The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting: A Corpus-Based Study

This book investigates a special genre of interpreting in the Chinese context, namely Government Press Conference (GPC) Interpreting. Drawing on the modality system from Systemic Functional Grammar and a corpus of 21 interpreting events, the project explores the regular patterns of modality shifts in Chinese-English GPC interpreting and seeks explanations in the sociocultural context. As a corpus-based project, the book covers qualitative analysis of the sociocultural context, qualitative analysis of the interpersonal effects of modality shifts, and quantitative analysis of modality shifts. This book will contribute to the understanding of the distinctive features of GPC interpreting in China, shed new light on the rendition of modality between Chinese and English in specific contexts, and also inspire new thoughts on the nature of interpreting in general.

108 downloads 7K Views 3MB Size

Recommend Stories

Empty story

Idea Transcript


Corpora and Intercultural Studies 1

Xin Li

The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting A Corpus-Based Study

Corpora and Intercultural Studies Volume 1

Series editors Kaibao Hu, School of Foreign Languages, Shanghai Jiao Tong University, Shanghai, China Hongwei Ding, School of Foreign Languages, Shanghai Jiao Tong University, Shanghai, China

This book series publishes original monographs and edited volumes in the investigations of different types of corpora (including text, speech and video) with a particular focus on intercultural studies. The differences in language use expressed in comparable corpora can be analyzed from an intercultural perspective. The emphasis is on excellence and originality in scholarship as well as synergetic interdisciplinary approaches and multicultural perspectives. Books exploring the role of the intercultural studies in the research fields of translation, linguistics, and culture, with a corpus-based approach will be especially welcome. The series publishes books that deal with emerging issues as well as those that offer an in-depth examination of underlying issues. The target audiences of this series include both scholars and professionals who are interested in issues related to intercultural communication across different cultures and social groups, which are reflected by the investigation in comparable corpora. Corpora and Intercultural Studies book series is published in conjunction with Springer under the auspices of School of Foreign Languages (SFL), Shanghai Jiao Tong University (SJTU). The first series editor is the Dean of SFL at SJTU, and the book series editorial board consists of leading scholars in the research field of corpora and intercultural studies in the world.

More information about this series at http://www.springer.com/series/15588

Xin Li

The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting A Corpus-Based Study

123

Xin Li School of Foreign Languages Shanghai Jiao Tong University Shanghai China

ISSN 2510-4802 Corpora and Intercultural Studies ISBN 978-981-10-5168-5 DOI 10.1007/978-981-10-5169-2

ISSN 2510-4810

(electronic)

ISBN 978-981-10-5169-2

(eBook)

Library of Congress Control Number: 2017943234 © Springer Nature Singapore Pte Ltd. 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Printed on acid-free paper This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. part of Springer Nature The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Acknowledgements

The present work is based on my doctoral thesis completed at Shanghai Jiao Tong University (China). Therefore, my deepest gratitude goes to my supervisor Prof. Hu Kaibao1 for his constant encouragement, insightful comments, and unfailing guidance throughout all phrases of this research project. I would also like to express my sincere gratitude to other experienced translation scholars in China, including Prof. Zhang Boran from Nanjing University, Prof. Chai Mingjiong and Prof. Sun Huijun from Shanghai International Studies University, Prof. Huang Libo from Xi’an International Studies University, Prof. Wang Jiankai from Fudan University, Prof. Wang Zhenhua, Prof. Liu Huawen, and Prof. Yang Bingjun from Shanghai Jiao Tong University, all of whom offered critical and constructive suggestions on my research proposal and writing. My special thanks go to Mr. Chen Mingming, the former Director of Department of Translation and Interpreting, Ministry of Foreign Affairs, who kindly accepted my invitation for a telephone interview and responded to my questions with patience despite his busy schedule at work. I am grateful to my colleagues and friends at the Center for Translation and Intercultural Studies, including Prof. Tao Qing, Dr. Zhu Yifan, Dr. Kyung Hye Kim, Dr. Li Tao, Dr. Pan Feng, Dr. Liu Huidan, Dr. Meng Lingzi, Dr. Xie Lixin, Dr. Li Xiaoqian, Dr. Hei Yi, and Dr. Sheng Dandan, who have given me encouragement and support in different ways. I have also been blessed with the help and support from many respectable scholars in the field of Translation and Interpreting Studies home and abroad. In addition to warm encouragement and positive suggestions they gave me through personal communication, Prof. Qin Hongwu from Qufu Normal University kindly

