (台灣閩南語常用詞辭典) published by the Ministry of Education (hereafter MOE) in 2008 proposes that the underlying word is the verb normally written using the character 落 (see the next section for more information). But since most Taiwanese people do not practice written Taiwanese regularly and are not familiar with the standard suggested by MOE, the slang term varies in its casual written forms. The latter two characters, 英文 (Romanized as yīngwén when pronounced in Mandarin), refer

The Discourses of lào yīngwén  231 to the English language. In everyday usage, lào yīngwén generally denotes code-mixing or switching into English either in a predominantly Chinese conversation or in a context where Chinese is expected, often in a jocular and at times disapproving manner. The very existence of such a term and the increasing popularity of its written forms indicate that Taiwanese generally are acutely aware of the English-Chinese code-mixing practice and its associated social meanings/ evaluations. The slang term provides a discursive site where attitudes toward English code-mixing are expressed and thus is a prime locus for exploring the relationship between code-mixing, local identity, and language ideologies in this increasingly globalized world where English is seen as playing a major role. The existing literature on language attitudes toward English in Taiwan generally focuses on those of learners in educational settings. There is little research on the tension between local identity and global competitiveness associated with English and on the language ideologies concerning the code-mixing practice in everyday life. This chapter, therefore, is an attempt to explore the subversion of the normative discourse about English as a global language in the Taiwanese context. Along with four other chapters in this volume addressing the influence of English in the sinophone world— including Chen, Goh and Fong, Meierkord, and Wasserfall—this chapter presents one aspect of such influence that has previously been little explored. More specifically, this chapter explores online discussions about lào yīngwén and analyzes these in terms of Bucholtz and Hall’s (2004, 2005, 2011) framework of language and identity construction and Blommaert’s (2010) notion of sociolinguistic scale. It is also more broadly situated within the studies of language ideologies (e.g. Schieffelin et al. 1998; Woolard and Schieffelin 1994) and indexicality (e.g. Ochs 1992; Silverstein 2003). It aims to address the following questions. First, how is lào yīngwén evaluated in related online discourses? Second, how are code-mixers discursively constructed as “others” in lào yīngwén-related online discourses? Third, how does the tension between local identity and global/cosmopolitan aspiration play out in related online discourses?

The Definition of lào yīngwén/làu ing-bûn As an emergent written slang term, the origin of lào yīngwén is little researched. Most of the Taiwanese with whom I have discussed the term believe that the verb lao/lau has its origins in Taiwanese and was later borrowed into Mandarin. According to the Taiwanese-Japanese Dictionary (台日大辭 典, Ogawa 1931), the Taiwanese verb láu (老) has a range of meanings, two of which are especially relevant: one entry is ‘to master’ something (老練, 熟 練), with the example “speaking the standard language fluently” (官話不止 láu). The other entry is ‘to speak with the tongue rolling’ (捲舌講話), with the examples “speaking the standard language (Beijing speech)” (láu 正字), and “can’t speak English fluently” (英國話 láu bōe來). In the aforementioned

232  Hsi-Yao Su Taiwanese Southern Min Dictionary of Common Expressions (MOE 2008), làu (落) has the meaning of ‘showing off one’s language ability’ (炫耀語言能 力). An example given is “Every time he speaks to me, he likes to show off ̍ kái kah guá kóng-uē lóng ài làu Ing-gí 伊逐改佮我講話攏 his English” (I tak 愛落英語).2 The dictionary entries reveal that the combination of láu followed by a standard language or a foreign language has existed for at least several decades in spoken Taiwanese. Yet its formulaic written versions— 落英文, 烙英文, and 撂英文—are relatively new. As mentioned previously, even though the current written suggestion by MOE is 落英文, the three initial characters, 落 (lào), 烙 (lào), and 撂 (liào), are used interchangeably because many Taiwanese are not familiar with MOE’s suggested use of written Taiwanese. The three characters have been adopted largely because of their phonological resemblance to the targeted Taiwanese verb lau. The prescriptive standard, 落 (lào), can be pronounced the same in both Mandarin and Taiwanese. The second one, 烙 (lào), involves using the character’s Mandarin pronunciation as the transliteration of the targeted verb’s Taiwanese pronunciation. The third one, 撂 (liào), is adopted possibly both because of its semantic and phonological proximity to the intended verb: in Mandarin there is a common phrase liào hěn huà (撂狠 話, ‘talk rough’) and one of the basic meanings of the verb 撂 is ‘to drop’. The verb 撂 thus carries the meaning of ‘dropping certain words in a provocative manner’, which is implied in the slang term; and the verb 撂 also resembles the phonology of the intended verb. The lack of a unified written practice illustrates a common phenomenon in Taiwan in that even though government efforts are made to standardize the Taiwanese language, the majority of Taiwanese speakers do not usually write Taiwanese, and when they do, most of them just rely on intuition. A further piece of evidence on the emergent status of the written slang forms comes from Google search results. A Google search of the three versions of the slang term shows that most of the search results are dated within the past decade, and there has been an increase in the use of the term in news headlines within the past five years, which indicates growing acceptance of the borrowed slang term in Taiwan Mandarin.3 In everyday usage nowadays, lào yīngwén refers to code-switching to English in a Chinese-based conversation. But what lào yīngwén describes is not always just a simple, neutral switch. It can also imply a certain degree of showiness. The phrase lào yīngwén can thus invoke a negative evaluation of such a code-mixing practice, as illustrated by the dictionary entry mentioned earlier and the following two excerpts from online sources. The first example is the featured response to a question “What does 撂英文 mean?” in HiNative, “a global Q&A platform for language learners” in which one can “ask and answer questions about cultures and languages with native speakers around the world,” as described by their own website (HiNative 2018).

The Discourses of lào yīngwén  233 Excerpt (1) 講英文的意思 但是很口語化 而且比較負面一點 例如母語是中文的時候 你用英文和別人說話 別人說你撂英文 他表達的意思比較像是你故意講英文而不講中文就好 It means ‘speaking English’. But it’s very colloquial, and has a negative connotation. For example, in a Chinese-as-native-language context, if you speak English with others, others can say that you lào yīngwén. What he/she implies is that you choose to speak English on purpose when you can simply speak Chinese. The second example is a post in a popular blog titled Sān fēnzhōng rèdù (三 分鐘熱度; Short-lived enthusiasm; Pixnet 2011): Excerpt (2) 台灣的職場上，有一句俗諺叫「烙英文」。這是指有人喜歡在非必要的時候 講幾個英文單字(通常都講不出一句完整的英文)，目的是顯示出自己的英 文程度不錯，還有讓週圍英文不好的人感覺很不爽。這些烙英文的人，通 常中文也講得很好，但不知道為什麼有些時候，他就是想講英文。烙英文 的人，常會被罵做作、愛現。 In the Taiwanese workplace, there is a common slang term, lào yīngwén. It refers to some people’s tendency to insert a few English words when the use of English is not called for (and they usually cannot produce a complete English sentence). The goal is to show their English proficiency and to irritate those whose English is not as good. These people who lào yīngwén usually speak Chinese well, but for some unknown reason, they just want to speak English. They are often criticized as pretentious and showy. The two excerpts reveal the at times negative sentiments toward the use of English code-mixing in Taiwan, which is in sharp contrast with the discourses that positively associate English with career advancement and global competitiveness (see Lee 2012; Lin 2014, 2018 for detailed discussions). Before I turn to a detailed discussion of the subversive discourses about English, I shall first lay out the theoretical frameworks used to examine such discourses in the following section.

Theoretical Frameworks Taiwanese are not intolerant of code-switching in general. In fact, code-switching between Mandarin and Taiwanese is an extremely common

234  Hsi-Yao Su practice and people are usually quite accepting of it in everyday contexts.4 The negative evaluation of English code-mixing, therefore, cannot be attributed to an intolerant attitude toward language mixing in general. Rather, it should be approached from the relationship between language and identity. Bucholtz and Hall (2004, 2005, 2010) propose a theoretical framework which investigates identity as constituted in linguistic interaction. The central premise of their framework is the emergence principle, which views identity as “the emergent product rather than the preexisting source of linguistic and other semiotic practices and therefore as fundamentally a social and cultural phenomenon” (2010: 19). In other words, identity is treated not as a given, but as constructed in discourse. This premise about the discursive nature of identity is highly relevant to this study because my focus in this chapter is on how people who code-mix are evaluated and constructed as a certain type. To explore how such a discursive construction of identity takes place, we need two other principles in Bucholtz and Hall’s framework, the indexicality principle and the relationality principle. The indexicality principle states that identity relations emerge in interaction through several related indexical processes, including (a) overt mention of identity categories and labels; (b) implicatures and presuppositions regarding one’s own or others’ identity position; (c) displayed evaluative and epistemic orientations to ongoing talk, as well as interactional footings and participant roles; and (d) the use of linguistic structures and systems that are ideologically associated with specific personas and groups (2010: 21). The indexicality principle points to possible loci to examine the indexical processes through which linguistic elements and identity categories or stances are linked. The relationality principle complements it with three sets of common discursive strategies through which identities are constructed. It states that identities are intersubjectively constructed through several, often overlapping, complementary relations, including similarity/difference, genuineness/artifice and authority/delegitimacy (2010: 23). In other words, identities and stances are often created and made explicit in discourses through the processes of adequation/distinction, authentication/denaturalization, and authorization/illegitimation, respectively, which correspond to the three complementary relations mentioned earlier. Bucholtz and Hall’s framework can be used to explore how the practice of English code-mixing is discursively constructed as the language of an “other,” i.e. an unpopular type of Taiwanese different from the general Taiwanese and the authors of lào yīngwén-related discourses. But in this case, the identity construction is also intertwined with tension between local and global identity/orientation, which Bucholtz and Hall’s framework does not readily address. Blommaert’s (2010) framework of sociolinguistics of globalization is a useful tool to explore the local-global ambivalence implicit and explicit in the data. Blommaert’s framework includes three independent, but closely related concepts: sociolinguistic scales, orders of indexicality, and polycentricity. The concept of sociolinguistic scale starts from the observation that “social events and processes move and develop on a continuum of

The Discourses of lào yīngwén  235 layered scales, with the strictly local (micro) and the global (macro) as its extremes and several intermediary scales (e.g., the level of the state) in between” (2010: 32). In social interactions, we often encounter “scale-jumping.” For example, in discussing the role of English in Taiwan, we may sometimes jump from orientation toward the local scale/TimeSpace (e.g. the minimal usefulness of English in the everyday context in small towns) to orientation toward the global scale/TimeSpace (e.g. English as a resource in international arenas in current time). These moves often index norms and social order. Different scales are related to different patterns of normativity, but these patterns of normativity are not valued equally. Some forms of semiosis are systematically taken as more valuable, others as less valuable, and still others are not taken into consideration at all. This is the stratified nature of orders of indexicality, and it is closely tied to a third element: center and polycentricity. Centers, whether real or perceived, are sites that produce authority, power, and norms. In brief, centers are evaluating authorities. When humans communicate, we are often oriented toward more than one center; thus interactions are almost always polycentric. These two theoretical frameworks provide us with useful tools to examine the discourses of lào yīngwén, but any meaningful analysis needs to be positioned in relation to Taiwan’s unique historical and social background. In light of the data analyzed in this study, it is important to highlight Taiwan’s historical trajectory from a pre-Qing periphery to a Japanese colony to a post-World War II, Cold War outpost with close connections with the United States, and to a fast-developing country with advanced medical resources and social welfare. Below I turn to methods and data analysis.

Data Analysis Data used in this study were collected from well-known online bloggers’ posts about lào yīngwén. The first example is from Wanwan 彎彎, a famous cartoonist/blogger in Taiwan. The post is entitled lào yīngwén (烙英文; Wanwan 2009), and is composed of a five-frame comic strip and author’s commentary. The comic strip portrays how a Taiwanese man, who likes to include a few English expressions in Chinese conversations, fails to communicate in English with a foreigner. Excerpt 3 presents the comic strip (Figure 11.1), while excerpt 4 presents the author’s commentary immediately following the comics in the post. In the translation of excerpt 3, the expressions originally in English are in bold. Excerpt (3) Translation (Right side, first frame) Taiwanese man (left): Wanwan (right):

 Pa-jie – Fa-wo.

236  Hsi-Yao Su

Figure 11.1 A cartoon about lào yīngwén. Source: Wanwan (2009), reprinted with permission by Wanwan.

(second frame) Taiwanese man: Wanwan: (third frame) Taiwanese man:

I’ll cue you when the time comes. That cut. Q? Ka.. I asked the waiter for a menu. Way over the top. Taiwanese are too…

The Discourses of lào yīngwén  237 Wanwan: (Left side, first frame) Western man (right): (second frame) Taiwanese man (left): Wanwan (far left):

Wei-te, mian-niu, o-fu-er Excuse me…? Uh, uh… (thinking “aloud”) So those who like to lào yīngwén do not necessarily speak English well.

Excerpt (4) 出社會之後漸漸遇到好多英文單字阿 比較尷尬的是大家都在講然後自己聽不懂又要裝懂 (雖然也常遇到這種很愛烙英文，假ABC 但卻根本就只是半瓶水阿…) Since entering the working world, I’ve encountered so many English words. It’s embarrassing when everyone’s using them, and I have to pretend that I know what they mean. (although I also often encounter people who like to lào yīngwén, the fake ABCs. In fact, they are just empty vessels that make the most noise.) This multimodal post, composed of a five-frame comic strip and author’s commentary, is rich in acts of identity construction, and the identities constructed are closely tied to a sense of localness and translocalness. In the first three frames at the right side, a clear distinction is made between the author, Wanwan (represented as a round head), and the Taiwanese man who codemixes frequently. A gap was created between the man’s fluent code-mixing and Wanwan’s fragmented understanding (or lack of understanding). Such a gap is textually created both by sentence structure (the man’s fluent, full sentences versus Wanwan’s single phrases) and by the contrast between the representations of the code-switched English words—one in the English alphabet and the other using Chinese characters as transliteration of the English words unfamiliar to the author. The nonsensical string of Chinese transliteration further highlights the lack of mutual understanding and invokes a sense of incongruity (and thus playfulness) in the three frames. The three frames clearly involve a process of distinction (Bucholtz and Hall 2010), in which the man who code-mixes and the author are constructed as two different types of persons through both the aforementioned linguistic details and other graphic elements, including the color contrast of the words of the code-mixer and Wanwan. In the third frame, the man complains that “Taiwanese are too…” (台灣人都太…), setting himself apart from Taiwanese in general. The three frames together construct two disparate identities: A code-mixer who distances himself from the general Taiwanese and sees himself as distinctive from (and possibly superior to) general Taiwanese, and

238  Hsi-Yao Su general Taiwanese who do not understand such a practice and thus are excluded from an effective communication. The three frames involve activities on a local scale and normativity oriented toward a local center (Blommaert 2010). As Wanwan’s own commentary (excerpt 4) reveals, the first three frames of the comic strip allude to the Taiwanese workplace in which English is often seen as a symbol of professionalism and career advancement, and to the not uncommon practice of inserting English into a Chinese-based conversation. But in frame four at the left side, as the situation changes, we see a scale jump from a local to a translocal one. In this more translocal scale, single English words are not seen as true English, but as fragmentary English, which does not aid international communication. Wanwan’s comment at the very end of the comic strip “so those who like to lào yīngwén do not necessarily speak English well” (原 來愛烙英文的人英文不一定強啊) further constructs lào yīngwén as not “real English.” A final act of “othering” English code-mixers appears after the comic in the author’s own commentary, as shown in (4), in which she labels code-mixers as “fake ABCs (American-born Chinese/Taiwanese)” and “a half-full glass of water” (which alludes to “Empty vessels make the most noise” 半瓶水響叮噹). In this comic post, it is quite clear that lào yīngwén is equated with English-Chinese code-mixing and is evaluated negatively. The negative evaluation is achieved through (1) distinguishing code-mixers from general Taiwanese and (2) deauthenticating code-mixers as neither “the general, modest local Taiwanese” nor the “real global citizens.” These processes of distinction and deauthentication (Bucholtz and Hall 2010) are enabled by scale jumping (Blommaert 2010) from a more local to a more translocal one in the comic strip. The code-mixer, thus, is depicted as a “global citizen wannabe” rather than someone who is truly globally oriented. The second example is a post on PTT, a popular BBS site in Taiwan (MicroGG 2014). The author is a well-known PTT figure nicknamed līngbēi gǔkē (拎杯骨科, hereafter LB for short), literally ‘your-father-orthopedist’. The first two characters, līngbēi, is a nonsensical combination literally meaning ‘carry’ and ‘cup’. Here the Mandarin Chinese pronunciations of the two characters are manipulated as the transliteration of the Taiwanese phrase “your father,” a vulgar self-address term that elevates the status of the speaker to that of the father of the addressee. The latter two characters stand for gǔkē ‘orthopedic’. The author is an orthopedist known for both his medical knowledge and his local and confrontational writing style. The post in question is a response to a PTT discussion about a news article which reported on the abuse of the Taiwanese national health insurance by some overseas Taiwanese. According to the original news report, some overseas Taiwanese suspend their health insurance while they are abroad and resume it only temporarily when they come back to Taiwan. They can thus get away with paying a small amount of money yet enjoy full national health insurance coverage. The news report was discussed on the gossip board of PTT, and LB responded with his first-hand experience as a practicing medical doctor with such patients. Although the core of the discussion was not

The Discourses of lào yīngwén  239 about lào yīngwén, language plays an important role in LB’s construction of such patients as an unpleasant category. In the interest of space, I summarize the content below and only provide excerpts when they are relevant to the analysis. LB begins the post with the signature opening “Fuck, I am Your-Father Orthopedic (LB),” followed by the explanation that he had not posted anything for a while but had to say something upon seeing the news about overseas Taiwanese abusing national health insurance. He then describes his experiences with such patients, who are depicted as not seeking treatment for a particular medical condition, but as demanding expensive physical examinations. He complains about such overseas Taiwanese patients, especially those from the United States, and criticizes their acts of lào yīngwén. He then addresses his fellow physicians and advises them not to bother persuading these patients that their conditions do not warrant the medical examinations they demand. Just insist on a physician’s professional judgment. Finally, he directly addresses such overseas Taiwanese, claiming that he has no need of them as patients to enhance his performance, and if they can afford it, please see a doctor in the United States. In this post, the overseas Taiwanese patients are clearly constructed as different from the “real, authentic Taiwanese.” We can analyze the construction of identity with Bucholtz and Hall’s indexicality principle, which identifies several indexical processes through which identities can be discursively and linguistically constructed. Below I discuss the relevant indexical processes evident in LB’s post one by one. The first indexical process involves overt mention of identity categories and labels. The overseas Taiwanese are labeled as “the mother-fucking overseas compatriots” (僑你媽的 雞八胞), “these people” (這些人), “those from the US” (美國來的), “banana people” (香蕉人), and “the overseas Taiwanese (僑胞) who only want a physical check-up” (只想健康檢查的僑胞), for instance. The term, qiáobāo (僑 胞) ‘overseas compatriot’ already has the implicit meaning of otherness. On top of that, these identity labels either are directly aggressive (such as “the mother-fucking overseas compatriots” and “banana people”) or explicitly set apart the overseas Taiwanese in question from the general Taiwanese by categorizing them as another type of people (“these people”), as coming from a different place (“those from the US”), and as having a particular purpose (“the overseas Taiwanese who only want a physical check-up”). In addition to direct identity labels, the overseas Taiwanese patients are further categorized as “other” through discursive means. Excerpt 5 below is an example in which LB describes such patients as people lacking medical knowledge yet attempting to exploit Taiwanese national health insurance. Excerpt (5) 這些人根本就不是來看病的 是存心來做健康檢查的 第一次進來門診，一開口就是要電腦共振/核磁斷層的 幹拎娘，名詞連講都講不對，拎杯為什麼要開給你

240  Hsi-Yao Su These people are not here to see a doctor They are here for a physical check-up First time here for a consultation, as soon as they open their mouths, they begin to ask for CRI/MAT scan (meant to mean CAT scan/MRI) Fuck your mother. You can’t even get the names right, why should I approve it? Here the two medical procedures, CAT scan (電腦斷層) and MRI (核磁共 振), are deliberately misused by LB in quoting the overseas Taiwanese patients. That is, the first two characters of each procedure are mismatched with the final two characters of the other procedure. This misuse effectively presents these overseas Taiwanese as lacking medical knowledge and as demanding more than they can understand. This image is in sharp contrast with LB’s identity as a physician. Through such an indirect quote, these overseas Taiwanese patients are simultaneously de-legitimized as people who have knowledge of medical issues and are distinguished from LB and other good, conscientious patients. Right after excerpt 5, LB criticizes the overseas patients’ use of lào yīngwén, as shown in excerpt 6. In this excerpt, the overseas Taiwanese are further portrayed as members of an out-group distinct from the local Taiwanese. Their otherness is at least partially achieved through LB’s negative evaluation of their acts of lào yīngwén. Excerpt (6) 尤其是美國來的啦，幹拎娘很愛烙英文是不是 阿我的 knee 很 pain，幹，pain 啥小 要 pain 懶叫出來，去隔壁泌尿科看啦 拎杯看病都只用台語看啦，幹 Especially those from the United States, you fucking love to lào yīngwén, right? Ah, my knee is in pain. Fuck. What freaking pain. If you are in pain, you can take out your penis and go to the urology department next door. I practice medicine only in Taiwanese, for fuck’s sake. In this excerpt, the overseas Taiwanese are constructed as code-mixers from the United States, while LB presents himself as a local Taiwanese physician, who speaks Taiwanese with his patients. Whether it is true that LB only uses Taiwanese in his medical practice is not known, but discursively the excerpt constructs the overseas Taiwanese as clearly distinct from a local Taiwanese identity through the association of the two identities with English and Taiwanese, respectively.

The Discourses of lào yīngwén  241 A strong and negative evaluation of these patients is achieved alongside the construction of the two identities as distinct and distant from each other. Some of the key elements of negative evaluations include the explicit mention of lào yīngwén (烙英文), the frequent use of strong language such as the aforementioned “fuck your mother” (幹拎娘), “what freaking pain” (pain啥 小), “penis” (懶叫) , and the self-referential term līngbēi.5 The constructed dialogue (Tannen 2007) that LB uses to exemplify lào yīngwén ‘Ah, my knee is in pain’ (阿我的knee很pain) is especially interesting, as it is rather unusual to switch language to and fro so quickly. It also violates English grammar, with “pain” used as an adjective, although this type of grammatical violation is not uncommon in code-mixing practice in Taiwan. It is not known whether it is a faithful copy of a particular patient’s language practice or a dialogue constructed by LB, but this reported remark has a salient discursive effect: it presents the language of the overseas patient as both disorderly and obnoxious. A further example of how the overseas patients are “otherized” is provided in excerpt 7. This excerpt occurs when LB addresses his fellow physicians and advises them not to bother to be nice to such patients. There is very little likelihood that these patients will come back for further consultation or file a complaint, since their stay in Taiwan is short. Excerpt (7) 我跟你講，這些香蕉人最缺的就是時間 你沒看他一個早上掛了三科門診要看 每一診的跟診護士電話打來打去追他的病歷在哪個診間 你要是真有辦法惹到他 逼他花他的美國時間來寫你一張院長信箱 幹，那是屬於兩千三百萬台灣人的重大勝利阿 ! ! ! (lines 18–20) Let me tell you—what these banana people need the most is time Don’t you see that he/she has three appointments in one morning Nurses in each consultation room call here and there to trace his medical record If you are so good as to irritate him Force him to spend his American time to write a complaint to the Dean Fuck, that would be a victory for the 23 million Taiwanese people!!! In addition to the label “banana people,” which alludes to their white interior despite an Asian appearance, the overseas Taiwanese in this excerpt are constructed as distinct from the local Taiwanese in at least two ways: (1) their time is American time (a common expression in Taiwan Mandarin referring to “time in an unrealistic world,” but here it is meant literally), and (2) wasting their “American time” would be a victory for the 23 million Taiwanese. These discursive strategies effectively depict the overseas patient in question as more “American” than authentically “Taiwanese.”