1

For the sake of authenticity, the Chinese scholars’ full names in most parts of the book (including acknowledgements and appendices) are presented in accordance with the Chinese naming convention, i.e. family name before given name. For the convenience of citation in English, the author’s name is put in reverse order (i.e. the western naming order) in the front cover.

v

vi

Acknowledgements

provided technical support for my data coding and statistical analysis, Prof. Mona Baker from the University of Manchester, Prof. Xiao Xiaoyan from Xiamen University, Prof. Miriam Shlesinger from Bar-Ilan University and Dr. Sun Tingting from Beijing Foreign Studies University all generously shared otherwise unobtainable academic resources that I needed for my research. Last but not least, I would like to thank my amazing family for their support, understanding, and unconditional love over the years. In particular, I would like to dedicate the present work to my son Yuchen, whose warm hugs and brilliant smiles have lit my life and pulled me through whatever difficulties in my research and writing.

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 Corpus-based Interpreting Studies as an Offshoot of Corpus-based Translation Studies . . . . . . . . . . . . . . 1.1.2 GPC Interpreting in the Chinese Context . . . . . . . . . . 1.1.3 Modality in GPC Interpreting . . . . . . . . . . . . . . . . . . . 1.2 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Organization of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Previous Literature on Interpreting and Modality . . . . . . . . . . . 2.1 Corpus-based Interpreting Studies. . . . . . . . . . . . . . . . . . . . . . 2.1.1 An Overview of Corpus-based Interpreting Studies. . . 2.1.2 Corpus-based Studies on the Sociocultural Aspects of Interpreting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Corpus-based Studies on GPC Interpreting in China . . 2.2 Studies of Modality in Translation Between Chinese and English. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Studies on the Translation of Modality from English to Chinese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Studies on the Translation of Modality from Chinese to English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Contrastive Studies on Modality in Chinese and English Political Speeches . . . . . . . . . . . . . . . . . . 2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.... ....

1 2

. . . . . .

. . . . . .

2 3 5 6 7 8

.... .... ....

11 11 11

.... ....

14 18

....

23

....

24

....

26

.... .... ....

28 29 30

. . . . . .

. . . . . .

vii

viii

3 Theoretical Framework and Methodology . . . . . . . . . . . . . . . . . . 3.1 Theoretical Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Toury’s Framework of DTS . . . . . . . . . . . . . . . . . . . . 3.1.2 The Semantic System of Modality . . . . . . . . . . . . . . . 3.1.3 The Sociocultural Context of Interpreting . . . . . . . . . . 3.1.4 The Working Model for Descriptive-Explanatory Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Data and Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 The Corpus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Identifying High-Frequency Modality Markers . . . . . . 3.2.3 Coding Modality Shifts . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4 Describing Modality Shifts in the Corpus . . . . . . . . . . 3.2.5 Exploring the Relationship Between Modality Reconstruction and Its Sociocultural Context . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Modality Shifts in the Use of English Modality Markers . . . . . . 4.1 General Features of Modality Shifts . . . . . . . . . . . . . . . . . . . . 4.1.1 Overall Degree and Distribution of Modality Shifts . . 4.1.2 Modality Shifts with Different ST Speakers . . . . . . . . 4.2 Interpersonal Effects of the Major Types of Modality Shifts . 4.2.1 Weakening and Subjectivization . . . . . . . . . . . . . . . . . 4.2.2 Strengthening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3 De-obligation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Contents

. . . . .

. . . . .

. . . . .

. . . . .

35 35 35 38 43

. . . . . .

. . . . . .

. . . . . .

. . . . . .

47 48 49 50 52 55

.... ....

57 57

. . . . . . . . . .

61 61 62 66 68 70 75 80 86 87

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

5 Modality Shifts in the Translation of Chinese Modality Markers . . . 5.1 General Features of Modality Shifts . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1 Overall Degree and Distribution of Modality Shifts . . . . . . 5.1.2 Modality Shifts with Different ST Speakers . . . . . . . . . . . . 5.2 Interpersonal Effects of the Major Types of Modality Shifts . . . . . 5.2.1 Strengthening and Objectivization . . . . . . . . . . . . . . . . . . . . 5.2.2 De-obligation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 The Reconstruction of Modality in Relation to Its Sociocultural Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Modality Reconstruction in GPC Interpreting in Relation to Its Meta-Discourse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 The Meta-Discourse on Diplomatic Interpreting in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Juxtaposing the Reconstruction of Modality in GPC Interpreting with the Meta-Discourse . . . . . . . . . . .

.

89 89 89 94 96 97 105 109 111 113 113 114 117

Contents

ix

6.2 Modality Reconstruction in GPC Interpreting in Relation to its Typological Features . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 Participants and Goals of Communication . . . . . . . . . . 6.2.2 Topics and Speech Type . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 Temporal and Spatial Setting . . . . . . . . . . . . . . . . . . . 6.2.4 Working Mode and Directionality . . . . . . . . . . . . . . . . 6.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Major Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Contributions and Implications . . . . . . . . . . . . . . . 7.3 Limitations and Suggestions for Future Research . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

120 120 123 124 126 129 130

. . . . .

. . . . .

. . . . .

. . . . .

133 133 137 138 139

Appendix A: Top 350 Wordlist of the English TTs . . . . . . . . . . . . . . . . . 141 Appendix B: Top 350 Wordlist of the Chinese STs . . . . . . . . . . . . . . . . . 151 Appendix C: Coding Sample of Modality Shifts: I/We think . . . . . . . . . . 161 Appendix D: Telephone Interview with Mr. Chen Mingming . . . . . . . . . 201

Abbreviations

CECIC CI CIS CMM CPPCC CTS DTS EMM F. GPC IS LL LT NPC PLA SFG SI ST TS TT

Chinese-English Conference Interpreting Corpus Consecutive interpreting Corpus-based Interpreting Studies Chinese modality marker Chinese People’s Political Consultative Conference Corpus-based Translation Studies Descriptive Translation Studies English modality marker Frequency Government Press Conference Interpreting Studies Log-likelihood Literal translation National People’s Congress The People’s Liberation Army Systemic Functional Grammar Simultaneous interpreting Source text Translation Studies Target text

Note: This list does not include those abbreviation forms used in one particular table only, or used for the coding of modality shifts. They are explained in footnotes or in a specific table (e.g., Table 3.6). Commonly used abbreviations for countries and organizations (e.g., the USA, ROK, WTO) are also excluded here.