242  Hsi-Yao Su In sum, through these indexical processes, the overseas Taiwanese patients are constructed as “others,” obviously different from the local Taiwanese. In addition to being distinguished from the local Taiwanese, the overseas patients are further deauthenticated as neither Taiwanese enough nor American enough in that their English code-mixing is not “good English” and that they are seen as unable to afford the medical expenses in the United States (thus exploiting the health insurance in Taiwan). This process of distinction and deauthentication is achieved through multiple strategies, but LB’s explicit mention of lào yīngwén practice and his mocking performance of lào yīngwén certainly play an important role. The post also exemplifies the polycentric nature of language/discourse. The single post itself is oriented toward multiple centers (Blommaert 2010), revealed through different levels of tension emergent in the discourse. The first level of tension concerns medical practice, with LB as a practicing physician and overseas patients as people asking for medical examinations that they cannot even name correctly (see excerpt 5). The second and third levels concern the status of English as an internationally powerful language and Taiwanese as a marker of local identity, respectively. The ideology of English as an internationally powerful language can be further scrutinized from two main perspectives, one more historical and the other more recent, though they co-exist and are deeply interwoven in contemporary Taiwan. The “older” discourse concerns Taiwan’s post-World War II socio-political development and its tight connection with the United States, to which many Taiwanese migrated; the “newer” discourse concerns the ideological connection between English and competitiveness in an increasingly globalized world both at the policy level and the level of everyday discourse. The postwar Taiwan-US relationship is beyond the scope of this study, but it suffices to say that post-war Taiwan served as a cold-war outpost and had close political and economic links to the United States. Partly due to the political and military uncertainty concerning Taiwan’s national identity and security since the mid-twentieth century, and partly due to the status of the United States as a wealthy and powerful country, Taiwanese who migrated to the United States have generally been viewed as more privileged than the general Taiwanese populace, and this privilege can be met with both positive admiration and negative sentiments. In both Wanwan’s commentary and LB’s post discussed earlier, attitudes toward such emigrants (and their descendants) can be spotted. Wanwan’s commentary adopts the term ABC (shorthand for “American-born Chinese”) and seems to make a distinction between the “real” ABCs and the “fake” ones, who only use a few English words and cannot carry on an actual conversation. In this regard, the privilege of the “real” ABCs is not challenged; it is the “fake” ones, who show off their limited English proficiency, that she negatively evaluates. In LB’s post, negative sentiments toward the emigrants who abuse the Taiwanese health care system are apparently strong, and, unlike in Wanwan’s case, the privileged position of these emigrants is explicitly questioned, as LB attributes

The Discourses of lào yīngwén  243 the cause of such abuse to their inability to afford the medical expenses in the United States. LB’s negative stance toward these emigrants reveals a gradual societal change and power shift in Taiwan: as Taiwanese society becomes wealthier and more educated with greater access to English, and as the discourses of Taiwanese identity become more prevalent, the privileged position of the emigrants is no longer unquestionable. Yet living experiences and connections with American society remain a much sought-after resource, as it is the “fake” ABCs or those who are not actually well-off in the United States that are mocked and criticized; the privileged status of the “real” ABCs and those who can live comfortably according to the American standards is not challenged. And, interestingly, it is the former group to which the lào yīngwén label is attached in Wanwan’s and LB’s posts. In contrast to the aforementioned “older” discourse that associates English with American connections and migration in post-war Taiwan, a more recent discourse that connects English with international competitiveness also informs this study. Lin (2014, 2018) explores the ideological construction of English in non-English-speaking East Asian countries, Taiwan being one of them, and identifies prevalent discourses that link English with national development and manpower, and internalization and globalization as revealed through government policy, language education-related discussions, and advertisements. Lee (2012), focusing on attitudes toward English in Taiwan, identifies an ambivalence: on the one hand, her interviewees express the same ideology on the connection between English and market competitiveness as identified in Lin (2014, 2018); on the other hand, English use in daily contexts is often seen as showing off and is negatively evaluated. Lee’s observation, along with the lào yīngwén discourse investigated in this study, reveals resistance to the normative status of English that has been little researched. Yet another center (Blommaert 2010) to which LB’s post is oriented concerns Taiwanese as a marker of local identity. Interestingly, in many of the lào yīngwén discourses, the Taiwanese language or a Taiwanese accent (in English or Mandarin) is often mentioned to create a discursive contrast with lào yīngwén, as excerpt 6 reveals (also see excerpt 8 as another example). The interaction of the ideological construction of English as a global language and that of Taiwanese (or a Taiwanese accent) as a marker of local identity is clearly illustrated by the paragraph in excerpt 6 in which the overseas patients are constructed as English-speaking and demanding. While the demanding image is achieved largely through LB’s portrait of the gap between these patients’ requests for medical procedures and their lack of medical knowledge, part of the negative image also appears to originate from their use of English. Such a connection between English and a sense of superiority is oriented to the status of English as a global, powerful language. However, while LB alludes to this translocal norm of English, he also responds to and subverts the norm by stating that he practices medicine only in Taiwanese. This is clearly an act of distancing himself from the overseas patient and a

244  Hsi-Yao Su call for orientation to a local center where Taiwanese has its symbolic status as a marker of Taiwanese identity. In the previous examples, lào yīngwén is represented as the insertion of single English words in Chinese-based conversations. This contrasts with the next example, a blog posted by surgeon and blogger Lisa Liu 劉宗瑀 (Liu 2015), in which longer stretches of English appear in a context where Chinese is expected. In this comical post, Liu detailed an incident of workplace miscommunication caused by the attending physician’s insistence on the use of English in the operating room when all the staff involved were Taiwanese and when most of the time there were no non-Chinese-speakers around. In this particular incident, a foreign physician was visiting and observing, and the attending physician, Dr. Wan, was excited to have a visitor to showcase the team’s medical and language proficiency. During the operation, he requested the nurse to give him Opsite, a type of wound dressing, in English, while the nurse mistook it as a request in Taiwanese to pick Wan’s nose for him during the operation. Wan’s request was comically portrayed as English spoken with a Taiwanese accent, both by the author’s narration and by the use of Chinese characters as a transliteration of the English pronunciation of Opsite (as opposed to representing it in English). Excerpt (8) ….. 老萬醫師則是超高興有難得的國外觀眾，想到了他引以為傲的接軌國 際，大聲又自豪的轉頭要旁邊流動護士遞東西給他：「普利斯give me O～P site！」 OP site是一種軟薄膜敷料，用來貼在傷口最外側。 護士則是一臉登愣： 「花特？」 老萬不耐：「O～P site! O～P～site!」 護士回： 「re……really?」 老萬甩了器械轉頭： 「歐～批～賽！」有點動氣到最後變成一整個台語腔 都出來…… 只見……護士戴上手套（疑？）、一個箭步上前，把老萬臉上的口罩拔下， 然後，護士手指伸進老萬的鼻孔裡…… 「挖……鼻……屎……」 （台語） 我們全場都驚呆了。 …… 直到老萬的三魂七魄從火星回到地球，默默把護士的手移開、戴上口 罩，離開刀房前交代一句： 「小劉醫師……妳關剩下傷口。」 ….. Dr. Wan was so thrilled to have a foreign audience. Thinking of the emphasis on international standards of which he had been so proud, he loudly and proudly asked the circulating nurse to pass him something. “Pu-li-si (please) give me O~P site!”

The Discourses of lào yīngwén  245 OP site is a type of wound dressing, placed on the outer layer of the wound. The nurse was lost. “Hua-te (what)?” Dr. Wan was impatient. “O~P site! O~P~ site!” Nurse: “re….really?” Dr. Wan waved the instruments and turned his head. “Ou–Pi–Sai!” He was so irritated that his Taiwanese accent emerged. What we saw next was the nurse putting on a glove, moving forward, taking Dr. Wan’s mask off, and inserting her finger into Dr. Wan’s nostril…. “O-Pĩ-Sai” (“Pick the nose” in Taiwanese) We were all in shock. ……. Finally when Dr. Wan regained his composure from the Mars to Earth experience, he silently moved the nurse’s hand away and put on his mask. Before he left the operating room, he said, “Dr. Liu, you close the remaining wounds.” Unlike the previous examples that connect lào yīngwén with code-mixing, in this chapter, lào yīngwén refers to Dr. Wan’s attempt to create an allEnglish work environment when such a practice is not expected. Although this post about miscommunication is mostly framed as a joke rather than a direct comment on the use of English in the Taiwanese workplace, implicit evaluation can still be observed. The title “the (unfortunate) consequence of lào yīngwén” (撂英文的下場) directly associates the use of English with an unpleasant outcome. The content of the post also reveals the author’s disapproving stance toward Wan’s order to use English with Taiwanese staff during a surgical operation. Like the previous examples, scale jumping can also be observed in excerpt 8. Wan’s insistent use of English is geared toward international standards and is oriented to English as the global norm at the translocal scale. Yet his accented English and the miscommunication, as portrayed by Liu, index a local scale. After the embarrassing incident, Wan gave a final order in Mandarin instead of English, another indication of the scale jumping from an aspiration to the international standard to the practical consideration of the local linguistic reality. Across the examples discussed in this section, subversive discourses that challenge the usefulness and authority of English in the Taiwanese context can all be observed, revealing a certain degree of resistance to the normative status of English as a powerful language and as a symbol of market competitiveness.

General Discussion The consistency across the different online posts on the subject of lào yīngwén reveals some common language ideologies. First, the tension between local identity and global/translocal identity/aspiration is a recurrent theme in contemporary Taiwan, and language—or more precisely

246  Hsi-Yao Su metalinguistic and metapragmatic commentaries about language (Silverstein 1976; Verschueren 2000)—has become an important site where these identity-related ideologies are shaped and contested. The three examples examined—Wanwan’s comic strip, LB’s post criticizing overseas Taiwanese, and Liu’s post about an embarrassing workplace miscommunication—all reveal such tension. Several threads of ideologies or beliefs about language can be observed: i English indexes prestige and global identity/aspiration. ii Eagerness to show global identity/aspiration is often seen as incompatible with local identity. iii The overuse of English is thus seen as showing off and distancing oneself from one’s local identity. iv Code-mixed English or locally accented English is not seen as “real” English. v Thus code-mixers or locally accented speakers of English are doubly condemned: they are seen as showing off and distancing themselves from local identity, but the local traces in their language preclude a “truly” global, cosmopolitan identity. vi The contrast between local identity and global aspiration (rather than a “true” cosmopolitan, translocal identity) is often ideologically represented with the contrast between Taiwanese and lào yīngwén, such that a negative evaluation of lào yīngwén practice is often made in conjunction with reference to Taiwanese, as shown in LB’s comment on his practicing of medicine in Taiwanese and Liu’s comical depiction of workplace miscommunication. These observations lead us to the second point: English used to be a resource restricted to the highly educated or Taiwanese emigrants to Englishspeaking countries. As English education becomes more accessible both in the formal school curriculum and in private education, the symbolic role of English has also transformed. Its symbolic status remains but is manifested in quite different ways. Use of single English words may still carry a sense of foreignness and a cosmopolitan image (see Wasserfall’s chapter in this volume for a discussion on conventionalized code-switching in Taiwan). But when such use of single lexical items is not accompanied by fluency in English or is spoken in a nonlocal English accent, speakers of such practice are seen as falling short at both ends: they are lacking in local pride and identity, and thus subject to negative evaluation by the general Taiwanese populace; at the same time, their code-mixing practices are seen as a local (rather than translocal) practice and such speakers are perceived as lacking the necessary resources to be “true” global citizens. In both past and present, the phenomenon remains the same in that only those linguistic forms with restricted access are given the highest level of symbolic salience. The difference is that code-mixing in the past, when English was less widespread, was

The Discourses of lào yīngwén  247 probably enough to index a sense of globalness/translocalness and social superiority. But in the present time, English code-mixing with local traces does not index a high-class status, but rather just an aspiration. Third, one related development of this transformation is that certain forms of English do not necessarily index a global scale, and Taiwanese, a marker of local identity, is not necessarily seen as incompatible with a truly global image. In quite a number of the online posts that I have collected, a sense of localness is signaled by the authors (as exemplified by LB’s dense use of local colloquial language) alongside their negative evaluations of code-mixing as not authentic English. LB, along with others that I have not shown in this chapter, constructs himself as both local and translocal, as both Taiwanese and global citizens.

Conclusion Taking the increasingly widespread written use of lào yīngwén as a starting point, this chapter has explored the subversive reactions to the normative status of English as a global language and as a symbol of workplace competitiveness. The term lào yīngwén does not necessarily convey negative evaluations, but negative evaluations can often be spotted, sometimes blatantly as in LB’s post, sometimes more subtly as in Wanwan’s comic strip and Liu’s post. This chapter explores this little studied aspect of attitude toward English in Taiwan. As mentioned in the introduction, the Taiwanese government has recently re-opened the discussion regarding the possibility to designate English as a second official language, and such a policy, if implemented, may bring Taiwan closer to other post-colonial sinophone territories such as Singapore or Hong Kong. The government effort has met with varied reaction from the public. Some stress the importance of English as a global language, while others focus on practicality and national and local identity, thus closely echoing the competing discourses discussed in this study. In light of this renewed development in language policy, this study hopes to contribute to a general understanding of the various ways in which English is discursively constructed and ideologically represented as playing particular roles and indexing particular identities, personae, or images in various local contexts, and to illuminate the constant interaction between the normative and subversive discourses about English and the local languages that weave our everyday life experiences.

Notes 1 The statistics were issued by the Bureau of Education of the Kaohsiung City government by order of the Ministry of Education (via http://bsb.edu.tw), accessed on February 7, 2017. 2 The Taiwanese-Japanese Dictionary and Taiwan Southern Mǐn Dictionary of Common Expressions represent Taiwanese in slightly different ways. The former mainly adopts the character tradition. The latter attempts both character-only

248  Hsi-Yao Su and Romanization-only systems; thus, examples are given in both forms. The entries and examples quoted in the text follow the two dictionaries’ original representations. 3 Below is an example of the written slang term in news report titles. “Drunkdriving elderly man spoke English (lào yīngwén) and tried to escape. Police calmly responded and netizens cracked up” (酒駕伯撂英文想逃—警淡定回話笑 翻網友; accessed on April 21, 2019 via http://news.ltn.com.tw. 4 It is likely that code-mixing between Mandarin and Taiwanese may also generate negative evaluations in certain contexts, but given its commonness, it is generally well-accepted. 5 Many of the characters adopted here are used as transliterations of the intended Taiwanese phrases. The inherent meanings of the characters are often irrelevant to the meanings of the phrases.

References Blommaert, Jan. 2010. The Sociolinguistics of Globalization. Cambridge: Cambridge University Press. Bucholtz, Mary, and Kira Hall. 2004. “Theorizing Identity in Language and Sexuality Research.” Language in Society 33, no. 4: 501–47. doi:10.1017/s0047404504334020  ———. 2005. “Identity and Interaction: A Sociocultural Linguistic Approach.” Discourse Studies 7, no. 4–5: 584–614. doi:10.1177/1461445605054407 ———. 2010. “Locating Identity in Language.” In Language and Identities, edited by Carmen Llamas and Dominic Watt, 18–28. Edinburgh: Edinburgh University Press. HiNative. 2018. “What Does 撂英文 Mean?” Accessed on January 19, 2020 via https://hinative.com. Lee, Ann W.H. 2012. “Language Ideology of English: Its Relationship with Linguistic Ownership.” MA thesis, National Taiwan Normal University. Lin, Han-Yi. 2014. “The Ideological Construction of English: A Critical Review on the Discourse of English in East Asia.” Pan-Pacific Association of Applied Linguistics 18, no. 1: 219–40. ———. 2018. “The Promotion of English in East Asia at the Turn of the 21st Century: A Politico-Economic and Socio-Cultural Review.” Journal of Pan-Pacific Association of Applied Linguistics 22, no. 1: 1–17. doi:10.25256/paal.22.1.1 Liu, Lisa 劉宗瑀. 2015. “Lào yīngwén de xiàchǎng” 撂英文的下場 [Consequences of lào yīngwén]. Accessed on April 21, 2019 via http://bw.businessweekly.com.tw. MicroGG. 2014. “1.8 Wàn qiáobāo qùnián shàng bànnián huā jiànbǎo qīqiānwàn” 1.8萬僑胞 去年上半年花健保七千萬 [18,000 overseas compatriots spent 70 million on health insurance in the first half of last year]. Accessed on April 21, 2019 via https://www.ptt.cc. MOE (Ministry of Education, Republic of China). 2008. Jiàoyùbù Táiwān Mǐnnányǔ chángyòng cí cídiǎn 教育部台灣閩南語常用詞辭典 [Taiwanese Southern Mǐn dictionary of common expressions]. Accessed on April 21, 2019 at https://twblg. dict.edu.tw/holodict_new. Ochs, Elinor. 1992. “Indexing Gender.” In Rethinking Context: Language as an Interactive Phenomenon, edited by Alessandro Duranti and Charles Goodwin, 335–58. Cambridge: Cambridge University Press. Ogawa Naoyoshi 小川尚義. 1931. Tai-Nichi dai jiten 台日大辭典 [TaiwaneseJapanese dictionary]. Taipei: Taiwan Sōtokufu.

The Discourses of lào yīngwén  249 Pixnet. 2011. “Lào yīngwén (liào yīngwén)” 烙英文 (撂英文). Accessed on January 19, 2020 via https://sobakome.pixnet.net. Schieffelin, Bambi B., Kathryn A. Woolard, and Paul V. Kroskrity, eds. 1998. Language Ideologies: Theory and Practice. Oxford: Oxford University Press. Silverstein, Michael. 1976. “Shifters, Linguistic Categories, and Cultural Description.” In Meaning and Anthropology, edited by Keith H. Basso and Henry A. Selby, 11–56. Albuquerque: The University of New Mexico Press. ———. 2003. “Indexical Order and the Dialectics of Sociolinguistic Life.” Language and Communication 23, no. 3–4: 193–229. doi:10.1016/s0271-5309(03)00013-2 Tannen, Deborah. 2007. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse. Cambridge: Cambridge University Press. doi:10.1017/ CBO9780511618987 Verschueren, Jef. 2000. “Notes on the Role of Metapragmatic Awareness in Language Use.” Pragmatics 10, no. 4: 439–56. doi:10.1515/9783110907377.53 Wanwan 彎彎. 2009. Lào yīngwén 烙英文. Accessed on January 19, 2020 via https:// cwwany.pixnet.net/blog/post/32584159. Whorf, Benjamin. 1956. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf, edited by John B. Carroll. Cambridge, MA: Technology Press of MIT. Woolard, Kathryn A., and Bambi B. Schieffelin. 1994. “Language Ideology.” Annual Review of Anthropology 23, no. 1: 55–82.

12 Conventionalized Code-Switching in Taiwan English Insertions in Taiwan Mandarin Julia Wasserfall Abstract This chapter analyzes the concrete manifestations of English items in Taiwan Mandarin and speakers’ attitudes toward Mandarin-English code-switching (CS). It is pointed out that speakers of different backgrounds use and recognize a number of English items, while submitting them to different degrees of morphological and phonological integration. Many items perceived as widely spread are not systematically integrated into Mandarin speech and can therefore not be considered borrowings. With regard to speakers’ attitudes, the results show that speakers clearly differentiate between English and Chinese in their use of insertions and do so consciously. English words are used as markers of belonging and form an in-group code. Moreover, in informal domains, English is used consciously as a discursive strategy to mitigate negative directness in swearing and refusals as well as for emphasis or marking informality and playfulness.

Introduction As Hsi-Yao Su discusses in Chapter 11 of this volume, English language beliefs and ideologies in Taiwan differ strongly depending on whether the language is used alone or mixed with Mandarin Chinese. She argues that switching between the two languages is often seen as inauthentic, in terms of both local and cosmopolitan identity. Her chapter, however, shows that this phenomenon is so common that it has given rise to the phrase lào yīngwén, meaning the practice of inserting English words into Taiwan Mandarin. These insertions consist mainly of single words, as the cartoon (Figure 11.1) analyzed by Su aptly demonstrates. Complementing Su’s chapter, this chapter focuses on a closer linguistic analysis of language contact patterns between English and Mandarin in Taiwan. English as a global or international language has long since spread far beyond the traditional English-using countries. In the wake of its spread the English language has undergone nativization processes and a pluralization that scholars have studied under the umbrella term “World Englishes,” and the emergence of an eponymous field of research, in the 1980s (Kachru 2012:

Conventionalized Code-Switching in Taiwan  251 1–2; for a discussion of English[es] in the sinophone world, see Meierkord in this volume). With regard to the three broad types of English uses in the sinophone world, i.e. as first, second, or foreign language, Taiwan falls into the latter. In contrast to other sinophone spaces, such as Hong Kong and Singapore, Taiwan has no history of British colonialism and language planning vis-à-vis English has largely been restricted to language-in- education policies (see Chen in this volume for a detailed discussion). However, a good command of the language is a highly valued skill, as several studies have shown (see for example Lin 2012; Lin and Wu 2015; Oladejo 2006). On a personal level, English is mainly seen as a tool for educational or career advancement, as English instruction plays a crucial role in school and college curricula at various levels. An additional and less personal motivation for the learning of English can be explained by its importance as an international and global language. English is seen as a symbol for international competitiveness. Encouraged by the economic and political reasoning behind the implementation of a new English language policy (see Chen 2003: 161, 166–69), the learning of English has become a responsibility toward one’s country at least for some of its learners in Taiwan (Lin 2012: 70). In addition, Liao (2010) suggests there is a new emerging language ideology in the more urbanized north of Taiwan. While Mandarin, promoted by the KMT, was the sole means of educational and career advancement in the decades under martial law (see Klöter in this volume for a discussion of Mandarin language planning in the PRC and ROC), English has recently challenged Mandarin for this prestigious position, and has “become the new capital in Taiwan’s linguistic market, especially in Taipei” (Liao 2010: 133). However, the decentralized implementation of English language planning in Taiwan has led to inequalities in terms of access to language instructions, most notably between urban and rural districts (Chen, this volume). At the same time, a lively and often costly private market of English language teaching has developed in Taiwan, further underpinning the role of English as marker of social and equational distinction (Pan 2014; Su, this volume). With the spread of the English language in Taiwan, contact between Mandarin and English as well as an academic interest in the phenomenon has increased rapidly in the twenty-first century. Most studies, however, concentrate on written forms and/or take a purely functional, motivational, or attitudinal approach to the subject or are situated within classroom communication (see for example Chiu 2012; Hsu 2008; Nall and Nall 2010; Tien and Li 2014). However, English is also found in spontaneous and thus naturally occurring speech in the form of CS. In general, CS includes single words as well as phrases, entire sentences, and even the complete change of the language of interaction. Most switches between English and Mandarin in spoken language in Taiwan are, however, single lexical items or idiomatic expressions.1 Excluding English language teaching contexts, where language choices and CS patterns follow a separate set of rules and obligations, we can observe that in natural speech in Taiwan Mandarin is the matrix

252  Julia Wasserfall language in which English items are embedded. They constitute what Muysken (2000) calls “insertional code-switching” as opposed to alternation and congruent lexicalization. According to Myers-Scotton (1993, 2002) and her Matrix Language Frame Model, the two languages involved in CS acts do not contribute in equal shares to bilingual speech (Myers-Scotton 2002: 15). The matrix language is clearly identifiable as the dominant language into which single words or islands, i.e. larger units, of the embedded language are inserted. Those islands follow the grammatical structure of the embedded language, while the matrix language provides the morphosyntactic structure outside the islands. Single word switches and codified borrowings are, therefore, both integrated into the morphosyntactic frame of the matrix language and no difference can be made on the grammatical level. The only difference is their respective status in the mental lexicon of the speaker, i.e. established borrowings belong to the mental lexicon of both languages (ibid., 153). Other scholars, most notably Poplack and her associates (e.g. Poplack 1980, 2004; Poplack and Meechan 1998), identify single words as borrowings or nonce-borrowings due to their integration into the dominant language (see also Myers-Scotton 2002: 250–63). For Muysken (2000: 60–95) insertions of single words can be formally grouped with borrowings while, in certain circumstances, showing functional similarities to CS (or codemixing in his terminology). Although the debate is ongoing since the 1990s, the question regarding the status of single word insertions as borrowings or code-switches is still discussed among scholars of language contact today (see for example Grimstad 2017; Hadei 2016). Those distinguishing between CS and borrowing have proposed several criteria to distinguish between the two. Poplack and Sankoff (1984: 103–4), for instance, distinguish them by frequency of use, displacement of native language synonyms, grammatical integration, and acceptability by the speakers (see also Matras 2009: 110–14). Grosjean (1995: 268) suggests a phonological determination of his so-called “guest words.” If a foreign word is pronounced as in the original language, he proposes to define it as a switch, while phonologically integrated words should be understood as borrowings. A similar argument can be found in Myers-Scotton (2006: 265). At the same time, Grosjean (1995: 262) acknowledges the problem of unbalanced bilinguals, i.e. bilinguals who are not equally proficient in both languages. Interference might, hence, occur in the form of accent on the phonological level. As bilinguals are hardly fully balanced in their language competences, accent is an undefinable factor for both Poplack (1988: 220) and Riehl (2009: 22). Beyond the search for a universal grammatical framework for CS and/ or borrowing, the distinction between the two might seem futile. Many scholars in fact have proposed a continuum with codified borrowings at one end and single word code-switches at the other (Matras 2009: 110–14; Myers-Scotton 1992), in which code-switches might, but do not necessarily have to, develop into established and even codified borrowings. However, in the context of English-Mandarin language contact in Taiwan the issue takes

Conventionalized Code-Switching in Taiwan  253 on another importance as it results in an almost exclusive insertion of single words, begging the question of an “Englishization” of the lexicon of Taiwan Mandarin or the increasing use of English in everyday speech. The remainder of this chapter, therefore, takes a closer look at these English insertions and the processes they undergo. On the basis of the criteria proposed for a distinction between borrowings and code-switches, a small-scale study analyzes the morphological and phonological adaptation processes. The last part focuses on the speakers themselves, contextualizing this language contact phenomenon in speakers’ everyday lives as well as their attitudes toward the mixing of English and Mandarin in Taiwan.

Data Collection Starting from the continuum between CS and codified borrowings, as described earlier, this chapter analyzes how to locate English insertions in daily Mandarin conversation in Taiwan along the dividing line between codified borrowings at one end and single uses, i.e. code-switches, at the other end. Accordingly, only spoken language use was considered, leaving aside written usages and writing practices. In order to test English insertions in Taiwan against the criteria summarized earlier, several research tools were used to collect data. A variety of examples were collected through field observation and the help of informants. Observations were made during informal meetings of Taiwanese Mandarin speakers of different age groups and professions. Words and phrases were noted down in context, where possible. Informants included Taiwanese, also of different ages, and expatriates living in Taiwan. Information provided by the latter was double-checked with native speakers, to ensure English had not only been chosen in interaction because of the foreign origin of the interlocutor. Although considered to be a critical factor in the use of English insertions, the aspect of CS in intercultural interaction is not further explored in this analysis. A total of 58 words and two fixed expressions from those collected were chosen for a questionnaire survey. Due to the small sample size, the frequency of use and the perceived spread of these items will not be discussed in more detail in this chapter. However, the last section of the questionnaire offered open-ended questions regarding the participants’ use of English and space for further example sentences or frequently used words. Some of these examples are discussed below. A total of 122 valid questionnaires were collected in Taipei and Hsinchu city. In a second step, semi-structured interviews were conducted and recorded with ten questionnaire participants. Interview questions related to participants’ own usage of English insertions, their attitudes toward this language practice, and their opinion on the relation between English proficiency and the use of English in Mandarin. Participants were also asked to provide two example sentences and their translations into Mandarin for the words fu

254  Julia Wasserfall (colloquial for feeling) and hold. In all, 22 samples of fu were recorded from nine different speakers and 25 samples of hold from eight different speakers. The samples were analyzed for tonal assignments and translations were checked for semantic shifts. All interviewees gave their written informed consent to be recorded and personal data were anonymized for the analysis. Data collection was carried out in spring 2013.