xi

Chapter 1

Introduction

Despite the undisputable sibling relationship between Translation Studies1 (TS) and Interpreting Studies (IS), these two disciplines have followed divergent paths of development. Presumably, IS has a lot to borrow from TS as an “elder brother” with better established theoretical and methodological foundations, yet in reality, IS so far “has been shaped much more decisively by approaches from other than its sibling discipline” (Pöchhacker 2009, p. 48). Under the influence of Descriptive Translation Studies (DTS) developed by Toury (1978, 1980, 1995), TS since the 1980s “has largely focused on exploring literary translations in their social contexts” (Diriker 2008, p. 209), while IS has been primarily concerned about the “cognitive and psycholinguistic aspects” (ibid.) of simultaneous conference interpreting, which is considered as “the most salient type of interpreting in the 20th century” (Diriker 2004, p. 1). In fact, the impact of DTS and its central concept of “translational norm” has been so strong in the translation literature that some interpreting scholars also attempted to explore various norms governing interpreting behavior in the 1990s (Harris 1990; Schjoldager 1995; Gile 1999), following Shlesinger’s (1989) skeptical discussions on the possibility of extending the theory of translational norms to interpreting. However, as Straniero Sergio and Falbo (2012a, p. 25) points out, most of these norms “seem not to fit Toury’s definition of norms” because these norms are “cognitive-based” rather than socially situated. It was not until the 21st century that visions of IS and TS came to converge, with both translation and interpreting scholars calling for efforts to address more critically the relationship between the translating/interpreting practice and its sociocultural context. For TS, it means to move beyond the “objective” description of translation products and explore more systematically the sociocultural and ideological factors shaping translators’ choices. For IS, it means to shift away from the “Translation Studies” here is used in the narrow sense, to refer to studies of written translation. The broad-sense counterpart will be called “Translation and Interpreting Studies” to avoid confusion. However, “translation” is used either in the broad sense or the narrow sense depending on the specific context throughout the book and notations will be offered in cases of ambiguity.

1

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_1

1

2

1

Introduction

obsession with the cognitive process of simultaneous conference interpreting and analyze various types of interpreting practices in relation to their specific historical, institutional, and ideological contexts. This reciprocal dialogue on common interests between translating and interpreting scholars is most strongly evidenced in the 2006 volume Sociocultural Aspects of Translating and Interpreting co-edited by Pym et al. (2006), and the 2012 special issue of Translation and Interpreting Studies on socially-oriented translation and interpreting studies (see Angelelli 2012). The strength of this dialogue is also witnessed by the 2013 International Forum on Translation Studies with the theme “Translation and Interpreting as Social Practice” held at Henan Norman University, China. The significance of this sociocultural approach of IS lies not only in the contribution to a broader view of interpreting as a socially situated communicative practice, but also in challenging the geopolitical “bias towards prestigious forms of interpreting practice in developed countries” (Cronin 2002, p. 387) that constitutes only a minority sector of the interpreting reality. Among the relatively neglected forms of interpreting different from the Eurocentric simultaneous conference interpreting in the IS literature, there are “valuable, ideologically sensitive” (ibid., 391) interpreting activities being extensively practiced in the developing countries, which deserve scholarly attention. A case in point is the GPC interpreting in the Chinese context, which will be explored in the present study as both an inter-lingual translational activity and a social discursive practice.

1.1 1.1.1

Rationale Corpus-based Interpreting Studies as an Offshoot of Corpus-based Translation Studies

Corpus-based Translation Studies (CTS) has received growing concern and acknowledgement as a valid approach to translation studies since Mona Baker first advocated that applying corpus linguistics methods into translation studies would help “uncover the nature of translated texts as a mediated communicative event” two decades ago (Baker 1993, p. 243). Enlightening studies following this approach include empirical investigations into “translation universals” (Laviosa 1998; Olohan and Baker 2000; Xiao and Dai 2010), language-pair-specific features of translated texts (Huang 2008; Hu 2009), translation shifts (Munday 1998), translator style (Baker 2001; Han and Fan 2016), translation teaching (Zhu and Mu 2013; Xiong 2015) and diachronic change of translated language (Xia 2013). The fruitful findings reveal the strength of corpus methods in identifying generalizable linguistic features and behaviors about translation. Since Translation Studies in its broad sense include Interpreting Studies as its sub-discipline, there is no reason for restricting the corpus-based approach to studies of written translation. As Shlesinger (1998, p. 1) once argued, “corpus-based interpreting