Data Analysis Degree of Integration Morphological Integration The degree of morphological integration of single other-language items indicates how similar foreign words act in comparison with native ones. The main focus lies on the combination of native and foreign items as free and bound morphemes. While the former can act as words by themselves, the latter only occur in combination with other morphemes. Analyzing if and how foreign words are integrated into Mandarin can, thus, help to determine if a word is borrowed or switched. This includes both the integration of foreign verbs and nouns into the native system of inflection and the combination of foreign words with native affixes. Chinese words are, however, not inflected, i.e. grammatical aspects such as gender, number, case, or tense are not marked morphologically by suffixation. Even codified borrowings might not, therefore, show any modification. One possible way to determine morphological integration, however, is to investigate if foreign words and native morphemes combine to form compounds. For this purpose, the example of resultative verbs was analyzed in more detail. Resultative verbs consist of a resultative complement (R) which is traditionally understood as a verb and which indicates the resultative attainment of the verb (V) on the right. They are, therefore, often referred to as resultative compounds (Sun 2006: 54), resultative verb compounds (Li and Thompson 1989: 65), or resultative verbs (Wiedenhof 2015: 312). In Taiwan Mandarin, English items that combine with native resultative complements to form compounds are to be understood as integrated into the Chinese aspect marking of verbs. We will, therefore, take a closer look at English verbs inserted into Mandarin. Various verbs, such as hold, check, share, and tag, were found in spoken usage with varying frequency (listed here from most frequent to least frequent). The verb hold was selected for further analysis and interview participants were asked to provide sample sentences for the item; the following examples were obtained:2 1 這個動作要hold好了 zhège dòngzuò yào hold hǎo le (you) have to maintain this movement

Conventionalized Code-Switching in Taiwan  255



256  Julia Wasserfall between high level tone (55) and high falling tone (51). When fu was assigned a high falling tone, most interlocutors paused for a moment or asked to hear the word again, before they understood what was meant by the question. No such reaction occurred when fu was assigned the high level tone. The analysis shows that fu was assigned a high level tone (55) by all speakers. No deviating or unclear tonal assignment was found in the collected data (see Figure 12.1 for the tonal contour of all female participants). The tonal assignment of hold was not as clear (see Figure 12.2). Out of the 25 samples, the pitch contour of hold could not be clearly identified as one of the four tones in ten cases. For example, a rising and falling contour was assigned in two cases, which does not accord with any single tonal contour in Mandarin. Other such deviating contours were a level-falling and a falling-level contour. One possible explanation for these contours could be a re-analyzation of hold into two syllables due to phonotactic constraints. In general, the Chinese inventory of syllables is quite limited (DeFrancis 1986: 42). Moreover, Mandarin does not allow for consonant clusters or consonants in the final position of a syllable. The two nasals [n] (Pīnyīn -n) and [ŋ] (Pīnyīn -ng) are the only exceptions.4 These constraints significantly influence the perception and/or the production of the original features in loanwords in Mandarin (see Dong 2012; Kim 2019; Miao 2005). The English verb hold violates both the consonant cluster and syllable-final constraints. Consonant clusters C1-C2 in both English and German loanword adaptation in Mandarin were found to be most likely resolved by C1 deletion and C2 resyllabification through vowel insertion (Miao 2005: 106). However, both /l/ in C1 position and plosives in C2 position are more likely to be preserved by vowel insertion than to undergo deletion (ibid., 112). According to Kim (2014: 234), in C1-C2 clusters in final position the liquids in C1 position are more likely to be deleted while the non-liquid C2 is more likely to be

Figure 12.1 Tonal contour of fu by female participants.

Conventionalized Code-Switching in Taiwan  257

Figure 12.2 Tonal contour of hold.

preserved. It can therefore be argued that the monosyllable hold has been reanalyzed as a disyllabic word through final vowel epenthesis, explaining tone contours that are not permissible in monosyllabic words. Moreover, in 11 out of 25 samples, the pitch contour suggests a falling tone. Although in most cases the contour dropped slightly less steeply, this can be explained by optional tone sandhi phenomena in speech in “conversational tempo” where the fourth tone (51) is realized as high to middle (53) if followed by a second fourth tone (Wiedenhof 2015: 25). As hold was mainly recorded in combination with the complement -zhù and in natural speech, it can still be assumed that hold was assigned the fourth (53) tone in Mandarin. In addition, three examples of hold were found carrying a (high) level tone. Only a single instance of a dipping contour was found, in which the rise at the end was more marked than the initial fall. This sample, however, was uttered as repetition and can very likely, therefore, be understood as rising intonation, indicating a question. Therefore, in contrast to fu, no consistent pattern was found in the tone assignment of the English insertion hold. Further Adaptation Processes In addition to the aforementioned processes of integration, i.e. verb compounding and assignment of tones, other mechanisms for adapting English words in Mandarin speech can be observed. For example, some items are mainly used in their truncated form. Words such as con-call ‘conference-call’, fu ‘feeling’, fun ‘funky’,5 po ‘post’, and press-con ‘press conference’ were recorded in their truncated form only. In all cases the original is, in one way or another, incompatible with the phonotactics of Chinese or the preferred word length.

258  Julia Wasserfall Although there are words consisting of three, four, or even more syllables, the largest part of the lexicon of modern Mandarin is mono- and disyllabic (Schindelin 2017a). In addition, corpus studies have shown that monosyllabic words occur most frequently, followed by disyllabic ones. Words of greater length are, with few exceptions, among the least frequently occurring (Schindelin 2017b: 582–83). A corpus study of preferred word length in noun-noun compounds has, moreover, shown that among these, the disyllabic form is by far the most common (Duanmu 2012). Truncation of polysyllabic words is, thus, quite common in Mandarin. The English noun-noun compounds press-conference and conference-call both include a monosyllabic element in addition to the trisyllabic conference, creating a 3 + 1 and 1 + 3 syllable structure. Just as with Chinese two-word four-character expressions, conference is truncated to con in order to form the disyllabic press-con and concall. Truncation of English items, however, was also found in cases where the original words did conform to word length preferences. The truncation of funky, feel, and post can be partly explained by phonotactic constraints in Mandarin. As explained earlier, consonant clusters in general are not permissible and are resolved by vowel epenthesis and/or the deletion of one or both consonants in word final position. In the case of po as the truncated form of post, both consonants in the final consonant cluster have been deleted. Miao’s (2005: 108) corpus study found that the deletion of the entire consonant cluster in final position only occurs rarely in the adaptation of English source words. This phenomenon is, however, more widely spread in Taiwan Mandarin than in Pǔtōnghuà, or PRC Mandarin (Yip 2006: 955–56).6 The truncation of funky [ˈfʌŋki] to fun [fʌn], on the other hand, can be explained by the deletion of a syllable that does not accord with phonotactic constraints. Although the second syllable [ki] is not permissible in Mandarin, it is, nevertheless, often preserved in loanword adaptation. In order to satisfy the constraint of using only permissible syllables, faithfulness to the source word has to be compromised. Although the English vowel [i] and the plosive /k/ can be mapped faithfully into Chinese (Kim 2019: 61–64; Lin 2008: 372), their combination is not permissible. English [k] is therefore adapted through palatalization as Mandarin [ʨ], less frequently [ʨh] (Hui and Oh 2015: 412–13), while the vowel is faithfully retained. At the same time, the nasal [ŋ] is converted to [n] in final position. This is in accordance with the interchangeability of both nasals in final position in the adaptation of loanwords into Mandarin (Hsieh et al. 2009; Miao 2005: 70–71). It is also possible that the conversion of [ŋ] to [n] is semantically motivated as Mandarin [fʌn] is homophonous with the English word fun, indicating a playfulness in adapting the word. The use of the English spelling fun by the Taiwanese respondent is further indicative of a conscious conversion of both nasals. Finally, the English word feeling [ˈfiː.lɪŋ] is especially interesting. Its disyllabic form accords with the preferred word length in Mandarin; yet it is truncated to fu in Taiwan Mandarin, pronounced like the English word few

Conventionalized Code-Switching in Taiwan  259 [ˈfjuː]. Again, the final syllable is dropped while the first is retained. Interestingly, the second syllable [lɪŋ] is permissible in Mandarin and its retention therefore violates neither preferred word length nor phonotactic constraints. In the truncated form fu [ˈfjuː], on the other hand, the vowel is changed and the approximant [j] is inserted, resulting in a final that is not permissible in Mandarin.7 The diphthong [iu] can however be found in final position in Taiwanese.8 In this function, more precisely in the form of [k h iu˧] ‘chewy’, it has been borrowed into Taiwan Mandarin and is frequently seen in writing in the form of the Latin letter Q (on the use of Q in Taiwan and Hong Kong see Mair 2010). In combination with the initial /f/, however, the syllable violates both Taiwan Mandarin and Taiwanese phonotactic rules. It is therefore unlikely that the truncation of feeling results from phonological concerns.9 Another “nativization” process has been noted by Hsu (2008) in regard to written English mixing in Taiwan. In her corpus of advertisement texts, some English items have undergone word-class change through insertion into Mandarin. Similar patterns were found in the use of English insertions in spoken Mandarin. Examples include the words meeting, man, line, and lag. In example 3 the noun meeting derived from the verb to meet is used in its original function. In example 4, on the other hand, the verb to meet is used in the gerund form meeting as a verb. The marker -ing is retained without the need for indicating a progressive form in either Chinese or English. 3 我們今天有一個 meeting wǒmen jīntiān you yī ge meeting we have a meeting today 4 我要跟他 meeting wǒ yào gēn tā meeting I will/ have to meet him Examples 3 and 4 suggest that the noun meeting was introduced first, and that it was reanalyzed as a verb. The use of the gerund instead of the bare form of the verb also suggests a lexicalization of the item meeting as a verb. All other verbs are inserted in their uninflected form (e.g. hold, call). However, in written Mandarin there are cases in which the progressive marker -ing is used in combination with a Chinese verb, e.g. 流浪 ing liúlàng-ing, ‘roaming/ wandering’ or 準備 ing zhǔnbèi-ing ‘preparing’. This phenomenon seems to be restricted to written usage only. Other examples of English nouns which have undergone word-class change are listed in examples 5–7. 5 他/她很 man tā hen man s/he is very manly 6 line 我 line wǒ message me via line

260  Julia Wasserfall

In example 5 the English noun man is used as an adjective in the meaning of ‘manly, masculine’ and is frequently used in this manner, while no use in the nominal form was observed. The item has thus most likely undergone a word-class change in its use in Taiwan Mandarin speech. A second explanation for this usage could be a truncation of the English adjective manly analogous to the earlier examples of fun(ky) and po(st). However, since manly is disyllabic and the second syllable [li] is permissible in Mandarin, no phonotactic rules or preferences for word length would have been infringed by inserting the original form. The English noun line in examples 6 and 7, a proper name for a messaging app, is used as a verb. In example 6 (lit. ‘line me’) the noun is used directly as a verb, whereas example 7 shows the use of line as a main verb in combination with gěi ‘give’ indicating the beneficiary/recipient of the action of the main verb. In both cases, line takes on the meaning of ‘sending a message via the app line’ or ‘texting somebody via the app line’. The use of a proper name as a verb is not restricted to line or Taiwan Mandarin and can be observed frequently in the field of emerging technologies. Other examples involving applications or computer programs are the popular online search engine Google, as in to google something which can also be found in German (etwas googeln) and Mandarin (google or gu) as well as the computer program Photoshop as in to photoshop a picture ‘modify a picture with Photoshop’.

Semantic Change During the interviews, participants were asked not only to give example sentences using the words fu and hold but also to rephrase or translate them into Mandarin only. This exercise was used to check how readily available Mandarin equivalents were and to trigger comments about the use of Mandarin equivalents. Fu was generally translated as gǎnjúe (感覺) ‘feeling’ and once as qìfēn (氣氛) ‘atmosphere’ or ‘mood’. If used with a modifier such as Fàguó ‘France’ in hen yǒu Fàguó de fu (很有法國的fu), it usually translates as ‘atmosphere’ or ‘feeling’, i.e. ‘to have a French feeling’ or more colloquial as ‘vibe’, i.e. ‘to have a French vibe’. In these cases, fu can be translated as gǎnjúe (感覺) ‘feeling’ (很有法國的感覺 ‘to have a French feeling’). Used alone, as in yǒu fu (有fu ‘to have fu’) the translation into either English or Chinese is not as straightforward, as participant 4 (P4) explained: AUTHOR: What does it mean translated into Chinese? P4: Just … translated into Chinese … if you want to be precise … if you want

to say it appropriately… to have fu (yǒu fu 有 fu)… it’s … to have feeling (有感覺), I’d say. It’s “this music sounds very…,” because if you say it

Conventionalized Code-Switching in Taiwan  261 like that in Chinese, it doesn’t seem right, just doesn’t really convey [the meaning]. So, maybe fu is a little bit cheeky, just jesting, if you say “Hey, this music sounds like it has a lot of feeling” (yīnyuè tīng qilai hen yǒu gǎnjué 音樂廳起來很有感覺) it just seems wrong. […] AUTHOR: Is it difficult to translate into Chinese? P4: It depends on your context, if you say, let’s say… for example, like I just said “[this] music has a lot of fu” and “the fu is not the same” (fu bùtóng fu 不同)… “[this] music has a lot of fu” is “feels like… like it has a lot of feeling,” … like I just said “[this] music sounds like it has a lot of feeling” this translation is a bit odd, but if you say “the fu is not the same” and translate it to “this feeling is not the same” (gǎnjué bù tóng 感覺很不同) this way it’s right, the flavor of it, I think, is right, so the situation has to … one cannot say difficult … difficult or not. In this passage, P4 elaborates on how to best render different uses of fu in Mandarin only, discussing examples where fu is not modified by any adjunct. Not only is he generally hesitant in searching for an adequate translation, he also comments on the different contexts in which fu is used and how these affect the adequacy of translating it into Chinese. In his example of describing music, the Chinese word gǎnjúe ‘feeling’ is perceived as not fully expressing the meaning of fu. To (not) have fu seems to be difficult to translate directly as the feeling or atmosphere it is used to describe is rather fuzzy in its meaning. Fu is usually used in a positive sense, i.e. ‘creating a certain kind of atmosphere’, ‘being atmospheric’. The lack of fu was described as ‘not well’, ‘out of sorts’ (不爽), an equally vague description of something being amiss. 8 今天下雨了, 滿有 fu jīntiān xiàyǔ le, mǎn yǒu fu It is raining today. It’s quite atmospheric. The use of fu is furthermore perceived as less serious or carrying a lighter tone, described as ‘jesting’ by P4 or referred to as hǎowán 好玩 ‘playful’, ‘amusing’ by participant 5. In some contexts, the use of fu instead of gǎnjúe 感覺 ‘feeling’ seems to be the appropriate choice as participant 6 (P6) explains: For example, “this way it doesn’t feel right” (méiyǒu gǎnjué 沒有感覺, lit. ‘doesn’t have any feeling’), the second way to say it is “this doesn’t have any fu.” AUTHOR: So, you’d also say it this way? P6: I also would, because to say it this way it’s … shorter, it’s just, for example, sometimes it’s more smooth, not really smooth, a bit more comfortable this way, saying “this has no feeling” (méiyǒu gǎnjué 沒有 感覺) feels like, … a bit cumbersome, … just seems a bit complicated, long-winded, like that. P6:

262  Julia Wasserfall Trying to explain why the use of fu is preferred over gǎnjúe ‘feeling’, P6 refers to both language economy (“it’s shorter”) and the tone of the sentence. Using the Chinese equivalent seems to render the sentence unnatural, most probably more formal and is therefore perceived as inappropriate in certain contexts and situations. Participant 8 (P8) explained it as follows: Can you translate fu into Chinese? er… yes, well, in fact we often, well in fact, it’s to say “feeling,” “a great feeling,” but if you use Chinese like this to say “feeling” (gǎnjúe) it doesn’t feel adequate, it doesn’t really feel right (fu bùgòu hǎo fu bùgòu duì fu 不夠好 fu 不夠對), just doesn’t describe it adequately enough, so we often substitute “feeling” with “fu” to describe it more adequately.

AUTHOR: P8: OK…

In this passage, participant 8 describes the use of gǎnjúe ‘feeling’ as not adequate, because using the Chinese equivalent of fu “doesn’t feel right.” The substitution of fu for gǎnjúe ‘feeling’ seems, therefore, to be due less to differences on the semantic level than to new pragmatic meanings fu has acquired and which are not entirely congruent with its Chinese equivalents. The translation of hold did not cause participants to express similar difficulties. It was used and translated either in its literal sense of ‘to grab’ (zhuāzhù 抓住, example 2) or in the sense of ‘to maintain’ (wéichí 維持 in example 1 and chēng xiàqù 撐下去) as well as in the figurative meaning of ‘to control’ (kòngzhì 控制, example 2b). The latter seems to have been added in the process of inserting it into Mandarin, acquiring an additional semantic meaning. Other words that have undergone semantic change through their insertion into Taiwan Mandarin include high, low, and local. While the notion of “high” and “low” has been reduced to one of their possible original meanings, local has acquired new connotations. High is mostly used in the sense of ‘excited’ or ‘happy’, without, however, necessarily referring to a state of intoxication from drink or drugs. In this sense it is closest to the original meaning in combination with mood or spirits as in to be in high spirits. Its use in Mandarin with other meanings such as ‘extending far above the ground’ was not observed. In addition, no such usage was reported in the respective part of the questionnaire. It seems therefore that the English word high has been semantically narrowed in the sense of ‘excited’, while this meaning has been broadened to include excitement not connected with intoxication.10 Similarly, the meaning of low has been narrowed through its insertion into Mandarin. It is used in the sense of ‘coarse’, ‘vulgar’, or more generally ‘of low level’ usually morally or educationally. One participant in the questionnaire reanalyzed low as a truncated form of local, equating the latter with coarseness and vulgarity. Inserted into Mandarin, the adjective local usually describes something or someone as typically Taiwanese, mainly referring to southern Taiwan, and is often linked to local languages, especially Taiwanese. It stands in contrast to an international and metropolitan

Conventionalized Code-Switching in Taiwan  263 attitude and carries connotations of being rural and/or crude and is thus related to the concept of táikè (台客, lit. ‘Taiwanese visitor’), a frequently used and sometimes pejorative term for someone whose lifestyle, linguistic and social behavior is perceived as distinctly local (Su 2011).11 Despite the overlap of negative meanings, local is not necessarily pejorative. In a more positive sense, local also touches upon the notion of běntǔ (本土, lit. ‘native soil’) associated with authenticity and Taiwanese nationalism (ibid., 300).12 Therefore, local in the sense of ‘locally Taiwanese’ when inserted into Taiwan Mandarin has both positive and pejorative overtones depending on the speakers’ attitude and positioning toward these two different meanings of localism.

Speaker’s Attitudes While some words have thus acquired new semantic meanings through their insertion into Mandarin, they are largely considered as belonging to the English rather than the Mandarin lexicon. When asked directly about the correlation between English proficiency and the use of English insertions, participants were largely ambivalent. There was a general distinction between widely spread words, or simple terms, and other English items. Most participants agreed that some English words are so widely used that they are understood by almost everyone, suggesting that they can be considered uncodified borrowings. Participant 6, for example, stated that he did not use English insertions and when questioned about his own use of fu, however, he argued that since it had undergone phonological change and was already widely used in Taiwan, it could be regarded as part of the colloquial Taiwan Mandarin lexicon. Participant 4 further stated that interlocutors only need to understand the meaning of the English word inserted into Mandarin, which is then unrelated to overall English proficiency. Those items which are more widely used are of course easily learned even by those with no English proficiency through the Chinese context in which they are used. At the same time, only two participants stated that they unwittingly used English insertions with interlocutors with no knowledge of English. All others stated that they made a conscious effort to avoid English items depending on the other’s degree of English proficiency. This suggests that speakers are conscious of the English nature of these words. The status as a foreign language item in their mental lexicon is further supported by their explanations of the differences using words in English rather than Chinese as well as by their attitudes toward frequent use of these words. When talking about the differences between swearing in English and Chinese, participant 2 explained that using a foreign language instead of one’s mother tongue mollifies the harshness of the words, and thus makes the interlocutor or listener less uncomfortable.13 This phenomenon was also observed for other language pairs. Dewaele (2004), for example, found that multilinguals swear more frequently in their L1, unless they use a mitigating

264  Julia Wasserfall strategy, in which case they choose any of their non-dominant languages.14 In a similar way participant 3 described why the English word no is used in Taiwan Mandarin: P3:

I think “no” is more common, because if you say no it seems … if you use Chinese you’d say it like this “don’t want” (bù yào 不要) it seems … just … not very polite, very courteous, so if you use English and softly say no, it’s probably more … won’t hurt so much.

Just as with swearing, using the English word no helps both parties to mitigate a face-threatening rejection or refusal. Participant 5, on the other hand, stressed that this lightness of tone renders the meaning less serious and less formal (see above), restricting the use of English insertions to casual situations: P5: “I am so sad today,” this is just on purpose, because we usually don’t say

it this way, but just … just to switch one word, this way the other person will listen more attentively … but if I am really upset, if I am really upset today, I wouldn’t say “I am so sad today” I’d say “I don’t feel well” (xīnqíng hen chà 心情很差), just use detailed words that can express things in more detail, but if it’s just among friends chatting, just small things, and for fun, then it’s OK. Here, again, the playfulness of English insertions is foregrounded. Therefore, their informal use is in marked contrast to the seriousness of the utterance, marking it as less severe, i.e. mitigating the original meaning of the sentence. Participant 8 concluded that since English implies informality, it is better avoided in certain situations such as the more formal domain of work: P8:

Er … I most often use them [English words] when hanging out with friends, because in a work environment you’d have to choose your words carefully, so [indistinct] you’d better avoid those … er … casual words [laugh] [indistinct] I suddenly can’t think of the Chinese [word] [laugh].

In contrast to the concerns of participant 8 about the casual tone that English insertions convey, several participants related them with the work or university domain in particular. Participants 1, 2, 3, 4, 5 all used English insertions in their professional life. Examples included such words as e-mail, forward (as in to forward an e-mail), meeting, briefing, seminar, paper, and journal as well as specialized terms in the field of biotechnology, including among others DNA, enzyme, differentiation. All of the examples given have Chinese equivalents, i.e. standardized translations. The use of English in these domains is often referred to as a habit among the immediate speech community of the office or university. It thus functions as an in-group code,

Conventionalized Code-Switching in Taiwan  265 marking the belonging of its members through the use of certain English words as participants 3 and 5 explain: I think, in a work environment some colleagues just have the habit of using some [English] words, and then because you, you are in the same working environment, so you use the same words as them, it feels like blending in. P5: If in a working environment, it’s because [they’re] fixed [terms], people in the office will all say it this way, so you can only communicate by saying it this way, like saying “we have to go to a briefing now,” it’s “we have to go to a briefing” ( jiǎnbào 簡報), it’s just already become a habit with everyone to use this word to express this thing, so we all necessarily say it this way, no one would say “so let’s go to the briefing ( jiǎnbào 簡報) now.” AUTHOR: So if you’d say it in Chinese people would think “what is s/he talking about”? P5: No not that, but they would think … oh you’re new, you don’t know we say it this way here, only then you’re one of this culture, that is office culture, it’s taught to you this way. P3:

Although Chinese equivalents are readily available, habit or “office culture” (公司文化) is named as reason for using English. The comments of participant 5 suggest that there seems to be a certain number of fixed English terms that are inserted into Mandarin. The examples of other participants show that these fixed expressions vary to some degree, as they include both words related to the domain of office work or academia in general (e.g. e-mail, meeting, paper) and specialized terms depending on the profession itself. Participant 6, moreover, expressed the belief that the use of English is especially widespread among white-collar workers and that it increases with the professional status of the speaker. There is, of course, need for more detailed research in order to verify such claims. Previous studies (Chen 2003: 136; Lin 2012; Lin and Wu 2015) have, however, shown that English in Taiwan is closely related to professionalism, high levels of educational attainment, as well as social status.15 Under certain circumstances, the mixing of English into otherwise Mandarin speech is, therefore, negatively connoted (see Su’s analysis of lào yīngwén in this volume). All but three participants overtly rejected the all too frequent use of English insertions in Mandarin. Participants 2 and 3 were more lenient toward the habit, while participant 5 did not comment on the subject. Participant 2 explained that insertions were frequently used with the intention of making the conversation more casual or to demonstrate open-mindedness toward foreign language influences, while participant 3 said she did not mind other people using English in that way herself, but suggested that others might perceive it as pretentious. This was, in fact, the most common assessment of the overly frequent mixing

266  Julia Wasserfall of English and Mandarin. Participants characterized this language use as “showing off,” using various descriptions, as in the following three examples: P4:

P7:

P6: P7: P6: P8:

So if you insert some words, I think it’s OK but if you use, for example, English half the time [laugh] then I think you just want to show off, just show off (xuànyào 炫耀), wait a second show off [laugh], right, just feels like showing off (xuànyào) to me. I don’t mind, it’s very Taiwanese, sometimes when you speak English, it’s just, just a short bit, like just now “oh this has fu,” I think it’s quite adorable (kěài 可愛), but if it’s too much … pretentious (Taiwanese [ke˧paɪ˥]) [laugh] just, if it’s too much I don’t find it amusing (bù hǎowán 不好玩) … showing off (màinòng 賣弄) Er … you just have to get used to it [mixing English and Mandarin], but if it’s too deliberate then I just feel like this person is showing off, too arrogant, wants to display his/her English too much.

The Chinese expression xuànyào 炫耀 was used most frequently (four participants) followed by the English word show off (two participants). Other descriptions included ‘pretentious’ (Taiwanese [ke˧paɪ˥]), ‘showing off’ (賣 弄), and ‘arrogant’ (驕傲). Although field observations and anecdotal evidence suggest that more frequent use of English is common among Taiwanese Mandarin speakers when talking to a non-native speaker, participants in the study were much more conscious of using English due to the rather formal situation of being interviewed and recorded as well as the subject.16 It is therefore quite interesting that both participants 4 and 8 used the English insertion show off to describe their negative impressions. Participant 4 only realized the switch after the fact, stopped, laughed, and repeated the English word, while participant 8 consciously and playfully over- emphasized the word. The use of English to describe the negative connotations of using (too much) of the language is, thus, quite emblematic of the ambivalence about English in Taiwan. Moreover, four participants (3, 4, 6, 9) rejected the frequent mixing of English and Mandarin, directly stating their preference for either monolingual Mandarin or English use, suggesting that it is not necessarily English itself but the mixing of the two languages that is perceived in a negative way. However, multilingual language use such as CS and borrowing between Mandarin and Taiwanese (and to a lesser degree Hakka) is a widespread phenomenon in Taiwan. The use of CS in political campaigns (Wei 2003) and the popular media (Chiu 2012; Hsu 2008; Lin 2017) suggests that the phenomenon is not only widespread in formal and informal domains of Taiwan’s society but also positively perceived in general. The latter can be surmised from the fact that language use in both the contexts of political campaigning and popular media is designed to attract voters or an audience. Speakers would therefore be expected to shun language use that is not positively connoted with large parts of the society.