1.1 Rationale

3

studies offer a tool which is both viable and revelatory not only for the study of interpreting, per se, but for translation studies as a whole”. However, despite the flowering of CTS, literature in Corpus-based Interpreting Studies (CIS) so far is lagging far behind, due to the difficulty of compiling interpreting corpora and limited access to authentic data. Besides, the limited studies in CIS seem to have a predominant focus on simultaneous interpreting (SI) in the European setting, as shown in the special volume of papers in Breaking Ground in Corpus-based Interpreting Studies edited by Straniero Sergio and Falbo (2012b). Therefore, the recent attempts by Chinese scholars to explore norms and strategies of consecutive interpreting (CI) in the GPC setting (Wang 2009; Liu 2010; Hu and Tao 2012; Sun 2013; Pan 2014), although sporadic, strongly complements the picture of the Eurocentric CIS literature. As Straniero Sergio and Falbo (2012a, p. 22) rightly points out, “corpora findings may tell us how translators translate, but not why they translate the way they do.” Granted, pure descriptions of linguistic features of translation with corpus methods have in the past yielded fruitful insights into our understanding of the translation-specific phenomena such as “translation universals” and “interpretese”. However, in order to understand translation and interpreting as situated practices, we need to go beyond the mere observation of linguistic regularities in the translated texts and seek explanations for translators/interpreters’ choices in their sociocultural context. Therefore, the present study will attempt to explore both the linguistic reconstruction of modality (“how”) in Chinese–English GPC interpreting and the sociocultural motivations behind it (“why”).

1.1.2

GPC Interpreting in the Chinese Context

As China is stepping up its communication with the world in which interpreting plays an increasingly important role, recent years witness a blossoming of interpreting studies published in Chinese domestic journals (Mu and Wang 2009) and their influence has also reached the western world, as evidenced in the spectacular volume Interpreting Chinese, Interpreting China by Setton (2011) as well as the special issue of the renowned journal Interpreting dedicated to China (Setton 2009). However, as the “history, demographics and present-day demands are shaping a profession somewhat different from the one that grew up in Europe 50 years ago” (ibid., 109), more efforts need to be made for the world to understand the interpreting scene in China. For example, the frequent practice of “to B” interpreting (“B language” refers to a working language different from the interpreter’s mother tongue) in China may strike the western interpreting scholars as not professional, as the general norm there is working only to the interpreter’s mother tongue, i.e., “A” language. Different from European interpreters who are balanced bilinguals or multilinguals well exposed to the cultures of the different working languages, “perhaps 98% of interpreters of Chinese are Chinese native speakers, and fewer than one in five have lived for more than a month or two outside China”, which means the foreign language they are often required to work into are “usually