Conventionalized Code-Switching in Taiwan  267

Conclusion This small-scale study has illustrated the distinct patterns of EnglishMandarin mixing in Taiwan. Although English is learned and used as foreign language, the use of English insertions is common in natural speech. In addition to the 56 words suggested in the questionnaire, 91 out of 122 participants suggested 224 different additional words, phrases, and sentences which they themselves used or have heard used in Taiwan. Despite the apparent diversity of words, many words were found to be recurrent in the language use of the participants. This suggests that most English insertions in Taiwan Mandarin should be classified as established borrowings. It is supported by the adaptation processes that English words undergo. The more detailed analysis of fu and hold, however, shows that these processes are gradual. The English word feeling, for example, shows truncation to fu, while at the same time undergoing vowel change. Moreover, tonal analysis of naturally occurring samples by different speakers provides evidence for regular tonal assignment of a high level tone. Due to the lack of gender, case, and number marking in Mandarin, morphological integration of English nouns is very hard to measure. The verb hold was found to be morphologically integrated through compounding with the resultative complement -zhù 住 (or by combination with the same to form a compound verb, depending on the definition of such combinations). Despite this morphological integration, hold was not integrated phonologically by regular tonal assignment. Other phonological changes such as reanalysis as disyllabic or dropping of the consonant cluster at the end can be explained by the influence of Mandarin as the native language of the participants, manifested as non-native accent in English insertions. Comparing the meanings of these two words shows that semantic specialization has taken place in the case of fu. This word was found to be used to describe either an atmosphere in particular, usually with a modifier, or situations, actions, locations as “atmospheric” or as “invoking an emotional response” mainly in a positive sense. It was not observed in the meaning of sensory perceptions (i.e. ‘a sense of touch’). Participants, moreover, commented on the differences of fu and gǎnjué (‘feeling, to feel’), used as closest Mandarin equivalent. Hold, on the other hand, was found with the additional meaning of ‘to control’ when inserted into Mandarin, while the meanings of ‘to maintain’ and ‘to grab’ were also observed in the samples provided by the participants. More significant, however, are the status in the mental lexicon of Taiwanese Mandarin speakers and their attitudes toward the mixing of English and Mandarin. While participants were ambivalent about the relation between English proficiency and English insertions, most used or avoided them consciously, depending either on the formality of the situation or the age and/or knowledge of English of their interlocutors. English was associated both with informality and professionalism. The latter was only expressed indirectly through comments on mixing habits in professional and academic domains. The use of English in

268  Julia Wasserfall these contexts was described as habitual and conventionalized, while also being an expression of membership in a certain speech community, or community of practice. Explanations of this phenomenon suggest that besides a set of English words related to office culture in general (i.e. email, meeting), which are spread across smaller communities of practice or even professions, another set of specialized terms is used according to the profession of the community. These fixed English terms thus behave like established borrowings within the immediate community of practice while remaining largely unknown to the wider speech community. They are used as markers of membership and form an in-group code. Whereas most of these insertions have codified translations, English is the preferred language for referring to the concepts in question. Moreover, in informal domains, English is used consciously as a discursive strategy to mitigate negative directness in swearing and refusals as well as for emphasis or marking informality and playfulness, all well acknowledged uses and functions of conversational CS (see for example Appel and Muysken 1987; Gumperz 1982). These findings confirm Muysken’s (2000: 69) claim that lone other word insertions can be formally grouped as borrowings, while functionally behaving like CS under certain circumstances. As English insertions in professional domains seem to be fixed according to the community of practice, they are, thus, listed (ibid., 71). Although Poplack et al. (1988) classify such single word insertions as borrowings, Muysken (2000: 71) calls this language contact phenomenon conventionalized CS (see also Lantto 2015 for a discussion of Spanish-Basque conventionalized CS). All in all, the results of the analysis of English insertions in Taiwan Mandarin confirm the view of CS and borrowing as a continuum along several dimensions (Matras 2009: 110–14), rather than a clearly distinguishable language contact phenomenon. Given the variable integration of such lone other-language items and their listedness or their accepted and recurrent use within smaller communities of practice, their status in the mental lexicon of Taiwan Mandarin speakers, and the conscious use of English words as discursive strategies, this mixing pattern can be most plausibly called conventionalized CS, while individual items might well be placed closer to either end of the borrowing and CS continuum.

Notes 1 Nevertheless, phrases and complete sentences have been observed in individual cases and classroom interactions (see also Chen 2003; Tien and Li 2014). 2 The sentences marked with “a” and “b” are the rephrased examples provided by the participants. 3 The tone scales were first used by Chao Yuan Ren in the 1930s (see Wiedenhof 2015: 15). 4 Another exception is the rhotacization with the semivowel -r [ɻ] or érhuàyīn 兒 化音 in Mandarin. The addition of this suffix is typical of northern Mandarin speakers and mainly restricted to speech, i.e. it remains largely uncodified in dictionaries. It is, however, very rare in Taiwan Mandarin (Wiedenhof 2015: 48–53).

Conventionalized Code-Switching in Taiwan  269

270  Julia Wasserfall Approaches, edited by José A. Álvarez Valencia et al., 105–21. Charlotte, NC: Information Age Publishing. Chen, Su-Chiao. 2003. The Spread of English in Taiwan: Changing Uses and Shifting Attitudes. Taipei: Crane. Chiu, Miao-Chin. 2012. “Code-Switching and Identity Constructions in Taiwan TV Commercials.” Monumenta Taiwanica 台灣學誌5: 27–49. Coulmas, Florian. 2005. Sociolinguistics: The Study of Speakers’ Choices. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511815522 DeFrancis, John. 1986. The Chinese Language: Fact and Fantasy. Honolulu: University of Hawai’i Press. Deterding, David. 2015. “Segmentals.” In The Handbook of English Pronunciation. Vol. 26, edited by Marnie Reed and John M. Levis, 67–84. Malden, MA: Wiley Blackwell. doi:10.1002/9781118346952.ch4 Dewaele, Jean-Marc. 2004. “Blistering Barnacles! What Language Do Multilinguals Swear in?!” Estudios de Sociolingüística 5, no. 1: 83–105. doi:10.1558/sols.v5i1.83 ———. 2016. “Thirty Shades of Offensiveness: L1 and LX English Users’ Understanding, Perception and Self-Reported Use of Negative Emotion-Laden Words.” Journal of Pragmatics 94: 112–27. doi:10.1016/j.pragma.2016.01.009 Dollinger, Stefan. 2015. The Written Questionnaire in Social Dialectology. Amsterdam: John Benjamins. doi:10.1075/impact.40 Dong, Xiaoli. 2012. What Borrowing Buys Us: A Study of Mandarin Chinese Loanword Phonology. Utrecht: LOT. Duanmu, San. 2012. “Word-Length Preferences in Chinese: A Corpus Study.” Journal of East Asian Linguistics 22, no. 1: 89–114. doi:10.1007/s10831-011-9087-y Edwards, John. 2006. “Foundations of Bilingualism.” In The Handbook of Bilingualism, edited by Tej K. Bhatia, 7–31. Malden, MA: Wiley Blackwell. doi:10.1002/9780470756997.ch1 Grimstad, Maren. 2017. “The Code-Switching/Borrowing Debate: Evidence from English-Origin Verbs in American Norwegian.” Lingue e Linguaggio 16: 3–34. doi:10.1418/86999 Grosjean, François. 1995. “A Psycholinguistic Approach to Code-Switching: The Recognition of Guest Words by Bilinguals.” In One Speaker, Two Languages: Cross-Disciplinary Perspectives on Code-Switching, edited by Lesley Milroy and Pieter Muysken, 259–75. Cambridge: Cambridge University Press. doi:10.1017/ cbo9780511620867.012 Gumperz, John J. 1982. “Conversational Code Switching.” In Discourse Strategies, edited by John J. Gumperz, 59–99. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511611834.006 Hadei, Marzieh. 2016. “Single Word Insertions as Code-Switching or Established Borrowing?” International Journal of Linguistics 8, no. 1: 14–24. doi:10.5296/ijl. v8i1.8899 Hamers, Josiane F., and Michel Blanc. 2000. Bilinguality and Bilingualism. 2nd ed. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511605796 Hsieh Feng-fan 謝豐帆. 2014. “Jiècé yīnxìxué yǔ Hànyǔ jiècí yánjiū” 借词音系学与 汉语借词研究 [Loanword phonology: A Chinese perspective]. Dāngdài yǔyánxué 16, no. 3: 358–71. ———. 2017. “Loanword Phonology.” In Encyclopedia of Chinese Language and Linguistics, Vol. 2, edited by Rint Sybesma et al., 624–29. Leiden: Brill.

Conventionalized Code-Switching in Taiwan  271 ———, Michael Kenstowicz, and Xiaomin Mou. 2009. “Mandarin Adaptations of Coda Nasals in English Loanwords.” In Loan Phonology, edited by Andrea Calabrese and W. L. Wetzels, 131–54. Amsterdam: John Benjamins. doi:10.1075/ cilt.307.05hsi Hsu, Jia-Ling. 2008. “Glocalization and English Mixing in Advertising in Taiwan: Its Discourse Domains, Linguistic Patterns, Cultural Constraints, Localized Creativity, and Socio-Psychological Effects.” Journal of Creative Communications 3, no. 2: 155–83. doi:10.1177/097325860800300203 Hui, Yang, and Mira Oh. 2015. “Adaptation of English Stops into Mandarin Chinese.” Linguistic Research 32, no. 2: 403–17. doi:10.17250/khisli.32.2.201508.005 Kachru, Braj B. 1985. “Standards, Codification and Sociolinguistic Realism: The English Language in the Outer Circle.” In English in the World: Teaching and Learning the Language and Literatures, edited by Randolph Quirk, H. G. Widdowson, and Yolande Cantù, 11–30. Cambridge: Cambridge University Press for the British Council. ———. 2012. “History of World Englishes.” In The Encyclopedia of Applied Linguistics, edited by Carol Chapelle. Hoboken: Wiley. doi:10.1002/9781405198431. wbeal0513 Kachru, Yamuna, and Cecil L. Nelson. 2006. World Englishes in Asian Contexts. Hong Kong: Hong Kong University Press. Kim, Tae-Eun. 2014. “Preservation and Deletion in Mandarin Loanword Adaptation.” International Journal of Chinese Linguistics 1, no. 2: 214–43. doi:10.1075/ ijchl.1.2.02kim ———. 2019. Mandarin Loanwords. New York: Routledge. doi:10.4324/9781351253406 Kubler, Cornelius C. 1985. “The Influence of Southern Min on the Mandarin of Taiwan.” Anthropological Linguistics 27, no. 2: 156–76. Lantto, Hanna. 2015. “Conventionalized Code-Switching: Entrenched SemanticPragmatic Patterns of a Bilingual Basque-Spanish Speech Style.” International Journal of Bilingualism 19, no. 6: 753–68. doi:10.1177/1367006914552830 Li, Charles N., and Sandra A. Thompson. 1989. Mandarin Chinese: A Functional Reference Grammar. Berkeley: University of California Press. Liao, Sze-wei. 2010. “Identity, Ideology, and Language Variation: A Sociolinguistic Study of Mandarin in Central Taiwan.” PhD diss., University of California, Davis. Lin, Han-Yi. 2012. “Local Responses to Global English: Perceptions of English in Taiwan. Personal Responses toward the Spread of Global English and Its Impact in Taiwan.” English Today 28, no. 3: 67–72. doi:10.1017/s0266078412000119 Lin, Han-Yi, and Chi-Shu Wu. 2015. “English, Competitiveness, and Deficiency: A Critical Discourse Analysis of English Education in Taiwan.” International Journal of English and Education 4, no. 1: 160–75. Lin, Yen-Hwei. 2008. “Variable Vowel Adaptation in Standard Mandarin Loanwords.” Journal of East Asian Linguistics 17, no. 4: 363–80. doi:10.1007/ s10831-008-9031-y Lin, Yi Shiuan. 2017. “Switching to Taiwanese in Mandarin-Dominant Spoken Media Discourse in Taiwan: Evidence of Association as the Main Motivation.” Lingua 198: 53–72. doi:10.1016/j.lingua.2017.07.002 Mair, Victor H. 2010. Is Q a Chinese Character? Language Log. Blog. Accessed January 31, 2019 via http://languagelog.ldc.upenn.edu/nll/?p=2252.

272  Julia Wasserfall Makeham, John, and A-chin Hsiau, eds. 2005. Cultural, Ethnic, and Political Nationalism in Contemporary Taiwan: Bentuhua. New York: Palgrave Macmillan. doi:10.1057/9781403980618 Matras, Yaron. 2009. Language Contact. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511809873 Miao, Ruiqin. 2005. “Loanword Adaptation in Mandarin Chinese: Perceptual, Phonological and Sociolinguistic Factors.” PhD diss., Stony Brook University. Muysken, Pieter. 2000. Bilingual Speech: A Typology of Code-Mixing. Cambridge: Cambridge University Press. Myers-Scotton, Carol. 1992. “Comparing Codeswitching and Borrowing.” Journal of Multilingual and Multicultural Development 13, no. 1–2: 19–39. doi:10.1080/014 34632.1992.9994481 ———. 1993. Duelling Languages: Grammatical Structure in Codeswitching. Oxford: Clarendon. ———. 2002. Contact Linguistics: Bilingual Encounters and Grammatical Outcomes. Oxford: Oxford University Press. doi:10.1093/acprof:oso/9780198299530.001.0001 ———. 2006. Multiple Voices: An Introduction to Bilingualism. Malden, MA: Blackwell. Nall, Timothy M., and Shu-Pin Nall. 2010. “Motivations for Code-Mixing English Words into Spoken Taiwanese and Mandarin Chinese in Taiwan: Qualitative Survey Results.” Guólì Jīnmén Jìshù Xuéyuàn xuébào 4: 57–67. OED (Oxford English Dictionary). “lag, N.1 and Adj.” Online Edition. Oxford University Press. Accessed March 12, 2020 via http://www.oed.com. Oladejo, James. 2006. “Parents’ Attitudes towards Bilingual Education Policy in Taiwan.” Bilingual Research Journal 30, no. 1: 147–70. doi:10.1080/15235882.200 6.10162870 Pan, Yih-Fang. 2014. “English Language Education (ELE) as a Mark of Social Distinction in Taiwan: Buxibans, ELE Teachers Discourse, and Social Reproduction.” PhD diss., University of New Mexico. Park, Joseph Sung-Yul. 2009. The Local Construction of a Global Language: Ideologies of English in South Korea. Berlin: Mouton de Gruyter. doi:10.1515/9783110214079 Poplack, Shana. 1980. “Sometimes I’ll Start a Sentence in Spanish Y TERMINO EN ESPAÑOL: Toward a Typology of Code-Switching.” Linguistics 18, no. 7–8: 581–618. doi:10.1515/ling.1980.18.7-8.581 ———. 1988. “Contrasting Patterns of Codeswitching in Two Communities.” In Codeswitching: Anthropological and Sociolinguistic Perspectives, edited by Monica Heller, 215–45. Berlin: Walter de Gruyter. doi:10.1515/9783110849615.215 ———. 2004. “Code-Switching/Sprachwechsel.” In Sociolinguistics: An International Handbook of the Science of Language and Society, edited by Ulrich Ammon, 589–96. Berlin: Walter de Gruyter. doi:10.1515/9783110141894.1.4.589 Poplack, Shana, and Marjory Meechan. 1998. “How Languages Fit Together in Codemixing.” International Journal of Bilingualism 2, no. 2: 127–38. doi:10.1177/136700699800200201 Poplack, Shana, and David Sankoff. 1984. “Borrowing: The Synchrony of Integration.” Linguistics 22, no. 1: 99–136. doi:10.1515/ling.1984.22.1.99 Poplack, Shana, David Sankoff, and Christopher Miller. 1988. “The Social Correlates and Linguistic Processes of Lexical Borrowing and Assimilation.” Linguistics 26, no. 1: 47–104. doi:10.1515/ling.1988.26.1.47

Conventionalized Code-Switching in Taiwan  273 Riehl, Claudia Maria. 2009. Sprachkontaktforschung: Eine Einführung. 2nd rev. ed. Tübingen: Narr. Schindelin, Cornelia. 2017a. “Word Length.” In Encyclopedia of Chinese Language and Linguistics, Vol. 4, edited by Rint Sybesma et al., 584–89. Leiden: Brill. ———. 2017b. “Word Frequency.” In Encyclopedia of Chinese Language and Linguistics, Vol. 4, edited by Rint Sybesma et al., 580–84. Leiden: Brill. Seargeant, Philip. 2009. The Idea of English in Japan: Ideology and the Evolution of a Global Language. Bristol: Multilingual Matters. doi:10.21832/9781847692030 Su, Hsi-Yao. 2009. “Code-Switching in Managing a Face-Threatening Communicative Task: Footing and Ambiguity in Conversational Interaction in Taiwan.” Journal of Pragmatics 41, no. 2: 372–92. doi:10.1016/j.pragma.2008.10.007 ———. 2011. “The Meaning Contestation of Tai: Language Ideologies and the Global-Local Ambivalence.” Concentric: Studies in Linguistics 37, no. 2: 283–308. Sun, Chaofen. 2006. Chinese: A Linguistic Introduction. Cambridge: Cambridge University Press. doi:10.1017/cbo9780511755019 Tien, Ching-Yi, and David S.C. Li. 2014. “Codeswitching in a University in Taiwan.” In Codeswitching in University English-Medium Classes: Asian Perspectives, edited by Roger Barnard and James McLellan, 24–42. Bristol: Multilingual Matters. doi:10.21832/9781783090914-005 Wei, Jennifer M. Y. 2003. “Codeswitching in Campaigning Discourse: The Case of Taiwanese President Chen Shui-bian.” Language and Linguistics 4, no. 1: 139–65. Wiedenhof, Jeroen. 2015. A Grammar of Mandarin. Amsterdam: John Benjamins. doi:10.1075/z.197 Yip, Moira. 2002. Tone. Cambridge: Cambridge University Press. doi:10.1017/ cbo9781139164559 ———. 2006. “The Symbiosis between Perception and Grammar in Loanword Phonology.” Lingua 116, no. 7: 950–75. doi:10.1016/j.lingua.2005.05.007

13 Ubiquitous but Unplanned The Utterance-Final Particle ê in Taiwan Mandarin1 Chin-hui Lin

Abstract Utterance-final particles (UFPs) are particles without referential meaning that fulfill several pragmatic functions in Taiwan Mandarin. Taking the Taiwan Mandarin UFP ê as an example, this chapter places the analysis of actual language use in the context of language contact and language planning. A brief overview of Mandarin UFPs shows the degree of ubiquity of this phenomenon. It is argued that there are historical reasons for this ubiquity, which can be related to complex patterns of migration from mainland China to Taiwan before and during the twentieth century and the resulting social structures of Taiwan’s society. It is also claimed that the UFP ê constitutes an especially noteworthy outcome of language contact in Taiwan, since it does not involve the usual candidate languages as donor languages, i.e. standard Mandarin or Southern Mǐn, but Jiāng-Huái Mandarin instead.

Introduction Whenever sinophone studies relate to language, Mandarin takes center stage. Interestingly, however, with respect to the question of which features belong to Mandarin and which ones do not uncertainty often remains. This has to do with the fact that the English term “Mandarin” has different equivalents in Chinese, each of which has a different connotation. Mandarin can refer to ideal and abstract governmental standards promoted in different polities of the sinophone world. In mainland China, this standard is known as Pǔtōnghuà ‘common language’, in Taiwan as Guóyǔ ‘national language’, and in Singapore as Huáyǔ ‘(ethnic) Chinese language’. These abstract standards need to be distinguished from the Mandarin varieties actually spoken in the various regions of the sinophone world. In this chapter, I will use Pǔtōnghuà and Guóyǔ to refer to the official standards in mainland China and in Taiwan, respectively, and “Taiwan Mandarin” to the Mandarin spoken in Taiwan. The term “Mandarin” denotes the actual spoken Mandarin in both mainland China and Taiwan unless otherwise specified. Taiwan Mandarin not only differs from standard Guóyǔ. Due to different historical developments and long-term separation, Taiwan Mandarin is

Ubiquitous but Unplanned  275 also considerably different from Pǔtōnghuà and other Mandarin varieties (Kubler 1985). For example, Diāo (1998: 387–90) points out that compared to Pǔtōnghuà, Taiwan Mandarin has been strongly influenced by Japanese and Southern Mǐn. If we look at the commonly used lexicon on both sides of the Taiwan Strait, significant linguistic differences become obvious (see Klöter 2017; Qiú and van den Berg 1994). Among the most frequently cited examples is the ubiquitous use of UFPs. Taking the Taiwan Mandarin UFP ê as an example, this chapter places the analysis of actual Mandarin language use in the context of language contact and language planning. The main question I will address is: How has the UFP ê emerged in Taiwan Mandarin? A brief overview of Mandarin UFPs shows the degree of ubiquity of this phenomenon in Taiwan Mandarin. I argue that there are historical reasons for this ubiquity, which can be related to complex patterns of migration from mainland China to Taiwan before and during the twentieth century and the resulting social structures of Taiwan’s society. I also claim that the UFP ê constitutes an especially noteworthy outcome of language contact in Taiwan, since it does not involve the usual candidate languages as donor languages, i.e. standard Mandarin or Southern Mǐn, but Eastern Mandarin (Jiāng-Huái Mandarin) instead. This hypothesis, which is fully reconcilable with migration history, is then analyzed on the basis of two types of evidence: spoken language data attested in early radio plays and interviews with Mandarin native speakers of different regional origins.

Utterance-Final Particles in Language Contact UFPs exist in many Sinitic varieties such as Mandarin, Southern Mǐn, and Cantonese. They occur at the utterance-final position, and, as Wu (1997: 98) states for Mandarin, are “essentially discourse-dependent: they often do not have a definite denotative or referential meaning, but are mainly used, among other things, to convey speaker’s attitude, feeling, stance, and/or disposition in a discourse context.” In other words, UFPs in different Sinitic varieties usually have no referential meaning and carry pragmatic/discourse functions. According to Appel and Muysken (1987), Curnow (2001), and Matras (2000), this type of word appears to be easily borrowed or transferred from language to language.2 In addition, Cheng (1997: 149) argues that the linguistic features which serve to indicate the speaker’s emotive feelings are transferred to the speaker’s second language naturally, even when s/he is a proficient secondlanguage user. A case in point is the English spoken by Chinese people in Singapore, which is mixed with particles and interjections from Southern Mǐn (known as Hokkien in Singapore; for a discussion of Colloquial Singapore English or Singlish, I refer to see Siemund and Li, this volume). Comparing language use in Taiwan and mainland China, Chén (2008: 116) points out that the deployment of UFPs in Taiwan Mandarin is a common phenomenon. There are not only quite a few distinct UFPs, the frequency of

276  Chin-hui Lin their occurrence in spoken language use is also considerably higher than in standard Mandarin. The excerpt below illustrates this. In this short Taiwan ̊ Here, Mandarin conversation, 4 UFPs are used: ma, la, a, and hoNn [hɔŋ]. hoNn is a Southern Mǐn syllable, which shows that Southern Mǐn UFPs are easily transferred to Taiwan Mandarin. 1

1

D

cóng

zǎoshàng

kāishǐ

guàng

ma?







→ 4

M

xiǎng-

lèi

le

huílái

xiūxí

a

hoNn.

     

D1: Does your shopping start in the morning? M2–4: It- it doesn’t matter. If I want to, then I go shopping. If I want- if I am

tired, I go home and take a rest. Table 13.1 compares the UFPs in my Taiwan Mandarin conversational data with those listed in standard Mandarin dictionaries, textbooks, or grammar books in mainland China and Taiwan.3 The comparison confirms my claim that the number of UFPs in Taiwan Mandarin conversation is much higher than that in standard Mandarin. How can we account for the high number and distinct use of UFPs in Taiwan Mandarin in comparison to standard Mandarin? Except for the first five UFPs, which exist both in standard Mandarin and Taiwan Mandarin, Table 13.1 A comparison of utterance-final particles in standard Mandarin and spoken Taiwan Mandarin UFPs in standard Mandarin

UFPs used in Taiwan Mandarin conversation

ma [ma] ba [pa] ne [nə] a [a] o [ɔ]

ma [ma] ba [pa] ne [nə] a [a] o [ɔ] la [la] ̊ hoNn [hɔŋ] ̊ haNn [hɑŋ] hioh [hɪɔʔ] lê [lɛ] nê [nɛ] ê [ɛ]

Source: Data collected from CASS (2010), Hé (1987), MOE (1994), and Tseng (2013).

Ubiquitous but Unplanned  277 the most obvious explanation for the high number of the UFPs may be the language contact between Mandarin and Taiwan’s local languages (for example, Southern Mǐn or Hakka). Against the background of Taiwan’s migration history and language policy after 1945, it can be assumed that when local Southern Mǐn speakers learned Mandarin, some linguistic features of their first language (for instance, the UFPs) were transferred to the second language (i.e. Mandarin). This scenario may explain the existence of la [la], ̊ haNn [hɑŋ], ̊ hioh [hɪɔʔ], lê [lɛ], and nê [nɛ] in Taiwan Mandarin. hoNn [hɔŋ], However, a closer look at Table 13.1 shows that this explanation is not sufficient. Most importantly, the UFP ê exists neither in standard Mandarin nor in any other local language spoken in Taiwan. How, then, has it emerged in Taiwan Mandarin? Does it originate from another Sinitic variety? If so, which one? In order to answer these questions, I shall first introduce the linguistic setting of Taiwan and the historical background to the formation of today’s Taiwan Mandarin.

A Short History of Taiwan’s Languages According to Zhōu (1996: 174–75), immigration of Chinese people from mainland China to Taiwan started in the tenth century, but the number of immigrants remained low in the first centuries. After the mid-sixteenth century, more and more southern Fujianese fishermen and merchants settled in Taiwan. Zhōu reports that by 1926, “Fujianese made up some 73.5 percent of the population in Taiwan” (ibid., 177).4 As a result, Southern Mǐn, the regional variety spoken in the southern part of Fujian province, became widespread throughout Taiwan. From 1895 to 1945, Taiwan was a colony of Japan. The Japanese government launched a Japanese language promotion campaign. Huang (1993: 96) estimates that by 1944, 71 percent of the Taiwanese population had become proficient in Japanese. However, Japanese was only used in public and not in private domains. Huang also concludes that the promotion of Japanese made most Taiwanese people into Japanese-Southern Mǐn bilinguals (ibid., 99). In 1945, Taiwan was ceded to the government of the Republic of China (ROC). For ideological reasons, the ROC government decided to promote the use of Guóyǔ in Taiwan through the Taiwan Provincial Committee for the Promotion of the National Language (台灣省國語推行委員會) which was established in 1946. Its task was to replace Japanese with Mandarin within a short period of time (see also Klöter, this volume). According to P. Chen (1999: 31–32), “up until 1987, schoolchildren in Taiwan could be penalized for speaking anything other than guóyǔ. The local dialects were either banned from mass media, or highly restricted in terms of time and budget allocation until quite recently.” This campaign has attained “considerable achievements in converting Mandarin into the lingua franca in Taiwan” (Li 1985: 123). Prior to its import to Taiwan, the national language had been promoted in mainland China since the early twentieth century. Since the 1920s, it had

278  Chin-hui Lin been based on the pronunciation of the Mandarin variety spoken in Beijing (cf. Chen 1999: 22; Klöter, this volume; Wáng 1995: 277) and its grammar was based on literature written in the vernacular literary language.5 It has to be emphasized, however, that Guóyǔ had originally not been the native language of speakers of any one region in mainland China, including Beijing, nor of Taiwan. The national language needed to be “acquired” as a second language, as Tung has claimed: Guóyǔ is not the Beijing dialect. … It is never the case that a country’s standard language is equivalent to one of its local dialects. … I would like to say some words to guóyǔ learners: you can always learn guóyǔ very well. People who grow up in Beijing also need to study to acquire good guóyǔ. These people have some advantages because their native language is closer to guóyǔ. However, if they do not study, what they speak is always the Beijing dialect, and never guóyǔ. (Tung 1974: 367–68) Since the national language is learned as a second language, interference from the learner’s first language (i.e. native language) is inevitable. As Li and Thompson explain: [B]oth Putonghua and Guoyu are far from being “uniform,” for China has a large population spread over a vast geographical area, and consequently numerous other dialects inevitably influence and affect the versions of Putonghua and Guoyu spoken by people from different regions. Thus, a truly uniform language in a country such as China can exist only in theory, not in reality. … There will always be some variation between “the Mandarin language” of one person and “the Mandarin language” of another person. (Li and Thompson 1981: 1) A local Mandarin variety can be described as an “inter-language occurring in the process of which a dialect speaker learns the non-native common language” (Chén 1991: 13). Such a local Mandarin variety has been termed “local variant of the common language” (地方普通话; ibid.). These varieties have thus emerged naturally. Taiwan Mandarin can likewise be seen as a local Mandarin variety: an inter-language generated by local dialect speakers. As mentioned earlier, political campaigning led to the successful establishment of Guóyǔ as a lingua franca in Taiwan. It is, however, important to note that similar to the promotion of Japanese before 1945, ROC language policies did not lead to the extinction of local languages, but created a “diglossia with bilingualism” society (Tsao 2000: 280). That is, Mandarin came to be used in the public domain, whereas the local languages such as Taiwan Southern Mǐn or Hakka are still used in the private domain. Sandel confirms that despite the fact that Mandarin Chinese has been promoted as

Ubiquitous but Unplanned  279 the language of instruction in schools since 1945, “a majority of the island’s inhabitants also speak one of a number of ‘local languages’” (2003: 527). Among these local languages, Southern Mǐn, due to its large number of speakers, has undoubtedly been the most influential source of today’s Taiwan Mandarin. As Teng asserts, “[c]ontributions towards the formation of Taiwanese Mandarin [i.e. Taiwan Mandarin] came mostly, if not entirely, from Southern Min” (2002: 232). Southern Mǐn, together with Taiwan Mandarin, is thus the most widely spoken Sinitic variety on the island. It is especially widespread in southern Taiwan where it is used as the main language of communication in private settings, in shops, and on the streets (cf. Qiú and van den Berg 1994). Although Taiwan is home to many languages, Southern Mǐn is often referred to as Taiwanese or Táiyǔ (lit. ‘Taiwanese language’), suggesting that it is the linguistic representative of the entire language community. In reality, however, the idea underlying the collective term Táiyǔ is misleading, since it conceals the fact that Taiwan Mandarin, the lingua franca, is used all over the island as the main language of media, education, and government administration. The successful promotion of Mandarin, furthermore, does not imply that the language has been acquired successfully from a prescriptive point of view. Instead, the language which most local Taiwanese people have acquired is a kind of “non-native Mandarin” (Cheng 1985: 354). This may be attributable to historical factors: after losing the Chinese civil war to the Communist Party in 1949, the ROC government withdrew to Taiwan. During 1949 and 1950, refugees and immigrants, including many army officers and their families, moved from mainland China to Taiwan. Citing the 1956 population census (PCO 1959), Huang (1993: 22) estimates that the total population of these mainland immigrants in Taiwan at the end of 1956 was about 1.21 million (the total population: 9.37 million; cf. PCO 1959). However, according to this 1956 census, less than 1 percent of these mainland immigrants came from Beijing, the heartland of standard Mandarin. Many came from different provinces all over China and spoke non-Mandarin dialects as their mother tongue. Applying the dialect classification in Yuán (1989) and Norman (1988), Kuo (2005: 76–78) points out that these dialects included various Mandarin varieties as well as all other Sinitic dialect groups (Xiāng, Yuè, Wú, Mǐn, Kèjiā [Hakka], Gàn). Furthermore, Kuo’s figures imply that more than 40 percent of these first-generation mainland immigrants came from non-Mandarin speaking areas. Among the Mandarin speakers, only 20 to 25 percent spoke Northern or Northwestern Mandarin (whose Mandarin is closer to the norm); Southwestern and Eastern Mandarin speakers accounted for ca. 30 percent. Her (2009: 27) also claims that “half of the first-generation mainland immigrants come from southern language areas [i.e. non-Mandarin speaking areas], including Wú, Yuè, Hakka and Mǐn, etc.” In the light of this enormous quantitative disparity between native Beijing Mandarin speakers or speakers who could more or less speak “standard”

280  Chin-hui Lin Mandarin on the one hand and the total population on the other hand, the shortage of qualified personnel required for the promotion of Mandarin was an obvious problem. As a consequence, many non-native Mandarin mainland immigrants taught Mandarin in schools. An editorial of the China Daily News (中華日報) in 1947 describes the problem as follows: “Some teachers who teach Guóyǔ cannot speak Mandarin well themselves. Some speak ‘Cantonese Mandarin’, some speak ‘Zhèjiāng Mandarin’, some even teach Mandarin in Shanghainese” (Zhōnghuá Rìbào 1947). As mentioned earlier, for these mainland immigrants who were non-native speakers of the national standard, Mandarin can be considered their “second language.”6 Among the group of mainlanders, speakers of Wú dialects dominated. LaPolla writes that these “speakers attempted to teach the Taiwanese population Mandarin, and forced the Taiwanese to speak it even amongst themselves. The Taiwanese did not generally have access to native speakers” (2001: 234). Moreover, the Wú speakers, as I will explain in more detail presently, were quite powerful in the realms of politics, economics, culture, education, and communication after 1949 (cf. Cheng 1985; Tāng 1999). From the outset, Mandarin also fulfilled an important function as a lingua franca of speakers with different regional backgrounds. Most of the first-generation mainlanders spoke their own dialects in private settings (see also Chen, this volume). Since many Sinitic dialects (especially southern dialects) are mutually unintelligible, these people had to resort to Mandarin when communicating with people with a different linguistic background. Chén Hào, a second-generation mainlander, describes his childhood as follows: When we lived in the dormitory of teachers, there were only five or six households of mainland immigrants. They came from Sichuan, Fuzhou, Shandong, and Hubei. When we met, we normally talked in Mandarin. But when we listened to the other mainlanders speaking with their family members at home, we could not understand a single word. It was a mystery to me… (Chén 2004: 79) Significantly, except for school education, non-native Mandarin speakers exercised a huge influence in the mass media. Before the lifting of martial law in 1987, the mass media in Taiwan were controlled by the ruling party, the Kuomintang (KMT). According to a survey conducted by United Daily News (聯合報) in 1987 (quoted by Chu 1998: 54), 76.28 percent of employees of the main TV channels were mainlanders. Cheng (1988: 99) points out that only 7.7 percent of employees of the Broadcasting Corporation of China (中國廣播公司), the main radio station in Taiwan, were non-mainlanders. He also reports that until around 1975, in broadcasting, “it was not uncommon to have PM [Beijing Mandarin] speakers read manuscripts written by non-PM speakers from southern China. In written mass media, the writings of non-PM speakers constitute a far larger volume than those of PM

Ubiquitous but Unplanned  281 speakers” (Cheng 1985: 354). He further claims that although Beijing Mandarin was the designated standard and supposed to be the common model, “in daily language contact, non-native Md [Mandarin] has been the actual model” (ibid.). Non-native Beijing Mandarin speakers were “more influential in affecting the grammar of spoken TM [Taiwan Mandarin] than those of PM [Beijing Mandarin] speakers” (ibid.). Although Southern Mǐn has certainly been the most obvious and the most influential linguistic source of today’s Taiwan Mandarin, Kubler (1981:  2) notices that there are other sources as well. His implicit argument that the quantity of speakers is not the sole factor determining outcomes of language contact situations is in line with Siemund’s claim that aside from the number of speakers, “the relative social status of the groups involved as well as the relative prestige of the language to a great extent determine the linguistic outcome of language contact” (2008: 4). If we take the other social parameters mentioned by Siemund into consideration, the influence of people from Jiangsu and Zhejiang provinces is striking. Cheng likewise mentions the special social status of Wú speakers in Taiwan society: [A]mong non-Tw [Taiwanese] speakers of Md [Mandarin] on Taiwan, those not originally PM [Beijing Mandarin] speakers are much more numerous, and are politically and economically more powerful than PM [Beijing Mandarin] speakers. Especially influential are the Wú speakers—who include the political elite from Zhèjiāng, President Chiang’s [Chiang Kai-shek’s] home province, and the financial tycoons and textbook writers from Shanghai. (Cheng 1985: 354) Cheng (1990: 17–18) furthermore points out that when people who speak Taipei Mandarin go abroad and meet people from Beijing, they are often regarded as Shanghainese. The reason is that Taipei Mandarin has been greatly influenced by the people from Shanghai and Zhejiang (who speak Wú dialects).7 In a similar vein, on the basis of his investigation of dialect loanwords in Taiwan Mandarin, Tāng (2002: 259) concludes that 879 out of 1080 dialect loanwords in his data are from Southern Mǐn, 116 are from Wú, and 68 are from Hakka. The reasons for the high portion of loanwords from Southern Mǐn are self-evident. Yet it is interesting to see that Wú loanwords account for around 10 percent of the total number of loanwords, which is even more than Hakka, one of the local languages in Taiwan.8 Tāng argues that the influence of Wú dialects on Taiwan Mandarin is the result of immigration (2001: 365). In Tāng (1999) he underpins this claim by pointing to the following political factors: Nanjing had been the seat of the ROC government during the two decades prior to its relocation to Taiwan in 1949. Thus, people from surrounding Jiangsu and Zhejiang provinces were not merely powerful and

282  Chin-hui Lin influential in political and economic realms, but also in education (Ang 1992; Tāng 1999). Tāng points out that after the ROC government’s relocation to Taiwan, “people speaking Wú dialects or Wú-style Mandarin have a higher status and are more powerful in the realms of politics, economics, culture, education and communication in Taiwan society” (2001: 366). He further argues that “the influence of a language does not always depend on the number of its speakers, but its social status and social value” (ibid.), which is in line with Siemund’s (2008) argument quoted earlier. Tāng’s arguments concerning the influence of Wú on the formation of Taiwan Mandarin (Tāng 1999, 2001) are convincing. It must, however, be pointed out that people from Jiangsu and Zhejiang do not only speak Wú. In some regions, such as Nanjing, Eastern Mandarin (Jiāng-Huái Mandarin) is spoken (Norman 1988: 191). It is therefore necessary to consider the influence of Jiāng-Huái Mandarin when analyzing external influence on today’s Taiwan Mandarin.

The Emergence of ê in Taiwan Mandarin Against this backdrop, we now return to our opening question: How has the UFP ê [ε] emerged in Taiwan Mandarin? Before addressing this question, I shall briefly describe its main functions in spoken discourse. In both standard Mandarin and Southern Mǐn, the particle ê can occur at utterance-initial position and functions as an interjection.9 Examples 2a–d are taken from a normative dictionary published in mainland China, the Contemporary Chinese Dictionary (XHC 2010: 358):10 2

ē,

a

nǐ kuài lái! 2SG quick come ‘Hey, come here quickly!’ é, tā zěnme zǒu le? prt 3SG why go prt ‘How come he left?’ ě, nǐ zhè huà kě bù prt 2SG this word but neg ‘Come on, what you said is not correct!’ è, wǒ zhè jiù lái! prt 1SG this just come ‘Okay, I will come in a moment!’ prt

b c d

duì right

ya! prt

As in standard Mandarin, the Southern Mǐn particle ê occurs in utteranceinitial position only and not in utterance-final position. Example 3 is taken from Tung (2001: 318): 3

E5,

i1 kong2 an3ne1 kam2 tioh8? 3SG say this.way whether right ‘Is it right that he said it this way?’

prt

Ubiquitous but Unplanned  283 In spoken language corpora of Taiwan Mandarin, however, the UFP ê occurs in four types of contexts: (i) storytelling/reporting, (ii) topic introduction/topic shift, (iii) qualified disagreements/upgraded agreements, and (iv) answers (cf. Lin 2014: 133). It is usually attached to a piece of information or to the speaker’s assessment. In Lin (2014: 132) I argue that the core function of ê is to induce a collaborative move (either by the speaker him/herself or elicited from the interlocutor) by foregrounding the utterance to which ê is attached. The use of ê is usually triggered when something in the previous context is counter to the speaker’s assumption, or when the speaker believes something in the previous context has been misconceived because some information is missing. With the deployment of ê, the speaker renders the utterance prominent and offers an opportunity for the interlocutor to respond. Hence, ê can mitigate a negative effect when it occurs with dispreferred moves such as a topic shift, disagreement, or a refusal. When it occurs in a reporting turn or in preferred responses such as agreement, it enhances the interaction and strengthens the positive effect of these moves (for a detailed discussion, see Lin 2014: 123–60). In the light of the historical contexts of Taiwan Mandarin formation, it seems reasonable to assume that ê was first used by the first-generation mainland immigrants whose major language was not Beijing Mandarin. If this assumption is correct, then ê must be attested in the Mandarin spoken by first-generation migrants from the mainland. Since no spoken corpora were established at that time, my search for early attestations of ê had to rely on radio plays. As mentioned earlier, immigrants from the mainland controlled the mass media in Taiwan before the lifting of martial law in 1987, and a huge majority of the employees of the Broadcasting Corporation of China were local non-mainlanders (see Cheng: 1988: 99). In order to verify my assumption, I checked two episodes of a radio play recorded in the 1960s.11 It turned out that the UFP ê is used in both episodes. The fact that almost all speakers and authors of these plays were mainlanders supports my hypothesis that ê must have been in use in the initial phase of MandarinSouthern Mǐn language contact. In a second step, in order to locate the possible source language of ê geographically, I interviewed 13 mainland Chinese persons with distinct dialect backgrounds.12 All these respondents confirmed that ê is not used in Pǔtōnghuà.13 Significantly, among the 13 informants, only Jiāng-Huái Mandarin speakers coming from Anhui, northern Jiangsu, and the city of Nanjing confirm that they use the UFP ê when speaking Mandarin. This corresponds to what Lǐ (1995: 107) and Chao (1926: 905) report about the use of the UFP ε in the Nanjing dialect. Examples 4a–c are quoted from Chao (ibid.):14 4

a

ʂɨ tsəmə tsɔ də ε! be so do de prt ‘It should not be done in this way!’

pu

neg

284  Chin-hui Lin b c

t‘a piŋ mε jɔ ki ε! 3SG at.all neg have go prt ‘He didn’t go!’ pu ʂɨ tɕ‘ɨ lə ʥiu suan lə neg be eat asp just count asp ‘Don’t think you can get away with eating it!’

ε! prt

Although Chao (1926) does not provide any conversational contexts for these examples, he mentions that these utterances are used as disagreements, which is similar to contexts where the Taiwan Mandarin ê occurs: qualified disagreements/upgraded agreements. The examples below are taken from Lǐ (1995: 107). He maintains that the UFP ε in the Nanjing dialect can be attached to declarative sentences (example 5a) or imperative sentences (example 5b). 5

a

b

ni ʂɨ kanpu ε, 2SG be cadre prt tsəmə nəŋ ʥiaŋ tsə tsɔŋ pu futsərən di hua? how can say this cl neg responsible de saying ‘You are a cadre. How could you say such irresponsible things?’ pu Jau kuaŋ ʥiaŋ hua, tɕhɨ tɕhai ε! neg must only say saying eat dish prt ‘Don’t just talk, eat!’

My informant from Nanjing confirms that the use of ε in examples 4a–c and 5a is quite similar to the use of ê in Taiwan Mandarin: ε is also triggered by a situation which diverges from the speaker’s assumption. By using ε, the speaker foregrounds the utterance to which it is attached. Although the ε-attached utterance in 5a is followed by a question, the following utterance need not be explicitly made. I therefore argue that example 5a can still be considered a case similar to the ê-attached disagreement in Taiwan Mandarin.15 Except for Nanjing, the UFP ε also exists in some other Chinese dialects: in Anhui, Jiangsu, and Zhejiang provinces, for example, it exists in the Yángzhōu 扬州 and Jīnhuá 金华 dialects. Examples 6a and 6b are taken from Huáng (1996a: 108; 1996b: 123). 6

a b

liɪ suoʔ sənmən? o t‘iŋ pəʔ tɕh iŋtshu 2SG say what 1SG hear neg clear ‘What did you say? I could not hear it clearly!’ kəʔ kəʔ tifɑŋ kɤsiŋ ε! this cl place clean prt ‘This place is clean!’

ε! prt

The earlier dialect data provide additional support for my claim that the UFP ê was very likely first imposed on Mandarin by people from the

Ubiquitous but Unplanned  285 Nanjing region and the area covering the provinces of Anhui, Jiangsu, and Zhejiang. It was then imported to Taiwan through the Mandarin spoken by these first-generation mainland immigrants. As mentioned previously, people from Jiangsu and Zhejiang had high positions in education, politics, and business (cf. Ang 1992: 239–40; Tāng 1999). In previous studies, Wú dialects have been mentioned as one important lexical contributor to Taiwan Mandarin (cf. Tāng 2002). In addition, Jiāng-Huái Mandarin, spoken “in central Ānhuī, and Jiāngsū north of the Yangtze, as well as in the region of Nánjīng” (Norman 1988: 191), has obviously also played a role in the formation of the Taiwan Mandarin UFP system.

Concluding Remarks A linguistic analysis of Mandarin yields very different results depending on the kind of data on which it is based. As the example of the Taiwan Mandarin UFP ê shows, its very existence would remain undiscovered if we look at prescriptive accounts of the standard language only. Due to its frequent use it can rightly be considered a ubiquitous feature of spoken Taiwan Mandarin. Its specific functions can only be identified through the analysis of actual usage. Within the long list of Taiwan Mandarin UFPs, ê is a special case: It is apparently the only UFP which is attested neither in standard Mandarin nor in Southern Mǐn, the two varieties that have contributed most significantly to the formation of Taiwan Mandarin. In this chapter I propose that the UFP ê was imported to Taiwan by speakers of Jiāng-Huái Mandarin. This assumption can be supported by historical and linguistic facts. From a historical perspective, after the retreat of the KMT government to Taiwan in 1949, people from Jiangsu and Zhejiang provinces (including Wú speakers and Jiāng-Huái Mandarin speakers) held important positions in government, educational institutions, and state media. Significantly, an examination of radio plays of the 1960s shows a strong presence of ê. Interviews with Mandarin native speakers of different regional backgrounds have revealed the Jiāng-Huái Mandarin identity of ê. To be sure, the planning of Mandarin in Taiwan has been extremely successful. The very fact that Mandarin is now the most widely used language in Taiwan is a result of language planning. But planning cannot explain actual usage in its entirety. Many features of actual language use have permeated the habits of speakers on different paths. In this respect, even “small” features like UFPs are closely linked to the complex social histories of their speakers.

Notes 1 This chapter is based on Lin (2014). 2 In Appel and Muysken (1987), Curnow (2001), and Matras (2000), “discourse marker” is used. For a detailed discussion about the equivalence of UFP and discourse marker, see Lin (2014: 23). 3 Most of the Taiwan Mandarin spontaneous spoken data in this chapter come from the Mandarin Topic-oriented Conversation Corpus collected by Academia

286  Chin-hui Lin

4

5 6 7

8

Sinica in Taiwan. The MTCC corpus consists of 29 spontaneous dialogues between two speakers who are familiar with each other. Their ages range from 14 to 63. Each pair of speakers was asked to choose a specific topic for the conversation. The total length of recording is c. 11 hours. I have excluded two dialogues, which are basically carried out in Southern Mǐn. My analysis is thus based on 27 Taiwan Mandarin dialogues involving 31 female speakers and 23 male speakers. In addition to the MTCC data, some examples come from my own recordings of Taiwan Mandarin made in 2007 and 2008. The settings of my own recordings are similar to those used for the MTCC. My own recordings made in Taiwan involve 12 female speakers and 4 male speakers. Their ages range from 25 to 55. I have also used four excerpts from another public Taiwan Mandarin corpus, the NCCU corpus of spoken Chinese which is similar to the other two corpora in the relevant respects. Zhōu (1996: 177) writes that in 1926, the total number of Hàn people (i.e. the major ethnic group of China) was 3,751,600, or 88.4 percent of the total population. Among these Hàn people, the total number of Fujianese people is 3,120,000, which amounts to about 83.1 percent of the population of Taiwan at the time. The earliest codification of the grammar of the vernacular literary language is Li (1924). However, the influence of this book in the Guóyǔ promotion campaign is unclear. Her defines the second language as “a local language that a person learns or acquires in one’s teens” (Her 2009: 4). The “Taipei Mandarin” mentioned by Cheng (1990: 17) is a relatively “standard” form of Taiwan Mandarin and mostly used in Taipei. Ang (1985: 97–98, 1992: 98–101) defines “Taipei Mandarin” as the mother tongue of the second- and third- generation mainlanders and some local third-generation non-mainlanders (Taiwanese people). It is different from “Taiwanese-accented Mandarin,” which is the second mother tongue of the local Taiwanese people. The influence of Wú dialects is not restricted to the formation of Taiwan Mandarin. They have also exerted considerable influence on Modern Standard Chinese, i.e. Guóyǔ. Quoting Hsü (1979), Davies (1992: 203) indicates that many of the most influential writers in the 1920s and 1930s were native speakers of regional dialects (i.e. non-Northern Mandarin): of the 213 writers who were active in that period of time, more than 80 percent were from non-Northern Mandarin areas and 40 percent from Wú dialect areas. Chen (2001: 56) also writes: Famous writers such as Lu Xun, Zhou Zuoren, Yu Dafu, Xu Zhimo, Mao Dun and Ye Shengtao were all native speakers of the Wu dialect. In their writings, all of them displayed features characteristic of the grammar and vocabulary of their native tongue. Because of the popularity of these writers, many Wu features subsequently became part of established Modern Written Chinese norms.

Davies (1992: 205) points out that the Wú dialect is the most dominant source concerning the part of Modern Standard Chinese which derives from non-Northern Mandarin regional dialects other than Northern Mandarin. It is thus reasonable to claim that Wú dialects to some extent shaped today’s standard Mandarin in both mainland China and Taiwan, i.e. Pǔtōnghuà and Guóyǔ. 9 Huáng et al. (1997: 152) claim that ê, when occurring in utterance-final position, is not used in standard Mandarin. 10 According to XHC (2010: 358), the utterance-initial ê [ε] can also be pronounced as ei [eɪ̯ ]. The normative dictionaries in Taiwan, for example, Hé (1987) and MOE (1994), likewise write that ê is used in utterance-initial position. 11 The titles of these two radio plays are: Wàn rén bǎotǎ 萬人寶塔 ‘A precious tower made by ten thousand people’ and Shēng cái yǒu dào 生財有道 ‘There’s a way to

Ubiquitous but Unplanned  287

12

13

14 15

make fortune’. They were both produced by the Broadcasting Corporation of China in the 1960s. Following Norman’s (1988) classification of Chinese dialects, the dialects used by my mainland Chinese interviewees can be broken down as follows: Northern Mandarin (one male speaker from Beijing and one male speaker from Shandong), Northwestern Mandarin (one female speaker from Shaanxi), Southwestern Mandarin (one female speaker from Sichuan), Jiāng-Huái Mandarin (one male speaker from northern Jiangsu, two female speakers from Nanjing, one female speaker from Anhui), Wú (one female speaker and one male speaker from Shanghai, one female speaker from Zhejiang), Mǐn (one male speaker from Fujian), and Yuè (one female speaker from Hong Kong). Their ages range from 25 to 50. Although some of them admit that they occasionally use short ê-attached phrases such as bù zhīdào ê ‘I don’t know’ or méiyǒu ê ‘no’, they believe that this kind of usage results from the influence of Taiwan TV drama series, which are very popular in mainland China. The examples in Chao (1926) are written in Chinese characters. I have transcribed the examples in International Phonetic Alphabet (IPA) without tones according to the pronunciation of my respondent from Nanjing. I have not found the “imperative” use in example 5b in my Taiwan Mandarin data. It is not easy to explain why only certain functions have been transferred and others not. The partial transfer of functions is in line with Matras and Sakel’s (2007: 835–36) claim that the “outcome [of contact-induced change] need not, however, be a one-to-one correspondence between form and function throughout the construction.” For Taiwan Mandarin ê, we can merely identify this incongruity. Since Taiwan Mandarin ê results from different stages of language contact involving different varieties, more data would be required to ascertain exactly when and under what circumstances a certain function of the source language, i.e. marking imperatives, has been dropped.