4

1

Introduction

acquired through formal schooling and personal study” (ibid., 109). Different from many western countries with a large number of immigrants, where interpreting is often employed by various organizations and businesses for different communicative purposes, in China, “[g]overnment (both central and local) is by far the largest employer, with most routine day-to-day interpreting done by civil-servant interpreter-translators with language degrees, but who often have other duties and other career plans” (ibid., 109). This distinctive Chinese context with its sociocultural and ideological implications needs to be explored for the world to understand the reality of interpreting in China. GPC interpreting to be explored in the current study refers to interpreting at the central-government-level press conferences during the annual “Two Sessions” (known as “两会” in Chinese) of National People’s Congress (NPC, the national legislature of China) and Chinese People’s Political Consultative Conference (CPPCC, the top political advisory body), which are considered the most important and influential GPCs in China. My choice of the GPC interpreting as the subject of study in China is motivated by three reasons. First, GPCs during the “Two Sessions” serve as in important window for the world to access first-hand information about the Chinese government’s policies and stances towards important domestic and diplomatic issues. As China grows into a strong economic power attracting increasing attention from the world, the government has realized the importance of a more proactive and effective public diplomacy. Because of “ideological differences and interest clashes”, “biased or seriously distorted reports about China” have been published by western media and unfriendly ideas such as the “China threat theory” have been promoted, seriously undermining China’s image (Zhao 2007). Therefore, the Chinese government cannot “expect foreign media to portray China justly, or close the opinion gap they have created” (ibid). Instead, “China must present an accurate picture of itself to the world” through better “public diplomacy” (ibid). This is why China has been promoting a “proactive public diplomacy campaign” (Sun 2013, p. 59) in the past decades, during which the GPC system was established and developed. Since GPCs play such an important role in communicating China’s visions to the world, it will be significant to explore how interpreters are mediating this dialogue between the Chinese government and the foreign media. Second, GPC interpreting during the “Two Sessions” is performed by well-trained in-house interpreters from the Department of Translation and Interpreting affiliated to the Ministry of Foreign Affairs, who represent the top-level of professional interpreting competence in China. As interpreting in China is mostly initiated by the government agencies instead of businesses and organizations, this top-level GPC interpreting can to a large extent reveal the interpreting reality and professional norms in China. The last reason is a practical one. As Shlesinger (1989, pp. 112–114) points out, the obstacles hindering descriptive interpreting studies include the “virtual nonexistence of textual corpora” for interpreting and the inevitable interference of “monitoring” in interpreting performance. Fortunately, these problems are solved with the availability of the corpus CECIC (Chinese–English Conference

1.1 Rationale

5

Interpreting Corpus), details of which will be discussed in Sect. 3.2.1. For one thing, the live-broadcast tradition of Chinese GPCs during the “Two Sessions” has made it possible for us to collect authentic data and build a fairly large interpreting corpus for descriptive analysis. For another, the GPC as an event is available to the public anyway through live broadcast, so the recording will not affect the actual performance of the interpreters.

1.1.3

Modality in GPC Interpreting

Despite the methodological contribution of corpora to translation and interpreting studies, we have to admit that “computerized corpus studies are still unable to tackle many translational phenomena” (Pérez 2002). Therefore, to conduct a corpus-based descriptive study on GPC interpreting, a lot of time-consuming manual analysis on the source text (ST)–target text (TT)2 relations is necessary. With such a large corpus as CECIC, it is impossible to explore all linguistic phenomena concerning translation within one research project, so I choose to focus on the semantic– grammatical category of modality for detailed analysis. Following Halliday and Matthiessen (2014), I define modality as the semantic category existing in most languages to cover various degrees/types of uncertainties between “yes” and “no”, in other words, different degrees/types of “possibility” and “necessity”.3 There are three main reasons for this choice. First, modality construes important interpersonal meanings such as the speaker’s “attitude” (Thompson 2008, p. 68) or “stance” (Biber et al. 2000, p. 979) in communication. Therefore, modality markers (modality markers in this study refer to all these lexical and grammatical markers of modality observable in a textual corpus) such as modal verbs and modal adverbs play a significant role in conveying the Chinese government’s attitude and stance on domestic and international issues in GPCs, as indicated by their high frequency in both STs and TTs of the corpus (for details of the frequency, see Appendices A and B). It will be meaningful to investigate how modality is reconstructed by interpreters in their inter-lingual and inter-cultural mediation. Second, modality as a lexico-grammatical system exists in both Chinese and English, but the rendition of modality from Chinese to English involves a considerable number of choices on the interpreter. It will be interesting to explore the regularities in the interpreters’ choices for translating modality and how these regularities relate to the sociocultural context of Government Press Conference Interpreting. Last, thanks to the availability of GPC interpreting data, so far three book-length corpus-based or corpus-assisted studies on GPC interpreting 2

ST and TT are generally used for written translation. But as this study is mainly based on the transcriptions instead of the audio data, I will use ST and TT to refer to the transcribed source speech and target speech respectively. 3 Note that “possibility” and “necessity” are used in the broad sense to cover all types of modality: epistemic, deontic and dynamic.