References Ang Ui-jin 洪惟仁. 1985. Táiwān Hélǎoyǔ shēngdiào yánjiù 台灣河佬語聲調研究 [A study of Taiwan Hoklo tones]. Taipei: Zìlì wǎnbào shè. ———. 1992. Táiwān fāngyán zhī lǚ 台灣方言之旅 [A tour of Taiwan dialects]. Taipei: Qiánwèi chūbǎnshè. Appel, René, and Pieter Muysken. 1987. Language Contact and Bilingualism. London: Edward Arnold. Chao Yuen Ren 趙元任. 1926. “Běijīng, Sūzhōu, Chángzhōu yǔzhùcí de yánjiū” 北京、蘇州、常州語助詞的研究 [A study of particles in Beijing, Suzhou, and Changzhou dialects]. Qīnghuá xuébào 3, no. 2: 865–917. Chén Hào 陳浩. 2004. Yī èr sān, dào Táiwān 一二三，到台灣 [One, two, three, this is Taiwan]. Taipei: Shíbào chūbǎn gōngsī. Chén Liǔfēn 陈柳芬. 2008. “Cóng Táiwān yúlè jiémù kàn dāngdì de yǔyán wénhuà tèsè” 从台湾娱乐节目看当地的语言文化特色 [Linguistic and cultural features in Taiwan from the perspective of Taiwanese TV entertainment programs]. Shǎnxī Shīfàn Dàxué xuébào (zhéxué, shèhuì kēxué bǎn) 37, no 1: 114–18. Chen, Ping. 1999. Modern Chinese: History and Sociolinguistics. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139164375 ———. 2001. “Development and Standardization of Lexicon in Modern Written Chinese.” In Language Planning and Language Policy: East Asian Perspectives, edited by Ping Chen and Nanette R. Gottlieb, 49–73. Richmond: Curzon.

288  Chin-hui Lin Chén Yàchuān 陈亚川. 1991. “‘Dìfāng Pǔtōnghuà’ de xìngzhì tèzhēng jí qítā” “地方 普通话”的性质特征及其他 [The features and characteristics of local Mandarin varieties]. Shìjiè Hànyǔ jiàoxué 1: 12–17. Cheng Jei-cheng 鄭瑞城. 1988. Tòushì chuánbō méijiè 透視傳播媒介 [Exploring communication media]. Taipei: Jīngjì yǔ shēnghuó chūbǎn gōngsī. Cheng, Robert Liang-wei. 1985. “A Comparison of Taiwanese, Taiwan Mandarin, and Peking Mandarin.” Language 61, no. 2: 352–77. doi:10.2307/414149 ——— 鄭良偉. 1990. Yǎnbiàn zhōng de Táiwān shèhuì yǔwén: duōyǔ shèhuì jí shuāngyǔ jiàoyù 演變中的台灣社會語文: 多語社會及雙語教育 [Taiwan society and languages in transition: A multilingual society and bilingual education]. Taipei: Zìlì wǎnbàoshè wénhuà chūbǎnbù. ———. 1997. Tái, Huáyǔ de jiēchù yǔ tóngyìyǔ de hùdòng 台、華語的接觸與同義 語的互動 [Contacts between Taiwanese and Mandarin and the interaction of synonyms]. Taipei: Yuǎnliú. Chu Chuang-Pin 朱全斌. 1998. Méitǐ, rèntóng yǔ chuánbō kējì: Guānxīn běntǔ chuánbō chuánméi fāzhǎn de wénhuà yánjiù 媒體、認同與傳播科技: 關心本土傳 播傳媒發展的文化研究 [Media, identity and communication technologies: Cultural studies of the development of local media]. Taipei: Yuǎnliú. Curnow, Timothy Jowan. 2001. “What Language Features Can Be ‘Borrowed’?” In Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics, edited by Alexandra Y. Aikhenvald and Robert M.W. Dixon, 412–36. Oxford: Oxford University Press. Davies, Peter. 1992. “The Non-Beijing Dialect Component in Modern Standard Chinese.” In Sociolinguistics Today: International Perspectives, edited by Kingsley Bolton and Helen Kwok, 192–206. London: Routledge. Diāo Yànbīn 刁晏斌. 1998. “Táiwānhuà de tèdiǎn jí qí yǔ nèidì de chāyì” 台湾话的特 点及其与内地的差异 [The characteristics of Taiwan Mandarin and the differences between Taiwan Mandarin and mainland Mandarin]. Zhōngguó yǔwén 5: 387–90. Hé Róng 何容, ed., 1987. Guóyǔ rìbào cídiǎn (dì sānshíliù bǎn) 國語日報辭典 (第 36 版) [Mandarin Daily dictionary, 36th ed.]. Taipei: Guóyǔ Rìbàoshè fùshè chūbǎnbù. Her One-Soon 何萬順. 2009. “Yǔyán yǔ zúqún rèntóng: Cóng Táiwān wàishěng zúqún de mǔyǔ yǔ Táiwān huáyǔ tán qǐ”  語言與族群認同: 從台灣外省族群的 母語與台灣華語談起 [Language and group identity: On Taiwan mainlanders’ mother tongue and Taiwan Mandarin]. Yǔyán jì yǔyánxué 10, no. 2: 375–419. Hsü, Raymond Shih-wen. 1979. The Style of Lu Hsün: Vocabulary and Usage. Hong Kong: University of Hong Kong. Huáng Jìlín 黄继林, ed. 1996a. Yángzhōu fāngyán cídiǎn 扬州方言词典 [Dictionary of the Yángzhōu dialect]. Nanjing: Jiāngsū jiàoyù chūbǎnshè. ———. 1996b. Jīnhuá fāngyán cídiǎn 金华方言词典 [Dictionary of the Jīnhuá dialect]. Nanjing: Jiāngsū jiàoyù chūbǎnshè. Huáng Lìlì 黄丽丽, Zhōu Shùmín 周澍民, and Qián Liánqín 钱莲琴. 1997. GǎngTái yǔcí cídiǎn 港台语词词典 [A dictionary of expressions from Hong Kong and Taiwan]. Hefei: Huángshān shūshè. Huang Shuan-fan 黃宣範. 1993. Yǔyán, shèhuì yu zúqún yìshí: Táiwān yǔyán shèhuìxuè de yánjiū 語言社會與族群意識: 臺灣語言社會學的研究 [Language, society, and ethnic identity: A sociolinguistic study of Taiwan]. Taipei: Crane. Klöter, Henning. 2017. “Táiwān/PRC Divide and the Linguistic Consequences.” In Encyclopedia of Chinese Language and Linguistics, vol. 4, edited by Rint Sybesma et al., 260–63. Brill. doi:10.1163/2210-7363_ecll_com_00000411

Ubiquitous but Unplanned  289 Kubler, Cornelius C. 1985. The Development of Mandarin in Taiwan: A Case Study of Language Contact. Taipei: Student Book. Kuo, Yun-Hsuan. 2005. “New Dialect Formation: The Case of Taiwanese Mandarin.” PhD diss., University of Essex. LaPolla, Randy J. 2001. “The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family.” In Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics, edited by Alexandra Aikhenvald and Robert M. W. Dixon, 225–54. Oxford: Oxford University Press. Li, Charles N., and Sandra A. Thompson. 1981. Mandarin Chinese: A Functional Reference Grammar. Berkeley: University of California Press. Li, David Chen-ching. 1985. “Problems and Trends of Standardization of Mandarin Chinese in Taiwan.” Anthropological Linguistics 27: 122–40. Li Chin-Shi 黎錦熙. 1924. Xīnzhù guóyǔ wénfǎ 新著國語文法 [New grammar of the national language]. Shanghai: Shāngwù yìnshūguǎn. Lǐ Róng 李荣, ed. 1995. Nánjīng fāngyán cídiǎn 南京方言詞典 [Dictionary of the Nanjing dialect]. Nanjing: Jiāngsū jiàoyù chūbǎnshè. Lin, Chin-hui. 2014. Utterance-Final Particles in Taiwan Mandarin: Contact, Context, and Core Functions. Utrecht: LOT. Matras, Yaron. 2000. “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4, no. 4: 505–28. doi:10.1177/ 13670069000040040701 ———, and Jeanette Sakel. 2007. “Investigating the Mechanisms of Pattern Replication in Language Convergence.” Studies in Language 31: 829–65. doi:10.1075/ sl.31.4.05mat MOE (Ministry of Education, Republic of China). 1994. Chóngbiān guóyǔ cídiǎn xiūdìngběn 重編國語辭典修訂本 [A revised dictionary of the national language]. Accessed March 1, 2014 via http://dict.revised.moe.edu.tw. Norman, Jerry. 1988. Chinese. Cambridge: Cambridge University Press. PCO (Population Census Office, Republic of China). 1959. Zhōnghuá Mínguó hùkǒu pǔchá bàogào shū (dì yī juàn: Tái-Mǐn dìqū) 中華民國戶口普查報告書 (第一卷: 台閩地區) [A population and housing census of the Republic of China, vol. 1: Taiwan-Fujian area]. Taipei: Population Census Office of Taiwan Province. Qiú Zhìqún 仇志群, and Fàn Dēngbǎo 范登堡 [Marinus E. van den Berg]. 1994. “Táiwān  yǔyán xiànzhuàng de chūbù yánjiū”  台湾语言现状的初步研究 [A preliminary study on the current linguistic situation of Taiwan]. Zhōngguó yǔwén 4: 254–61. Sandel, Todd L. 2003. “Linguistic Capital in Taiwan: The KMT’s Mandarin Language Policy and its Perceived Impact on Language Practices of Bilingual Mandarin and Tai-gi Speakers.” Language in Society 32: 523–51. :doi:10.1017/ s0047404503324030 Siemund, Peter. 2008. “Language Contact: Constraints and Common Paths of Contact-Induced Language Change.” In Language Contact and Contact Languages, edited by Peter Siemund and Noemi Kintana, 3–11. Amsterdam: John Benjamins. doi:10.1075/hsm.7 Tāng Zhìxiáng 汤志祥. 1999. Táiwān guóyǔ zhōng de fāngyán jiècí 台灣國語中的方 言借詞 [Dialect loanwords in Taiwan Mandarin]. Paper presented at the 6th International Symposium on Min Dialects, June 23–25, The Hong Kong University of Science and Technology. Accessed on July 31, 2014 via www.huayuqiao.org. ———. 2001. Dāngdài Hànyǔ cíyǔ de gòngshí zhuàngkuàng jí qí shànbiàn: Jiǔshí niándài Zhōngguó dàlù, Xiānggǎng, Táiwān Hànyǔ cíyǔ xiànzhuàng yánjiù 当代

290  Chin-hui Lin 汉语词语的共时状況及其嬗变: 90年代中囯大陆、香港、台湾汉语词语现状硏 究 [Synchronics and diachronics of contemporary Chinese words: A study on expressions from mainland China, Hong Kong, and Taiwan in the 1990s]. Shanghai: Fùdàn Dàxué chūbǎnshè. ———. 2002. “Táiwān guóyǔ zhōng de fāngyán jiècí” 台灣國語中的方言借詞 [Dialect loanwords in Taiwan Mandarin]. In Mǐnyǔ yánjiū yǔ qí yǔ zhōubiān fāngyán de guānxi, edited by Pang-hsin Ting and Song-hing Chang, 251–62. Hong Kong: The Chinese University Press. Teng, Shou-hsin. 2002. “Defining Taiwanese Mandarin.” In Education and Society in Plurilingual Contexts, edited by Daniel W. C. So and Gary M. Jones, 230–40. Brussels: VUB Brussels University Press. Tsao Feng-fu 曹逢甫. 2000. “Táishì Rìyǔ yǔ Táiwān guóyǔ: Bǎi nián lái zài Táiwān fāshēng de liǎng ge yǔyán jiēchù de shílì” 台式日語與台灣國語：百年來在 台灣發生的兩個語言接觸的實例 [Taiwanized Japanese and Taiwan Mandarin: Two case studies of language contact during the past hundred years in Taiwan]. Hànxué yánjiū 18: 273–97. Tseng Hsin-I 曾心怡. 2003. “Dāngdài Táiwān guóyǔ de jùfǎ jiégòu” 當代台灣國語 的句法結構 [The syntactic structures of contemporary Taiwanese Mandarin]. MA thesis, National Taiwan Normal University. Tseng, Shu-chuan. 2013.  “Lexical Coverage in Taiwan Mandarin Conversation.” International Journal for Computational Linguistics and Chinese Language Processing 18, no. 1: 1–18. Tung T’ung-he 董同龢. 1974. “Guóyǔ yǔ Běijīnghuà” 國語與北京話 [The national language and the Beijing dialect]. In Dǒng Tónghé xiānsheng yǔyánxué lùnwén xuǎnjí  [A collection of  Tung T’ung-he’s  linguistic essays], edited by Ting PangHsin, 367–69. Taipei: Shíhuò chūbǎnshè. Wáng Jūn 王均. 1995. Dāngdài Zhōngguó de wénzì gǎigé 当代中国的文字改革 [Script reform in contemporary China]. Beijing: Dāngdài Zhōngguó chūbǎnshè. Wu, Ruey-Jiuan Regina. 1997. “Transforming Participation Frameworks in MultiParty Mandarin Conversation: The Use of Discourse Particles and Body Behavior.” Issues in Applied Linguistics 8, no. 2: 97–118. XHC. 2010. Xiàndài Hànyǔ cídiǎn (dì wǔ bǎn) 现代汉语词典 (第5版) [Dictionary of modern Chinese, 5th ed.]. Beijing: Shāngwù yìnshūguǎn. Yuán Jiāhuá袁家骅. 1989. Hànyǔ  fāngyán gáiyào 汉语方言概要 [An outline of Chinese dialects]. Beijing: Wénzì gǎigé chūbǎnshè. Zhōnghuá Rìbào 中華日報 (China Daily News). 1947. “Shèlùn: Guóyǔ tuīxíng yùndòng de shíshī” 社論: 國語推行運動的實施 [Editorial: The implementation of the Mandarin promotion campaign], 4. January 26. Zhōu Chángjí 周長楫. 1996. Mǐnnánhuà de xíngchéng fāzhǎn jí zài Táiwān de chuánbō 閩南話的形成發展及在臺灣的傳播 [The formation of  Southern Mǐn and its spread in Taiwan]. Taipei: Táilì chūbǎnshè.

14 Diverse Language, Diverse Grammars On Quirky Phenomena in Mandarin Jeroen Wiedenhof Abstract This chapter expands on the argument that there are many phenomena in Mandarin which are left undocumented in descriptive accounts and reference grammars due to a number of social, cultural, and educational principles behind conventional grammar writing. The examples analyzed in this chapter are labeled as “quirky” phenomena, i.e. Mandarin usage commonly judged rare or exceptional, which therefore tends to escape linguistic attention. These include Mandarin locative markers, nominal predicates, endearment tones, and the /r/ phoneme. The existence of these “quirky” phenomena suggests that new conceptualizations are needed in the description of Mandarin. From the perspective of language standardization, the inherently dynamic qualities of language call for a critical reflection on the temporary limitations of a linguistic standard.

Introduction: The Study of Diversity in Mandarin The study of Mandarin has to balance two opposite forces, like yīn and yáng: on the one hand, enormous diversity due to a vast number of speakers; and, on the other hand, fixed standards due to its prestige as a national language (in China, Taiwan, and Singapore) and as an educational model (domestically and internationally). This state of harmonious discord in the study of language has parallels in many disciplines, for example in (a) geography and (b) biology. a

b

Consider a geophysical map, depicting mountains, rivers, creeks, and lakes. Language diversity, in this comparison, parallels the irregular features on the ground, and a linguistic account is like a physical map of the terrain. Straight lines are suspicious: they can be useful in a first survey, but must be abandoned as new details come to light. Headsprings will emerge, a river may shift its bed, or an entire mountain might collapse. There is a constant need for new maps. In the domain of biology organisms are, like language, in constant flux. The two domains are intimately related, to the extent that disciplinary distinctions between biology and linguistics are not always

292  Jeroen Wiedenhof viable. From an evolutionary perspective, language can be described as an organism, a symbiont co-existing with primate brains (Wiedenhof 1996, 2008). In other words, biological “parallels” with linguistics may actually constitute mutual identification. Consider the bird which I often spot singing in my city garden. A biologist uses descriptive categories to assess to what extent it belongs to the species of the European blackbird (Turdus merula), to what extent it looks and sounds like a blackbird native to the forest-dune area nearby, and whether the time has come to describe the genetic, morphological, and behavioral differences in these two habitats in terms of separate blackbird species (Ripmeester 2009). Biological diversity can be described in terms of these taxonomic distinctions, but any set of textbook criteria is bound to overlook some unusual individuals which may, in due time, require a new taxon. In linguistics likewise, grammatical categories serve as descriptive tools which, once established, are bound to overlook any number of quirky phenomena. Only if descriptive criteria are set out explicitly will it become possible to assess to what extent, say, the Mandarin expression dāngdì ‘local’ belongs to the adjectival word class (because it denotes a quality), to what extent it behaves differently from the adjective róngyi ‘easy’ (since it cannot form predicates by itself), and if such differences require distinct adjectival classes. Whether in linguistics, in physical geography, or in biology, the choice of analytical categories and procedures will leave different phenomena unaccounted for. Usually the question is not whether, but which phenomena are left unattended.

Descriptive Dilemmas The need to deal with the occurrence of rare or exceptional—in one word, “quirky”—phenomena leads to two dilemmas in any descriptive account. First, the problem of scope. Given a quirky phenomenon in a language, should we describe it, or turn our attention to statistically more representative matters? If we attend to one quirky phenomenon, then what about the next one? The scope of a grammar must be delimited somehow. The second dilemma is structural. Admitting a quirky phenomenon may necessitate a revision of our account of related phenomena, potentially undermining a carefully devised descriptive system. It may signal a break with well- established traditions and lead to terminological impasses. But, as we will see (below in the section on the /r/ phoneme), this principle works both ways, for quirky phenomena may also provide keys to analytic solutions. As a first example, consider the Mandarin neutral tone (輕聲), which is prone to some descriptive negligence. I define the neutral tone as short and unstressed, with no tonal trajectory. This tone is lacking in Xíng’s (2007)

Diverse Language, Diverse Grammars  293 treatment of the Beijing syllable, which must be considered an authoritative Chinese phonological account. I say “authoritative” because the account appears in a 37-page treatise on Hànyǔ (汉语) ‘Chinese’ which has pride of place among a host of monographs in a 2,638-page compendium on the languages of China, undertaken by three editors-in-chief, 20 editors, and 85 authors under the terms of China’s “Tenth Five-Year” National Key Publications Project (“十五”国家重点图书出版规划项目). And I use “Chinese phonological” to indicate that Xíng’s analysis of the Beijing syllable in terms of an initial, a final, and a tone positions his account firmly within the scholarly tradition of yīnyùnxué 音韻學, literally ‘the study of sounds and rhymes’—which, for want of a more convenient English term, I translate as “Chinese phonological studies.” In Xíng’s account, the shēngdiào xìtǒng 声调系统 ‘tone system’ comprises yīnpíng 阴平, yángpíng 阳平, shǎngshēng 上声, and qùshēng 去声 (2007: 111). As I have argued elsewhere, the phonemic status of the neutral tone in Mandarin is beyond doubt in terms of contrastive features (Wiedenhof 2015: 14–16, 55–56). The absence of this tone in Xíng’s paradigm may constitute a theoretical artifact, for which a number of rationales, more or less related, can be hypothesized. a b

c d

e

Prescriptivism: the Mandarin standard language is to represent an ideal norm which, however artificially, allows non-neutral citation tones only. Analytical deadlock: there is no place available for the neutral tone in a descriptive system which has been applied since pre-Middle-Chinese tonal distinctions appeared (píng 平 ‘even’, shǎng 上 ‘rising’, qù 去 ‘going’, and rù 入 ‘checked’).1 Terminological deadlock: the historical term sì shēng 四聲 ‘four tones’ has produced a predilection for a four-way tonal notion. The written-language bias (Linell 2005): confusion of script and language, or prioritization of script over language. Chinese characters rarely have neutral tones in citation, and the description of Chinese languages needs to reflect this state of affairs. Quirkiness: being exceptionally short, the neutral tone is different from all other tones, so it should not belong in any shēngdiào xìtǒng ‘tone system’.

The option of (a) prescriptivism and the concomitant risk of artificiality is mentioned here because it forms a recurrent theme in the history of Mandarin, as Klöter reminds us (this volume). For Xíng’s account, this option can be ruled out because the text mentions none of the usual terms for Mandarin as an educational standard (Pǔtōnghuà, Xiàndài Hànyǔ, etc.). On the contrary, the target of description is identified as the language of Beijing (“北 京话” on p. 111, also “the Beijing variety” in the English summary, p. 144), in which the neutral tone is well attested. Perspectives (b) through (e), and

294  Jeroen Wiedenhof their mutual relationships, may all have contributed to its absence in this phonological account. For the analytical and terminological dilemmas noted in (b) and (c), ample social and historical context is provided by other contributions in the current volume. Often, the sounds of language were regarded as the major focus, or the only focus, in linguistic standardization. As argued and illustrated by Simmons and Söderblom Saarela, the term zhèngyīn 正音 ‘correct pronunciation’ is frequently used in historical works describing or defining a national standard. Lexicon, morphology, and syntax have traditionally been less of a worry, perhaps partly because these domains allowed for a cumulative approach: one could always add more, or mark obsolete items as stylistic variants. Speech sounds were different because they required the establishment of a limited inventory. Accordingly, in the Chinese case, one manifestation of (d) the writtenlanguage bias is the historical tendency, still very influential today, to equate the definition of a standard language with the definition of a standard phonology. Lexical, morphological, and syntactical details are considered as data available in the form of characters, and the phonology of the standard specifies a consensus reached on the pronunciation of these characters. In this approach, characters plus speech sounds together constitute the fabric of a language. The quirkiness argument in (e) illustrates another practical problem of description. Xíng (2007: 111) defines each of the four non-neutral tones exclusively in terms of pitch levels, i.e. 1 (low) to 5 (high) in the conventional paradigm (Chao 1930: 25, 1968: 25–26): yīnpíng 陰平 55, yángpíng 陽平 35, shǎngshēng 上聲 214, and qùshēng 去聲 51. But tone is not pitch. Tones are characterized by a gamut of vocal characteristics including pitch, volume, tempo, and voice quality. Note that Xíng does use non-pitch criteria in his treatment of Cantonese (2007: 113), where tempo, in the form of contrastive length, is brought well inside the domain of tone. Contrastive length is precisely what characterizes the neutral tone among the other Mandarin tones. And Chinese phonological studies do accommodate features other than pitch to define tone, e.g. segmental material such as the unreleased-stop codas [p˺, t˺, k˺] for the category of rùshēng 入聲 ‘checked tones’.

Four Quirky Bits of Grammar The following examples illustrate rare or exceptional phenomena from different areas of linguistic description: the lexicon, syntax, morphology, and phonology. They are presented here to review and reconsider their treatment in the Mandarin grammatical record. Unless indicated otherwise, spoken Mandarin data were collected by the author over the years and in the context of different projects; for some details see Wiedenhof (1995: 1–2, 2015: xxi).

Diverse Language, Diverse Grammars  295 Lexical Localizers Among grammatical domains, the lexicon is arguably the most impervious to harm by negligence, for a vocabulary list can be shrunk or expanded with relative flexibility. Lexical publications may thrive despite omissions of all kinds, e.g. in the documentation of blasphemy, sacred words, clitic pronouns, and proper nouns. To illustrate, in my native Dutch, dictionaries typically neglect proper names. This creates a lack of documentation of simple linguistic facts, e.g. that the name of the town of Gorinchem, whose spelling would suggest a non-existing (though no doubt historical) reading *[ɣɔʀɪŋxəm], happens to be pronounced as [xɔɹkəm]. But lexicographical norms are not static either. In the Dutch context, a recent development is the acceptance of productive bound morphemes as separate entries, e.g. the suffix heid for situational or qualitative nominalization (blij+heid ‘happiness’, zeker+heid ‘certain-ty’). Still, even in the lexicon, some challenges to the integrity of a grammatical system will emerge, e.g. for compounds. Thus, Mandarin zhèrher ‘here’ can be described as occupying a position at the lower end of a formality scale for proximal demonstrative compounds, ranging from the more casual (zhèrher ~ zhèr ~ zhèrlǐ) to the more formal (zhèlǐ ~ zhèibiār ~ zhèibiān); and similarly for the distal counterpart nèrher ‘there’ (Wiedenhof 2015: 110–12). Neglecting one or more of these lexical items may seem merely to restrict this formality scale in terms of scope or gradient accuracy. But there is the rich system of Mandarin locative suffixation to consider: apart from her, r, rlǐ, lǐ, biār, and biān, we have suffixes such as miàn ~ miar ‘face’ and kuàr ~ kuer ‘chunk’, each of which combines with proximate zhè and distal nè to signal additional information such as shape or relative position of a location. The locative system is poorly documented, even in phonological terms (see below on patterns with and without the /r/ phoneme), but it is clear that present usage differs substantially from Chao’s detailed description (1968: 519–33, 620–27). Careful synchronic documentation is a prerequisite for diachronic insights, which by itself should clinch the argument in favor of documenting the quirky her morpheme. As a bonus, this will allow us to consider the plausibility of a common root for this her and the locative suffix xià ~ xia ‘down’ (also [hjə ·|], Chao’s [1968: 623] Gwoyeu Romatzyh-based “.hie”), as suggested by Middle Chinese reconstructions for modern Mandarin xià ~ xia ‘down’ such as haeH (i.e. hae with a qù tone) ‘descend’ and/or haeX (with a rù tone) ‘down’ (Baxter and Sagart 2014: 366–67). The Syntax of Nominal Predicates Mandarin nouns can be used to form nominal predicates, as in Tā yùndòngyuán. ‘He’s an athlete’. Complex noun phrases can likewise be made to serve as predicates. Examples 1–3 represent casual spoken registers. The

296  Jeroen Wiedenhof written language, represented in examples 4–6, makes use of the copula shi 是 ‘be’ in comparable cases, and formal spoken usage follows suit. Example 3 illustrates the predicative use of a quantifying noun phrase. The predicative use of such noun phrases is special in that it is fully acceptable in casual as well as formal spoken registers; and likewise in written Chinese, as in example 6. For the copula shi ‘be’, I prefer an analysis with a gradient scale, at least in terms of form: shì ~ shi ~ ri [ʐ ·|] ~ r [ɻ ] ~ ø ‘be’ (Wiedenhof 2015: 173). Here the symbol ø does not transcribe a zero morpheme: it represents the existence of the syntactic expression type of examples 1–3, in which the copular predicative meaning ‘be’ is correlated with both word order (final position) and prosodic order (no reversing strategies). This is, in other words, the nominal analogue to the predicative syntactic pattern for Mandarin adjectives: Méigui hóng. ‘Roses are red’. It is tempting to imagine a pathway of grammaticalization, with the fully syntactic pattern serving as a point of completion in a trajectory of copula attrition and deletion shì > shi > ri > r > ø. I will not go into this matter here for lack of evidence and uncertainty on how to interpret available data. 1 2 3 4

5

6

Zhèi [gāng mǎi de dōngxi]. this just.now buy sub thing ‘These are the things we bought just now.’ Zhèi ròu [gǒu chī de] a? this meat dog eat sub ec ‘So this meat is what the dogs are eating?’ Tā [sānshi liù suì]. 3 thirty six year.of.age ‘She is thirty-six.’ 這 是 剛 買 的 東西。 Zhèi shi gāng mǎi de dōngxi. sub thing this be just.now buy ‘These are the things we bought just now.’ 這 個 肉 是 狗 吃 的 啊 ? Zhèi ge ròu shi gǒu chī de a? this item meat be dog eat sub ec ‘So this meat is what the dogs are eating?’ 她 三十 六 歲。 Tā sānshi liù suì. 3.sg.feM thirty six year.of.age ‘She is thirty-six.’