6

1

Introduction

have been completed by Chinese scholars, but their linguistic observations are limited to the shifts of “addition”, “reduction” and “correction” in terms of ideational meaning (Wang 2009), and shifts of personal pronouns, terms of address, and intensifiers (Liu 2010; Sun 2013) with only occasional examination on modality-related hedging devices (see Sun 2013, pp. 148–166). In-depth analysis of modality shifts is yet to be conducted in GPC interpreting. Therefore, the present study attempts to explore this lacuna and hopefully will contribute to a deeper understanding of GPC interpreting in the Chinese context.

1.2

Research Questions

Drawing largely on the broad framework of Descriptive Translation Studies (DTS), the current project attempts to conduct a descriptive–explanatory study of modality in GPC interpreting based on the specialized corpus CECIC with the focus on the following overarching question: How are modal meanings reconstructed in Chinese–English GPC interpreting and what is the relationship between these interpreting choices and the sociocultural context? Inspired by Toury (1995, p. 82–85), I will explore the ST–TT relationships in terms of modality from two directions, first tracing the use of English modality markers from the TT to the ST, then the translation of Chinese modality markers from the ST to the TT. Borrowing the concept of translation shifts, i.e., “small linguistic changes occurring in translation of ST to TT” (Munday 2001, p. 55), I will explore the degree/types of shifts between the ST–TT “coupled pairs” [i.e., the pairs of ST-segments and corresponding TT-segments in translation, see Toury (1995, pp. 77–83)] involving modality markers to investigate the reconstruction of modality in GPC interpreting. Previous research on GPC interpreting in the Chinese context suggests that interpreters tend to adopt different strategies in translating the government officials’ speeches and the journalists’ questions, due to their solidarity with the official (Wang 2009; Liu 2010; Wang 2012). Although the STs in my data consist primarily of speeches by the Chinese government official, it also includes Chinese questions asked by journalists; therefore it will be interesting to see if the identity of the ST speaker affects the interpreter’s choices in reconstructing modality. In addition to the overall patterns of modality shifts, it is also necessary to investigate the interpersonal effects (i.e., the effects of modality shifts on the interpersonal relationship between the participants in the interpreting activity) of the major types of modality shifts through the analysis of specific examples, which will contribute to our understanding of the relationship between the reconstruction of modality and the sociocultural context of GPC interpreting. Since the sociocultural context is too broad to analyze in an exhaustive manner, I will focus on the two most relevant sociocultural factors for GPC interpreting: the institutional norms of Chinese–English GPC interpreting indicated in the meta-discourse, and its typological features. Therefore, the above overarching question can be broken into the

1.2 Research Questions

7

following sub-questions, with Research Question 1 and its sub-questions to be addressed from two directions. (1) How are modal meanings reconstructed in Chinese–English GPC interpreting? (1a) What are the regular patterns in the reconstruction of modality in Chinese– English GPC interpreting? (1b) Does the identity of ST speaker affect the degree or major types of modality shifts in Chinese–English GPC interpreting? (1c) What are the interpersonal effects of the major types of modality shifts? (2) What is the relationship between the interpreters’ linguistic choices of reconstructing modality and the sociocultural context of Chinese– English GPC interpreting? (2a) How do the regular patterns of modality shifts relate to the institutional norms of Chinese–English GPC interpreting indicated in the meta-discourse? (2b) How does the regular patterns of modality shifts relate to the typological features of Chinese–English GPC interpreting?