When we check the relative negligence of nominal predicate syntax in grammars of Mandarin after Chao (1968: 90–94) with the list of five suspected causes in the earlier section on descriptive dilemmas, the first one looms large: prescriptivism, in the sense that sentences 4 and 5 are considered to be more proper and civilized than 1 and 2. This is consistent with the fact that examples such as 1 and 2 are avoided in most textbooks of Mandarin.

Diverse Language, Diverse Grammars  297 But grammars are not textbooks, the relevant difference being that of description versus prescription. Zhōu’s (1998: 30) impression that nominal predicates are infrequent in spoken Beijing Mandarin seems extraordinary. For non-quantifying noun phrases the predicative pattern may be quirky, but it is nonetheless ubiquitous. The Morphology of the Endearment Tones Mandarin endearment tones occur in the type of language which adults (especially mothers: VanDam, De Palma, and Strong 2015) typically use to address babies, toddlers, and young children. This mode of speech is often accompanied by slow tempos, high volumes, and high-pitched intonation. Endearment tones are likewise common in child language, i.e. the language spoken by young children. 7a 7b 7c 7d

gǒu ‘dog’ > gǒugóu ‘nice doggy’ shūshu ~ shúshu ‘uncle’ (father’s younger brother) > shǔshú ‘dear uncle’ Tián ‘Tián’ (personal name) > Tiǎntián ‘Tián dear’, ‘my dear Tián’ tuō xié ‘take off one’s shoes’ > tuō xiěxié ‘take your shoes off, dear’

The segmental material in this pattern is usually a reduplicated syllable, as illustrated in example 7 (Wiedenhof 2015: 321–22). Endearment tones are a tonal pair consisting of a regular sandhi sequence of a third tone followed by a second tone [ ] substituting for the original tones. Endearment-tone morphology contributes affectionate connotations to the meaning of the original expression. The pattern is used productively. Lexicalized examples include the Taiwan Mandarin name for the clown mascot of the McDonald’s fast-food chain, Màidāngláo Shǔshú ‘Ronald McDonald’. Endearment tones seldom affect non-reduplicated forms, but ǎyí (< āyí ‘aunt’, mother’s sister) is used for ‘auntie’. There is some rare literary evidence for endearment tones thanks to dedicated character spellings, for instance gěgé 葛阁 ‘dear elder brother’, dǐdí 底笛 ‘dear little brother’ (Lung 1994: 54), and jǐjí 擠急 ‘willy’ (ibid., 174). Endearment tones must be a relatively new phenomenon in Mandarin. Chao does not mention it in his grammar (1968) or, perhaps more pertinently, in his extensive account of the idiolect of one of his young grandchildren (1951). Judging from Chao’s track record, he would have documented it if it had existed in his time. In Duanmu’s phonological study, endearment tones are described as a “fixed L-LH pattern” in Taiwan Mandarin (2007: 248). To be sure, endearment tones are frequent in Taiwan Mandarin, but they have a much wider distribution in the Mandarin area, including Beijing and Shenyang. Accounts are patchy, and the origins, spread, and current distribution of endearment tones need to be inventoried more extensively. Semantic details such as taboo connotations and transfers to adult speech likewise deserve further study.

298  Jeroen Wiedenhof Patterns with and without the /r/ phoneme The last item in this short inventory of neglected quirkiness is the Mandarin /r/ phoneme, which suffers some disregard in the phonological representation of Mandarin. I will endeavor to show how it may serve to solve an analytic dilemma. Terminologically, I distinguish between “r coda” to designate the phoneme /r/ in rhotacisms such as dērdēr ‘whoosh-whoosh’ (for the sound of skating); and “r suffix” for the morpheme r in compounds, e.g. guěr ‘devil’ < guǐ ‘ghost’ + r. Use of the r suffix implies the presence of an r coda, but the reverse need not be true. For the sake of convenience, I continue to quote from Xíng’s (2007) account, but identical and similar inventories of Mandarin speech sounds abound. As mentioned, Xíng’s syllable is a composite structure made up of an initial, a final, and a tone. There are 22 initials (including zero), chiefly consonants; four jièyīn 介音 ‘medials’, identified as glides: zero, i, u, and y; and five yùnwěi 韵尾 ‘rhyme codas’, distinguished as kāi 开 ‘open’, vocalic i and u, and consonantal n and ŋ. The medial and rhyme coda together constitute a yùnmǔ 韵母 ‘final’, as illustrated in Table 14.1. This table features a juǎnshé yuányīn 卷舌元音 ‘retroflex vowel’ which is transcribed as “ər.” In terms of arrangement, it is part of the final but comes after the coda. In my view, Xíng cannot have it both ways: either the retroflex vowel is part of the coda, or it belongs to a separate syllable. This analysis seems to rely unduly on the basic units of the Chinese character script (“written language bias,” point (e) in the earlier section on descriptive dilemmas). Indeed, Xíng’s explanation with regard to Table 14.1 reads: “There is a retroflex process for characters such as ér 儿 ‘child’, -ěr- 耳 ‘ear’ and èr 二 ‘two’, which acoustically resemble a compound vowel.” (“儿” 丶“耳”丶“二”等字的韵母有一个卷舌过程, 听起来像是一个复合元音。) It is unclear whether everyday expressions such as tiār ‘the weather’, bīnggùr ‘popsicle’, and zhèrher ‘here’ are regarded as reflecting the same process. Table 14.1 A Yùnmǔ xìtǒng ‘System of finals’ in Beijing Mandarin System of finals (韵母系统) Medials (介音)

Rhyme codas (韵尾) Open codas (开尾)

-i codas (-i 尾)

-u codas (-u 尾)

-n codas (-n 尾)

-ŋ codas (-ŋ 尾)

none i u y

ɿʅaə i ia ie u ua uo y ye

ai əi

au ou iau iou

an ən ian in uan uən yan yn

aŋ əŋ uŋ iaŋ iŋ yŋ uaŋ uəŋ

uai uei

Source: Bilingual representation by author of Xíng (2007: 111).

Retroflex vowel (卷舌元音) ər

Diverse Language, Diverse Grammars  299 My analysis is phonemic, defining the Mandarin syllable as a tonebearing unit (2015: 31, 55–67). Some contrastive examples are listed in Tables 14.2 and 14.3. The /r/ phoneme can occupy different positions within the syllable. There is an allophone [ʐ ] which occurs as an initial, and an allophone [ɻ] which occurs as a coda, not after a coda. There is no place for an extraneous retroflex vowel “ər” in this scheme, but retroflex consonants do occur in two varieties, distinguished by nasality.2 In Table 14.3, the distinction in (a) and (b) is discussed by Chao (1968: 653–54); as far as I can tell, it is preserved intact in Beijing Mandarin, apart from some current use of a variant with final rounding, shémo ‘what’, e.g. in pop songs. As always, there is no predicting how far acceptance of this variant will go. For the common variant shéme ‘what’ in (a), Chao records an unusual tone-sandhi pattern (high-pitch instead of middle-pitch neutral tone) which I prefer to relegate to the lexicon until similar cases are found to establish a recurring pattern. The lexicon, in other words, will document shéme ‘what’ with a high-pitched neutral tone, which is otherwise restricted to positions following a shǎng tone, here occurring after a yángpíng tone. This brings us back to the theme of quirkiness. In (a) and (b), does the contrast between schwa and non-schwa vowels deserve our attention? Judging by the Pīnyīn transcription, the answer should be no, for it does not distinguish gè [ɡə ] ‘item’ (Bié ná měi yi gè. ‘Do not pick up each one’.) from gè Table 14.2 Some contrasts with the /r/ phoneme (Pīnyīn r- and -r) Phonology

Phonetics

Pīnyīn

Meaning

/au1/ /rau4/ /aur3/ /raur4/ /raŋr2/

[ɑɔ ̜ [ʐɑɔ ̜ [ɑɔɻ ̜ [ʐɑɔɻ ̜ [ʐɑɻ̃ ̃

āo rào ǎor ràor rángr

‘boil, simmer, stew’ ‘revolve, go around’ ‘lined jacket’ ‘coil, winding’ ‘flesh of a fruit’

]

] ]

]

]

Source: Wiedenhof (2015: 51, 58).

Table 14.3 Some contrasts between /ə/ and /əa/ (both transcribed as Pīnyīn e)

a b c d e

Phonology

Phonetics

Pīnyīn

Meaning

/šə2mə0/ /šəa2mə0/ /šəar2/ /šər2/ /šəŋr2/

[ʂə mə ] [ʂɤʌ mə ] [ʂɤʌɻ ] [ʂəɻ ] [ʂəɻ̃ ̃ ]

shéme shé me shé’r shér shéngr

‘what’ ‘a snake?’ ‘tongue’ ‘fodder’ ‘string, rope’

Source: Wiedenhof (2015: 62–65, 419).

300  Jeroen Wiedenhof [ɡɤʌ ] ‘chromium’. Pīnyīn, in other words, ignores a phonological contrast between /gə4/ and /gəa4/ (Wiedenhof 2015: 64–65). Still, the number of cases where a Pīnyīn e final represents schwa appears to be quite small. There are some other nouns, such as bòhe [pɔ χə ] ‘peppermint’ and [kɤʌ pə ] ‘arm’, which may be transcribed as either gēbei or gēbo.3 Regardless of tone, Beijing speakers have this schwa vowel e and its rounded counterpart o as common allophones after labial initials, e.g. pè le ~ pò le ‘broken down’. Following other initials, consider cases such as zhè ‘this’ and nè ‘that’, where the final is pronounced as [ə ], not [ɤʌ ]. Pronouncing zhè ‘this’ as [ʈ͡ ʂɤʌ ] would, in other words, be a hypercorrection in Beijing Mandarin. Other high-frequency expression are the function words de [tə ·|] ‘sub’, le [lə ·|] ‘pf’, ne [nə ·|] ‘rlv’, and zhe [ʈ͡ ʂə ·|] ‘dur’. Many of these instances present not one but two counts of quirkiness, illustrating a neutral tone and a schwa vowel to boot. But in a grammatical description, must two counts of quirkiness mean out? The relevance of this vocalic contrast will be evident as soon as we consider rhotacisms. Table 14.4 is the rhotacized section of a larger table (Wiedenhof 2015: 66) representing the yùnmǔ ‘finals’ of Beijing Mandarin in terms Table 14.4 Rhotacized finals in Beijing Mandarin

♦ ♦

♦ ♦

♦ ♦ ♦ ♦

Phonology

Pīnyīn

/ar/ /aur/ /aŋr/ /ər/ /əar/ /əŋr/ /iar/ /iaur/ /iaŋr/ /iər/ /iəar/ /iur/ /iŋr/ /yŋr/ /əur/ /uŋr/ /ur/ /uar/ /uaŋr/ /uər/ /uəar/ /yar/ /yər/ /yəar/

-ar -aor -angr -er -e’r, -or -engr -iar -iaor -iangr -ier -ie’r -iur -ingr -iongr -our -ongr, wengr -ur -uar -uangr -uer -uor -üar -üer -üe’r

Source: Wiedenhof (2015: 66).

Diverse Language, Diverse Grammars  301 Table 14.5 Some contrasts with rhotacized finals Phonology

Phonetics

Pīnyīn

Meaning

Morphology

/gər1/ /gəar1/ /iər2/

[kəɻ ] [kɤʌɻ ] [jəɻ ]

gēr gē’r yér

‘bottom, background’ ‘ditty’ ‘aunt’ (mother’s sister)

/iəar2/

[jɛɻ

yé’r

/guər3/ /guəar3/ /yər/ /yəar/

[kwəɻ ] [kwɔəɻ ] [ɥyəɻ ] [ɥyɛɻ ]

‘senior male together with younger relative’ ‘a tumble’ ‘fruit’ ‘piece of jade’ ‘month in the lunar calendar’

< gēn ‘root’ < gē ‘song’ < yí ‘aunt’ (mother’s sister) < yé ‘old man’

]

guěr guǒr yuèr yuè’r

< gǔn ‘to roll’ < guǒ ‘fruit’ < yù ‘jade’ < yuè ‘month’

Source: Wiedenhof (2015: 49, 52, 65, 98).

of a phonemic analysis. The schwa vowel [ə] is analyzed as /ə/, the diphthong [ɤʌ] as /əa/; the diamond symbol ♦ marks where this phonemic contrast is relevant. Note that this distinction applies to eight, or one-third, of the 24 rhotacized finals. As illustrated in Table 14.5, these minimal pairs occur not just with the neutral tone but with the other tones as well. For each of these items, the r coda represents the r suffix, and the preceding morpheme is listed in the final column. To sum up, the problematic phonemic status of the schwa vowel in Mandarin turns out to be a by-product of a widespread tendency in grammatical accounts to disregard, for various non-linguistic reasons, the occurrence of the /r/ phoneme in syllable-final position. Recognizing /r/ as a full-fledged syllable coda enables us to describe the difference between a schwa vowel /ə/ and a diphthong /əa/ as a system-wide phonemic contrast in present-day Beijing Mandarin.

Conclusions: Preparing the Way for Linguists of the Future The phenomena described here are examples of attested features of Mandarin which tend to escape linguistic attention. They form, in Chin-hui Lin’s words (this volume), phenomena whose “very existence would remain undiscovered if we look at prescriptive accounts of the standard language only.” In lexical accounts, the rich locative system is poorly represented, leaving spoken forms such as zhèrlǐ, zhèrher, and zhèikuer for proximal ‘here’ scantily described or differentiated. In syntax, despite the high frequency of nominal predicates in spoken Mandarin, descriptions tend to favor literary equivalents requiring the use of the copula shi ‘be’. In Mandarin morphology, endearment tones represent a productive process which is all but absent from the grammatical record. And in phonology, lack of attention to the ubiquitous /r/ phoneme may lead to oddly skewed accounts of the Mandarin schwa vowel, and of the Mandarin syllable.

302  Jeroen Wiedenhof This state of affairs is induced by a number of conventions, some of which have been listed earlier, in the section on descriptive dilemmas. Linguistically speaking, none of these considerations should stop us from making our descriptions inclusive. Today’s exception might become tomorrow’s norm, and an accurate description now will keep the linguistic record maximally informative across time, helping future generations of linguists trace earlier developments. Inclusive documentation of linguistic diversity also forces us to continually rethink the adequacy of our grammatical models. The study of Chinese has come a long way, inspired by philosophers such as Gottfried Wilhelm Leibniz (1646–1716), naval merchants across the coasts of Asia, missionaries translating the Bible, diplomatic and military ranks, as well as software engineers. The developmental history of Chinese linguistics would be worth a careful meta-descriptive comparison. Even when restricted to Mandarin, sinophone studies stand to gain by a broad historical overview of the assumptions, requirements, and approaches presented by grammatical sketches and accounts. In my view, codified standard languages need to be included in such a survey. In the spirit of this volume, I would like to express my gratitude for a number of inspiring peer reviews for the current essay. One of these noted that my contribution “destabilizes the very notion of Standard Mandarin from a linguistic perspective.” This must be so, I think, because notions of the static and the constant are poor matches for the organic and evolutionary qualities of language. The idea of a standard language accordingly has different roles to play outside and inside linguistics. In language education and language policies, for instance, a standard language provides a practical basis for textbook benchmarks and administrative guidelines. But in strictly linguistic terms, a standard language is a randomly timed snapshot of the phonology, morphology, syntax, and lexicon of a living language, represented with a degree of formalization, often in inverse proportion to available resources. Another review characterizes this contribution as “nothing less than a call for a paradigm shift in the linguistic description of Mandarin.” The term “call” presupposes a mission. I am not good at missions, but if I had to declare one here, it would be to embrace the diversity found within Mandarin. Or perhaps, to stop pretending that Mandarin is a well-documented language. The latter is phrased in a rather negative tone, which would make it a nasty kind of mission. As I said: I am not good at missions. My concern as a linguist is with the data, their analysis, and their reporting. At each step, selections are made, depending on a range of historical, social, cultural, and educational contexts. Different approaches leave different type of materials undocumented, and being aware of these limitations helps. Linguistic expertise will remain an indispensable and powerful factor in the creation and codification of standard languages. After all, the successful introduction of any standard language depends on years of careful commitment to linguistic details, not to mention interconnections with script

Diverse Language, Diverse Grammars  303 culture. Educational and political demands for a standard language, in short, have often provided, and continue to provide, a professional lifeline for linguists. In one popular academic model, the linguistic profession may be regarded as a provider, and a standard language as a deliverable. In this scenario, the provision of fair product information by the field is long overdue. Linguists need to consider introducing expiration dates for standard languages. If this seems far-fetched, the intrinsically biological nature of language (see introduction to this chapter) will justify drawing inspiration from “Best-Before” labeling systems for nutriments and other organic produce. Being explicit about the temporary limitations of a linguistic standard, far from representing an admission of defeat, will actually make linguistic work more accountable because it focuses attention on the inherently dynamic qualities of language. A major distinction between codifying a standard language and publishing a grammar is that they represent language more prescriptively or descriptively, respectively. But there are also similarities. As argued above, a standard language is little more than a formalized description of a living language. Hence the same “snapshot” argument is applicable to any linguistic description, and to all of our grammars. Much like a standard language, a grammatical description reflects a very temporary state of affairs. Strictly speaking, it is outdated the very minute it is published, and will remain in constant need for updates and revisions under new circumstances. This is why there is a continuous need for new grammatical accounts—especially those which keep the ears wide open for quirky phenomena.

Abbreviations and Conventions text (a) sign; (b) phonological form /text/ phonological form -text, text-, -text- bound morpheme ‘text’ meaning [text] phonetic realization + morpheme boundary 1, 2… (a) person; (b) pitch level dur durative aspect ec expected continuation feM female gender pf perfective aspect rlv relevance sg singular number sub subordination

Phonetic transcription is in the International Phonetic Alphabet. Chinese characters are in traditional script unless quoted from sources in simplified characters. Pīnyīn transcription represents spoken usage, not character readings (Linguistic Pinyin, Wiedenhof 2015: 414–18). In particular, (a) in

304  Jeroen Wiedenhof the absence of audible differences, transcriptions are kept identical, e.g. for the r suffixed compounds huár (< huá) ‘slippy’, huár (< huái) ‘womb’, and huár (< huán) ‘little circle’; (b) nasal rhotacisms are spelled as ngr, e.g. huángr ‘egg yolk’; (c) following Hockett’s (1951: 45) spelling in the Yale transcription and Chao’s (1968: 51) convention in Gwoyeu Romatzyh, the letters e and r are separated by an apostrophe to indicate rhotacized non-schwa vowels in e’r, ie’r, and üe’r, and written together for rhotacized schwa in er, ier, and üer.

Notes 1 Rùshēng 入聲 is often rendered as entering tone, but this English label seems to represent a mistranslation. Rù 入 may mean ‘enter’, i.e. ‘get inside’, but is used here to mean ‘keep inside, hold back, check’. Since the term denotes an airstream “checked” by the unreleased final stop characterizing this tone, I prefer checked tone in English. 2 Ladefoged and Maddieson (1996: 153) object to the term retroflex on articulatory grounds in their description of sh- as a (laminal) flat post-alveolar sibilant. For the phonological account, I am using the traditional name. 3 The Pīnyīn spelling be is somehow avoided in modern dictionaries (e.g. XHC 2003, which allows me, at least for the neutral tone, but avoids be, fe, and pe). Dictionary spellings for [kɤʌ pə ] ‘arm’ include gēbei and gēbo.

References Baxter, William H., and Laurent Sagart. 2014. Old Chinese: A New Reconstruction. Oxford: Oxford University Press. doi:10.1093/acprof:oso/9780199945375.003.0002 Chao, Yuen Ren. 1930. “ə sistim əv ‘toun letəz’” [A system of “tone-letters”]. Le maître phonétique, third series, no. 30: 24–27. ———. 1951. “The Cantian Idiolect: An Analysis of the Chinese Spoken by a Twenty-Eight-Month-Old Child.” Semitic and Oriental Studies [University of California Publications in Semitic Philology] 11: 27–44. ———. 1968. A Grammar of Spoken Chinese. Berkeley: University of California Press. Duanmu, San. 2007. The Phonology of Standard Chinese. Originally published in 2000. 2nd ed. Oxford: Oxford University Press. Hockett, Charles F. 1951. Progressive Exercises in Chinese Pronunciation. New Haven, CT: Far Eastern Publications. Ladefoged, Peter, and Ian Maddieson. 1996. The Sounds of the World’s Languages. Malden, MA: Blackwell. Linell, Per. 2005. The Written Language Bias in Linguistics: Its Nature, Origins and Transformations. London: Routledge. doi:10.4324/9780203342763 Lung Ying-tai 龍應台. 1994. Háizi, nǐ mànmàn lái 孩子你慢慢來 [Easy now, baby]. Taipei: Huánguàn wénhuà chūbǎn yǒuxiàn gōngsī. Ripmeester, Erwin Adriaan Pieter. 2009. “Song and the City: A Comparison between Urban and Forest Blackbirds.” PhD diss., Leiden University. VanDam, Mark, Paul De Palma, and William E. Strong. 2015. “Fathers’ Use of Fundamental Frequency in Motherese.” Poster presented at the 169th Meeting of the Acoustical Society of America, Pittsburgh. May 18–22. Accessed on May 12, 2018 via http://www.vandammark.com.

Diverse Language, Diverse Grammars  305 Wiedenhof, Jeroen. 1995. Meaning and Syntax in Spoken Mandarin. Leiden: CNWS Publications. ———. 1996. “Nexus and the Birth of Syntax.” Acta Linguistica Hafniensia 28, no. 1: 139–50. doi:10.1080/03740463.1996.10416068 ———. 2008. “Language, Brains and the Syntactic Revolution.” In Evidence and Counter-Evidence: Essays in Honour of Frederik Kortlandt, Volume 2: General Linguistics, edited by Alexander Lubotsky, Jos Schaeken, and Jeroen Wiedenhof, 415–28. Amsterdam: Rodopi. doi:10.1163/9789401206365_026 ———. 2015. A Grammar of Mandarin. Amsterdam: John Benjamins. doi:10.1075/z.197 XHC. 2003. Xiàndài Hànyǔ cídiǎn 现代汉语词典 [Dictionary of modern Chinese]. Reprint of the Enlarged and Revised Third Edition of 2002. Beijing: Shāngwù yìnshūguǎn. Xíng Gōngwǎn 邢公畹. 2007. “Hànyǔ” 汉语 [Chinese]. In Zhōngguó de yǔyán, edited by Sūn Hóngkāi, Hú Zēngyì, and Huáng Xíng, 108–44. Beijing: Shāngwù yìnshūguǎn. Zhōu Yìmín 周一民. 1998. Běijīng kóuyǔ yúfǎ (cífǎ juàn) 北京口语语法（词法卷） [A grammar of spoken Pekingese: Morphology]. Beijing: Yǔwén chūbǎnshè.

Index

Note: Page numbers followed by “n” denote endnotes. ABC see American-born Chinese access policy 128, 132–33 alphabetic writing 27, 60, 64, 104 American Baptists 64 American-born Chinese 238, 242–43 Amoy 62–65; see also Xiamen ancestry 44, 48, 91, 175, 187, 194 Austronesian languages 80, 123–26, 170 Baba Malay 88, 173–74, 206 báihuàwén 104, 109, 110; see also vernacular(s) Bannermen 30–31, 39, 43, 46–47, 53–54; as authorities of Mandarin 46 Barclay, Thomas 64 Basic Law see Hong Kong, Macao Bazaar Malay 82, 88–89, 206 běihuà see Mandarin, Northern Beijing 54, 93, 279; as capital of the Ming dynasty 16; and Guānhuà 13, 19; language of 293; Manchu spoken in 52; as pronunciation standard 5, 6, 15, 26, 28, 31, 101, 107, 112–13, 115 Beijing dialect 13, 19, 21–24, 29–30, 52, 278–79 Beijing Mandarin 29, 111, 280–81, 283, 297–301 Běijīnghuà see Beijing dialect Běiyīn 27–28 Belt and Road initiative 154 Bible translation 49, 61–68, 74–76 bilingual education in Singapore 170, 172, 177, 208, 211 bilingual language policy in Singapore 179, 213 bilingualism: Chinese and Portuguese in Macao 146; during the Qing dynasty 44; in Singapore 164, 170, 176, 178,

205, 208–9, 216; in Taiwan 125, 278; see also dialect bilingualism bilinguals 252; in Hong Kong and Macao 145, 147; Manchu–Chinese 51; in Singapore 219; in Taiwan 277 biliteracy and trilingualism 142, 148 Blommaert, J. 199, 231, 234, 242–43 blue-green Mandarin see Mandarin, blue-green Bolton, K. 81–84, 90, 92–93, 145, 147 borrowing 83, 88, 207, 250, 252–53; into Cantonese from English 142, 147; into Colloquial Singapore English 207, 252–53; lexical 142, 147; into Mandarin 230, 254–57, 259, 263, 267–68; between Mandarin, Taiwanese, and Hakka 266; and morphological integration 254–55; of particles in Sinitic 275; and phonological integration 255–57; into Portuguese 83; into Singapore English 88 Buddhist missionaries 74–75 Canton see Guangdong, Guangzhou Cantonese 45, 88, 110, 142, 144, 151, 275; Bible translation 61; in Guangzhou 185, 192, 196; in Hong Kong 147, 155; and identity 158, 196; language activists 196; language attitudes in Hong Kong 152–53; language of government in Hong Kong and Macao 149; learned by speakers of other Chinese dialects 149; as lingua franca 149; in Macao 91, 146–47, 151, 154, 159; as medium of instruction in Hong Kong 89, 151–52, 155–56; as medium of instruction in Macao 151–52; number of speakers

308 Index in Hong Kong and Macao 89–90, 148–49; pop culture 192; proficiency 197, 199; protection 113; protests against suppression in Guangzhou 113; in Singapore 166, 167, 169, 174, 206–7, 209, 216; symbolic value 157; vernacular literacy in 117; written 27, 29, 75, 148, 152; see also Yuè dialects Cattaneo, L. 20 CCP see China Coast Pidgin, Chinese Communist Party Chao, Y. R. 106, 284, 295–97, 299 Chaoshanese see Teochew (Chaoshan) dialect Chaozhou 194 China Coast English see Chinese Pidgin English China Coast Pidgin 145 Chinese characters 27, 44, 50, 170, 294; Chinese pride towards 74; simplified 112; traditional 114, 151; used for the transcription of Manchu 52; and written vernaculars 61, 64, 69–70, 152 Chinese Communist Party 102, 111 Chinese Medium-of-Instruction 151 Chinese phonological studies 293–94 Chinese Pidgin English 83–84, 86; emergence in the nineteenth century 145 Chinese Repository 21 Chui, F. 143 civil service examination system 41, 49 clan (jiāzú) 194, 199 Classical Chinese 3, 13, 39; Chinese esteem of 69–71; difficulty according to Western missionaries 67, 73–74; official language in Qing empire 40–41, 48; replacement with written vernacular 104; source for translations into Manchu 50–53, 60 CMI see Chinese Medium-of-Instruction code-mixing 147, 152; code-switching 232–34; conventionalized 250–68; between English and Chinese in Taiwan 231, 238; between Mandarin and Taiwanese 233; negative sentiment toward 231, 233–34, 238; social meanings of 234, 237, 245; universal grammatical framework for 252 Colloquial Singapore English 88, 207–8; attitudes towards 222; as a national solidarity code 221; official rejection of 212–13; popularity of 223; see also Singlish

colonial language policy: in Hong Kong and Macao 145–46, 153–56; in Singapore 82, 167–68, 211 Common Chinese 17 Confucian classics 40, 63; Manchu translation of 49, 51 correct pronunciation 16, 105, 108, 294; academies for 45–46; and Mandarin 41 Court audience: and checking of pronunciation 43, 47–48; and language of Cantonese and Fujianese officials 44–45; and language of Hanlin Bachelors 43, 45; use of Manchu and Mandarin 47, 54 CS see code-switching CSE see Colloquial Singapore English de-Sinicization 123, 189 Democratic Progressive Party 122, 126–27 dialect bilingualism 157 Dialect Literature Movement 75–76 dialect(s) 14, 31, 80, 88, 101, 104–5, 110, 113, 149, 151, 188, 195, 197–99; crisis 117; evolution 15; and fāngyán 189; interactions across 16; northern group 46; and regional identities 201; Romanization of 27; scripts 110; in Singapore 166, 169–70, 174; status of 185, 190; in Taiwan 127; see also Beijing dialect; Cantonese; Holo; multidialectism; Nanjing dialect; Mandarin; Mǐn; Southern Mǐn; Taiwanese language diaspora 166, 170, 193 Dictionary of National Pronunciation 105–6 diglossia 43, 60, 76, 106; in late imperial China 43; in Singapore 168; in Taiwan 278 domains 135–36, 215; education 115, 157, 264; functional 103, 108; government and law 149, 157; informal 250, 266, 268; media 115, 157; private 116, 127, 192, 277; public 87, 112, 192; work 264 Douglas, C. 64–65 DPP see Democratic Progressive Party Duffus, W. 64–65 Eastern Mandarin (Jiāng-Huái Mandarin): influence on Taiwan Mandarin 275, 279, 282

Index  309 ecologies see multilingual ecologies Edkins, J. 19–23 Elman, B. 13–14 emergence principle 234 endearment tones 297 English Bible Society 65 English(es): in blog posts 235–45; in China 83–84, 92–93; in education in Hong Kong 153–55; in education in Macao 148, 155–56; in education in Singapore 169–71, 177, 207, 209–10; in education in Taiwan 123, 131–34, 230; as home language in Singapore 173–74, 178, 213–15; in Hong Kong 82, 89–92, 147; insertions in Taiwan Mandarin 253–63; Interactions across 81, 84–86, 93–94; planning in Taiwan 126, 130–31; proficiency in Singapore 218–20; rise and dominance in Singapore 163–66, 172, 178, 213; in Singapore 86–89; in the sinophone world 81–84; speaker’s attitudes in Hong Kong 152–53; speaker’s attitudes in Macao 153–54; speaker’s attitudes in Singapore 221–22; speaker’s attitudes in Taiwan 136–37, 263–66; status in Hong Kong 89–90, 149; status in Singapore 167–69, 207–10; status in Taiwan 123, 230; use in Taiwan 135–36, 231–35, 251–53; see also Colloquial Singapore English; Speak Good English Movement ethnic Chinese 1–2; in Singapore 166, 173–76 ethnic groups 193; in Macao 91; in Singapore 166, 169, 214, 218, 222 ethnic minorities 103; in Hong Kong 150 feature pool 85 field observation 253 Formosa 63–64; see also Taiwan four tones 20–22, 27–29; absence in Manchu script 50 four-corner system 28–29 Fujian 64, 66, 206, 277; and promotion of Mandarin under Yōngzhèng 44–46 Fuzhou 82, 280 Fuzhounese 61 Gibson, J. C. 64–65, 67–71, 74 grammatical categories 292 Guangdong 21, 44–46, 62, 144, 149, 173, 193, 197, 206; as Cantonese speaking

199–200; and promotion of Mandarin under Yōngzhèng 44–46 Guangzhou 81–84, 93, 114, 143, 185; and Cantonese language 149, 157, 192, 196; and local identity 195–97, 199–200; and migration 191; as a superdiverse city 191 Guānhuà 4, 13, 104; and Manchu 41; in the Ming 16–17; Nanjing 15, 23; norms of 15; northern 14–23, 27–29; orthographies 26–29; southern 14–21, 23, 26–27, 29–30; Western perception of 19–26; see also Mandarin Guóyīn see national pronunciation Guóyǔ see Mandarin, national language Hainanese 166, 169, 174, 206, 214 Hakka 27, 61, 75; in Guangdong 193; in Hong Kong 149; in Singapore 166, 169, 174, 206, 214; in Taiwan 114, 123–26, 135, 278–79, 281 Hanlin Academy: language training in 43, 45 Hànyǔ 103, 293; in Singapore 178 Hànyǔ Pīnyīn 29, 112, 212 Haugen, E. 103, 105–6 Hokkien: in Hong Kong and Macao 148; in Singapore 88–89, 166–67, 169, 174, 206, 211, 214, 216; see also Holo; Southern Mǐn; Taiwanese language Holo 122–25; in education 128; use and attitudes towards 135–36; see also Southern Mǐn; Taiwanese language hometown 190, 194–96; language 16; pronunciation 190; speech 190–94, 199–201 Hong Kong: attitudes towards Cantonese, English and Pǔtōnghuà in 152–53; Basic Law 149, 157; bilinguals 145, 147; Cantonese as medium of instruction 89, 151–52, 155–56; Cantonese spoken in 147–50, 155; census data 89–90, 150; colonial language policy 145–46; dialect bilingualism 157; economic miracle 158; English in education 153–55; ethnic minorities 150; Hakka spoken in 149; history in 1960s and 1970s 142; language-in-education 142, 148, 154–55; language of government 149, 157; language planning 154–55; language policy 89, 146, 152, 154–55, 157; language shift 149; mainland students in 153; official language of

310 Index 142, 149; Pǔtōnghuà used in 89–90, 148, 150, 157–58; size and population 142, 149–50; status and use of English 89–90, 149; traditional Chinese characters used in 149, 151 Hong Kong English 90, 147 household registry system 191 Huáyǔ: language of Singapore 177, 274 IaEs see Interactions across Englishes illiteracy 62–63, 67, 76; see also literacy IMLD see International Mother Language Day immigration: to Singapore 174, 206; to Taiwan 124–26, 277, 281 imperialism 66 indexicality principle 234, 239 Indians: in Hong Kong 150; in Singapore 166–69, 176, 205–6, 208, 218 indigenization: in Taiwan 126–27 Indonesians: in Hong Kong 86, 150 in-group code 250, 264, 268 inner circle 86, 178 Interactions across Englishes 86, 88–90, 95 International Mother Language Day 130, 190 Japan 14, 23, 60; as a model of nation building and language modernization 104, 107; and Portugal 144; Singapore under rule of 82; Taiwan as a colony of 66, 102, 124–25, 277 Japanese language: in Taiwan 114, 277 Jiāng-Huái Mandarin 274–75, 282–83, 285 Jiangxi, as origin of Mandarin teachers in Fujian 46 jiāxiānghuà see hometown, speech Jīnghuà 28 Jīngzhōu 30–32 Kachru, B. 86 kana 27–30 Kāngxī emperor 16, 30, 40, 43–44; and Jesuits learning Manchu 43; and proper pronunciation 44 Kāngxī imperial dictionary 16–18; and pronunciation 44 KMT see Kuomintang Kuomintang 66, 102, 111, 114–15, 124, 126, 251, 280; Singapore branch of 211

language activists 76, 189, 196, 200 language and identity in Taiwan 6, 67, 122, 127, 239–40, 242–44, 245–47 language attitudes: in Hong Kong 152–54; of migrant students 185–86, 191–92, 197–201; in Singapore 221–22 language barrier: between ethnic groups in Hong Kong and Macao 150–51 language contact: in Taiwan 250–53, 274–75, 277, 281 language ideology 5, 101, 111, 123, 127, 185, 187, 200, 242–43, 251 language law(s): of China 113, 192; of Hong Kong and Macao 149, 153, 156, 157; of Taiwan 114–16, 128 language of Beijing see Beijing dialect language of government: in Hong Kong 149, 157; in the PRC 113 language planning: governmental vs. non-governmental 104; in Hong Kong and Macao 89, 154–55, 157; indigenization-driven and internationalization-driven 126; in Japan 107; and lexicography 43; in the People’s Republic of China 111–14; during the Qing 44, 53; in the Republic of China 102–3, 105–7, 109–11; in Singapore 165, 167, 170, 177, 208–9, 211; in Taiwan 113–16, 126, 132, 135, 251, 275, 285 language policy: in Hong Kong 89, 146, 152, 154–55, 157; in Singapore 164–69, 170, 177–79, 208–9, 213, 224; in Taiwan 121–37, 230, 251, 277; see also bilingual language policy in Singapore; colonial language policy language purism 107–8, 116 language-in-education: in Hong Kong 6, 82, 89, 142, 148, 154–55; in Macao 91, 155–56; in Singapore 6, 86–87, 170–71, 174, 177, 208–11; in Taiwan 115–16, 123–24, 125, 128–30, 131–32, 137 language shift: in China 5, 6, 17, 23, 60, 101, 106; in Hong Kong 149; in Singapore 164, 165, 167, 172, 177, 178, 213–15; in Taiwan 101–2 Láo Nǎixuān 28–29 Lee K. Y. 7, 164, 165, 168–69, 170, 171, 173, 178, 179, 210, 211, 212 lingua franca 4; Bazaar Malay as 206–7; Cantonese as 149, 192; English as 81, 84–85, 89–90, 93–95, 146, 209, 269; Macao Portuguese Pidgin as 145;

Index  311 Mandarin as 13, 14, 16, 18, 20, 104, 166, 277; and the Mass Language Movement 110; Portuguese as 83 linguistic ecologies see multilingual ecologies literacy: and Cantonese in Hong Kong 151–52; and Christian missionaries 75; in Classical Chinese 73; in English 213; in English and Mandarin 142, 148; and language reform in China 27; in Mandarin 7, 76, 112, 117, 150; in Romanized Mǐn 60, 63, 65–66; Taiwan local language 129; see also illiteracy; vernacular literacy Literary Chinese see Classical Chinese local languages education see mother tongue, education in Hong Kong; education in Taiwan London Missionary Society 22, 62 Lú Zhuàngzhāng 28, 29, 30 Macanese see Macao Creole Portuguese (MCP) Macao: attitudes towards English in 153–54; Basic Law 149; bilinguals 145, 147; Chinese and Portuguese bilingualism 146; colonial language policy 145–46; English in education 148, 155–56; as gaming capital 143, 158–59; population of 142; as a Portuguese colony 146; size and population 91; traditional Chinese characters used in 149, 151 Macao Creole Portuguese (MCP) 84, 91, 144, 146, 148, 151 Macao Portuguese Pidgin (MPP) 144–45, 146–47 Mainlanders in Hong Kong and Macao 91, 143, 153, 155, 158 Malay language 7–8, 80–88, 146, 151, 164, 167–71, 205–8, 210, 214, 216–17 Manchu language 39; description of 41–42; in the late seventeenth and early eighteenth century 43; in the mid-to late eighteenth century 47; in the nineteenth century 54; and the teaching of Mandarin 45–47; translations from Chinese 48–51; as a vernacular language 51–53 Mandarin: blue-green 105, 106, 108; dominance of 4, 6–7, 101–2, 111–12, 126–27, 136–38, 164, 167; Eastern (Jiāng-Huái) 275, 279, 282, 283, 284–85, 287n12; in education in

Taiwan 125, 277; five-tone 13, 15, 18–19, 20–21, 29–30; four-tone 21, 28; history 15–19, 293; as an international language 178; Jiāngnán 15, 16–18, 20, 29, 32n6; lingua franca of the Chinese diaspora 166; lingua franca in Taiwan 277–78; local varieties of 166, 197, 278; in Macao 6–7, 91, 95n7, 142, 148, 151, 154; Northern 5, 14, 15, 17, 18–19, 20–21, 23, 27–28, 29, 30–32, 32n7, 33n9, 33n12, 33n16, 34n27, 52, 101, 111, 115, 192, 268n4, 279, 286n8, 287n12; Northwestern 279, 287n12; official language of Taiwan 125, 127; promotion in Taiwan 114, 123, 127, 135, 277–80; in Singapore 7, 86–87, 88, 164, 165–66, 167, 169–70, 172–73, 174, 175, 176, 177, 178, 179, 179n8, 200, 207, 208–11, 211–13, 214, 216, 217, 224, 224n2; Southern 13, 14, 15–19, 20–21, 23, 26–30, 30–32, 32n7, 33n13, 33n16, 105; see also Guānhuà; national language; Pǔtōnghuà Mandarin Phonetic Symbols 29, 108, 109, 114; see also orthographic schemes for Chinese Mass Language Movement 110–11, 117 matrix language 251, 252 Mayers, W. F. 25–26 Meadows, T. T. 21–23, 24, 34n21 migration 7, 80, 92, 93, 94, 166; within China 32n6, 189, 194; to Hong Kong and Macao 148; to Singapore 173–74, 189, 206; to Taiwan 124–26, 243, 274, 275, 277, 281 Mǐn 27, 29, 33n15, 46, 61, 62, 64–77, 102, 192, 279, 287n12; see also Southern Mǐn missionaries 5; and Manchu 43; and Mandarin Chinese 13, 16–17, 20, 22, 24–25, 26, 43; and Romanization of Chinese dialects 27, 33n15, 34n25, 60–77; see also Buddhist missionaries Morrison, Robert 20, 23–24, 25, 34n20 mother tongue 2, 7–8, 55n4, 67, 72, 185, 193, 263; education in Hong Kong 152, 154–55, 157; education in Taiwan 115–16, 127; as an ideological notion 185–89; movement in Taiwan 123, 130; vs. mǔyǔ 188, 193; and regional dialects 116, 189–92, 194–96, 199–201, 279, 286n7; in Singapore 168, 169–70, 171–72, 207, 208–11, 212, 213, 215,

312 Index 217, 218–20, 221; see also step-mother tongue multidialectism 192 multilingual ecologies 80–81, 91, 92, 94, 115, 123, 124, 147, 190, 191–92 Muysken, P. 251–52, 268, 275, 285n2 mǔyǔ see mother tongue nánhuà see Mandarin, Southern Nanjing 5, 14, 15, 22, 23, 34n23, 41, 281, 282–85, 287n12 Nanjing dialect 5, 13, 15, 21, 24–26, 34n24, 282–85 Nankingese see Nanjing dialect national language 60, 76; in the People’s Republic of China 92, 111–14, 150, 153, 188, 192, 291; in Qing China 14–15, 27, 32n3, 294; in Republican China 65, 104–11, 210; in Singapore 205, 206, 291; in Taiwan 114–16, 123, 125–26, 127, 128, 189, 274, 277–78, 280, 291 National Language Movement 109–10, 125 national pronunciation 31, 104–6, 109–10, 115 nationalism 66, 108, 116, 174, 211, 263 Nationalist government see Kuomintang nation-building and language 104, 188; in Singapore 165, 171, 178, 189 neutral tone 292–94, 299–301, 304n3 New Chinese immigrants 173, 175 Ningbo 61, 82 Non-Chinese ethnic groups see ethnic minorities nonce-borrowings 252 official language(s): of Hong Kong 142, 149; of the People’s Republic of China 111–14, 116, 117, 274–75; of Singapore 86, 168, 206–10; of Taiwan 124–26, 229–30, 247 One Belt One Road see Belt and Road initiative orthographic schemes for Chinese 5, 25–30, 34n25, 60, 104, 108, 109; see also Mandarin Phonetic Symbols Overseas Taiwanese 238–43 Pakistani in Hong Kong: 90, 150 Patuá see Macao Creole Portuguese (MCP) Pearl River Delta 81, 83, 142, 149, 157, 192

People’s Republic of China 1, 3, 5, 6, 9, 101–2, 103, 116, 117, 122, 142, 157, 192, 251, 258; language planning 111–14 Peranakans see Straits Chinese phonetic notation systems see Mandarin Phonetic Symbols; orthographic schemes for Chinese Pidginization see Macao Creole Portuguese (MCP), Macao Portuguese Pidgin (MPP) Portuguese language 7, 19, 20, 80, 91, 92, 142, 144–48, 149–51, 153–56, 159 PRC see People’s Republic of China Presbyterian Church 62–67 purism see language purism Pǔtōnghuà 66, 76, 92, 93, 151, 193, 195–99; as an educational standard 293; in Hong Kong 89–90, 148, 150, 152–53, 157–58; as an identity marker 200; in Macao 156; in the People’s Republic of China 111–14, 116, 117, 274–75; promotion of 66, 185, 188, 192; variation within 278; see also Mandarin; official language(s) Qiánlóng emperor 34n27, 39, 40, 47–48, 49, 50 Qièyīnzì yùndòng see script reform movement Qing dynasty: bilingualism 44; language planning 44, 53 Qing history: importance for sinophone studies 1 questionnaires 135, 152–53, 215–16, 253, 262, 267, 269n13 /r/ phoneme in Mandarin 9, 291, 292, 295, 298–301 Raffles, S. 82, 205, 210 Reformation: and vernacularization 51, 72 regional languages: marginalization of 6, 7, 76, 103, 106, 114, 116, 188, 190, 192, 195, 196, 211 Republic of China 114–15, 123–25, 174, 277; language planning 102–3, 105–7, 109–11; see also Taiwan resourcing policy 128, 130, 132, 134, 137 Ricci, M. 20, 33n18 ROC see Republic of China Roman alphabet 5, 20–21, 25–26, 27–29, 33n15, 34n21, 34n25, 60–77, 104, 148, 152, 170, 237

Index  313 Romanized Mǐn 61, 62, 65–75 rù tone 17, 19–20, 21, 27–29, 31–32, 33n11, 294, 295, 304n1 Schlegel, G. 25–26 script reform 111–12 script reform movement 26–27 SGEM see Speak Good English Movement Shanghai 23, 65, 82, 93, 94, 114, 143, 191, 281 Shanghainese 61, 76, 192, 280, 281 Shantou (Swatow) 61, 64, 194 Shàoxīng 47–48, 53 Shěn Qǐliàng 50, 52 Shùnzhì emperor 39, 40, 43–44, 53 Sibe 41 Singapore: attitudes towards English in 221–22; bilingual education 170, 172, 177, 208, 211; bilingual language policy 179, 213; bilingualism 164, 170, 176, 178, 205, 208–9, 216; bilinguals 219; colonial language policy 82, 167–68, 211; dialects spoken in 166; diglossia 168; English as home language 173–74, 178, 213–15; English in education 169–71, 177, 207, 209–10; English proficiency 218–20; ethnic Chinese 166, 173–76; ethnic groups 166, 169, 214, 218, 222; home language use 213–15, 217; immigration 174, 206; independence 82, 164, 167, 170, 173, 175, 209, 211; Indians in 166–69, 176, 205–6, 208, 218; individual multilingualism 215–17; language attitudes 212–13; language planning in 165, 167, 170, 177, 208–9, 211; language policy in 164–69, 170, 177–79, 208–9, 213, 224; language shift 164–65, 167, 172, 177, 209, 213–15, 224; official languages 86, 168, 206–10; rise and dominance of English 163–66, 172, 178, 213; status of English 167–69, 207–10; trilingualism 216, 217–19, 224n5; see also Speak Good English Movement; Speak Mandarin Campaign Singapore English 87–88, 207–8, 213, 218, 223; see also Colloquial Singapore English Singlish 213, 221–22; see also Colloquial Singapore English Sinicization 123, 148 Sinkeh Chinese 173, 174–75

sinophone world 2–3, 4–6, 7, 9, 14, 32, 60, 76, 80–81, 86, 93–94, 95n1, 101, 111, 115, 123, 131, 142, 150, 151, 152, 156, 166, 186, 189, 200, 231, 250–51, 274 SMC see Speak Mandarin Campaign social mobility and language 16, 136–37, 150, 152, 154, 229 sociolinguistic scale 199, 231, 234–35, 238, 245, 247 South Asians: in Hong Kong 90, 150–51; see also ethnic minorities; Hong Kong Southeast Asia: speakers of Cantonese 2, 88, 166–67, 169, 174, 206, 207, 209, 211, 214, 216 Southern Mǐn 2, 60–77, 116, 148, 275–77, 279, 281, 285; in Macao 91; see also Hokkien; Holo; Taiwanese language Speak Good English Movement 211–13 Speak Mandarin Campaign 177, 179n8, 211–13 Standard Written Chinese 109, 110–11, 142, 148–51, 153, 155, 286n8 step–mother tongue 196–99 Straits Chinese 88, 173–74, 175, 206, 211 substrate languages 89, 146, 205, 207–8 Suzhou 16, 22, 76 Swatow see Shantou SWC see Standard Written Chinese Sydenstricker, A. 26, 34n24 Taiping rebellion 22–24, 26 Taiwan: attitudes towards English in 136–37, 263–66; bilingualism 125, 278; bilinguals 277; code-mixing in 231, 238; dialects of 127; diglossia 278; English in education 123, 131–34, 230; English language planning 126, 130–31; English use 135–36, 231–35, 251–53; historical language situation 114–15, 124–25, 277–82; identity 6, 66–67, 122, 127, 189, 231, 239–40, 242–44, 245–47, 250; immigration to 124–26, 277, 281; indigenization in 126–27; language contact in 250–53, 274–75, 277, 281; language planning in 113–16, 126, 132, 135, 251, 275, 285; language policy 121–37, 230, 251, 277; martial law 66–67, 122, 251; mother tongue education 115–16, 127; mother tongue movement 123; official language(s) of 124–26, 229–30, 247; promotion of Mandarin 114, 123, 127,

314 Index 135, 277–80; status of English 123, 230; traditional Chinese characters used in 114; Wú dialect speakers in 280–82; see also Formosa Taiwan Mandarin 232, 241, 250, 251, 253, 254, 258–60, 262, 264, 267, 268, 268n4, 269n6, 274–85 Taiwanese language, written 61, 66–67, 117, 230–32, 247n2 Taiwanese Nativization Movement 122, 269n12 Taiwanese Southern Mǐn see Taiwanese language Taiwanization 123, 126, 189 Tamil language 7, 81, 86–87, 88, 164, 167, 168, 169, 170, 171, 206–7, 208, 210, 213, 214, 217–18 Teochew (Chaoshan) dialect 75, 166–67, 169, 174, 193–95, 201n5, 206, 207, 214, 217, 275 three circles model 86 tones: Mandarin 13, 15–22, 27–30, 31–32, 33n11, 33n12, 34n20, 52, 255–57, 267, 268n3, 291–94, 297, 299–301; Middle Chinese 50, 295, 304n1 traditional Chinese characters: in Hong Kong and Macao 149, 151; in Taiwan 114 translanguaging 147, 151–52 trilingualism: in Hong Kong and Macao 89, 142, 146, 148, 150; in Singapore 216, 217–19, 224n5 Ueda Kazutoshi 107 UFPs see utterance-final particles

urbanization 191 utterance-final particles 9, 207, 274, 275–77, 282–85, 286n9 vernacular(s): written 39, 40–41, 48–54, 55n3, 60–77, 104, 109, 111, 148, 151–52, 278, 286n5 vernacular literacy 117, 151–52 Vietnam 60, 75, 125–26 Vietnamese 61, 75 Wade, T. F. 21, 22–26, 34n22 Wáng Bǐngyào 27–28 Wáng Zhào 28–29, 109 Woodin, S. F. 61, 68, 70, 72 written vernacular Chinese 40, 104, 109–11, 117n3; see also vernaculars Wú dialect 102, 279–82, 285, 286n8, 287n12 Xiamen 5, 60, 62, 81, 82; see also Amoy Xíng Gōngwǎn 292–94, 298 Xinhai revolution 31, 72 yīnyùnxué see Chinese phonological studies Yōngzhèng emperor 39, 40, 44–45, 53 Yuè dialects 102, 144, 149, 206, 279, 287n12; see also Cantonese Yùzhì Kāngxī zìdiǎn see Kāngxī imperial dictionary Zhejiang 46–47, 281–82, 284–85, 287n12 zhèngyīn ‘correct pronunciation/sounds’ see correct pronunciation