1.3

Organization of the Study

The current study consists of seven chapters. Following this chapter that introduces the general background and research objectives, Chap. 2 presents a review of relevant literature on interpreting studies as well as modality. As the present study takes a corpus-based approach to investigate the reconstruction of modality in GPC interpreting and attempts to seek explanations in the sociocultural context, previous CIS studies with a focus on the sociocultural aspects, studies on GPC interpreting, studies of modality in translation will be given the most weight in the review. Chapter 3 discusses the research framework and methodology for the present study. First, the major theories that provides theoretical and analytical insights for the current research are presented, including the ST–TT comparative model of translation shifts from Toury’s (1995) DTS framework, the systemic functional view of modality from Halliday and Matthiessen’s (2014), and the integrated model of sociocultural context analysis inspired by Diriker (2004), Alexieva (1997) and Pöchhacker (2009). That leads to a summary of my integrated working model for descriptive–explanatory study of modality in GPC interpreting. Then, the research method is described through the discussion of the corpus data and the research procedure. Chapters 4, 5 and 6 constitute the core of the study, presenting the corpus-based analysis, findings and discussions in response to the research questions. Chapter 4 presents the corpus-based findings of modality shifts in Chinese–English GPC interpreting by proceeding from the English TT. In this direction, I will first identify the English modality markers in the TT, and then go back to the ST to explore what Chinese forms have triggered the use of English modality markers, and finally compare the ST–TT coupled pair for translation shifts. Chapter 5 presents a similar

8

1

Introduction

analysis from the ST direction. Proceeding from the Chinese modality markers in the ST, I will investigate what solutions are used by the interpreters to render them into English and then compare the ST–TT coupled pair for translation shifts. Analyses from the two directions will give a comprehensive picture of how modality is reconstructed in GPC interpreting. Following the descriptive studies, Chap. 6 maps the modality shifts in GPC interpreting onto the sociocultural context and discusses the motivational factors including the institutional norms and the typological features of GPC interpreting. As the concluding chapter, Chap. 7 summarizes the major findings of the study, discusses its contributions, implications and limitations, and offers suggestions for future research.

References Alexieva, Bistra. 1997. A typology of interpreter-mediated events. The Translator 3 (2): 153–174. Angelelli, Claudia V. 2012. The sociological turn in translation and interpreting studies. Translation and Interpreting Studies 7 (2): 125–128. Baker, Mona. 1993. Corpus linguistics and translation studies: implications and applications. In Text and technology: In honour of John Sinclair, ed. Mona Baker, Gill Francis, and Elena Tognini-Bonelli, 233–250. Amsterdam: John Benjamins. Baker, Mona. 2001. Towards a methodology for investigating the style of a literary translator. Target 12 (2): 241–266. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 2000. Longman grammar of spoken and written English. Beijing: Foreign Language Teaching and Research Press. Cronin, Michael. 2002. The empire talks back: Orality, heteronomy and the cultural turn in interpreting studies. The Interpreting Studies Reader, 386–397. London: Routledge. Diriker, Ebru. 2004. De-/Re-Contextualizing conference interpreting. Amsterdam: John Benjamins. Diriker, Ebru. 2008. Exploring conference interpreting as a social practice. In Beyond descriptive translation studies, ed. Anthony Pym, Miriam Shlesinger, and Daniel Simeoni, 209–219. Amsterdam: John Benjamins Publishing Company. Gile, Daniel. 1999. Norms in research on conference interpreting: A response to Theo Hermans and Gideon Toury. In Translation and norms, ed. C. Schäffner, 98–105. Clevedon: Multilingual Matters Limited. Halliday, Michael A.K., and Christian Matthiessen. 2014. Halliday’s introduction to functional grammar, 4th ed. Abington: Routledge. Han, Jianghong 韩江洪, and Qing Fan 凡晴. 2016. 基于语料库的路易•艾黎和许渊冲“三 吏”“三别”英译风格对比探究 [A corpus-based comparative study on the translation style of two versions of Du Fu’s “Sanli” and “Sanbie”]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (6): 93–100. Harris, Brian. 1990. Norms in interpretation. Target 2 (1): 115–119. Hu, Kaibao 胡开宝. 2009. 基于语料库的莎剧《哈姆雷特》汉译文本中“把”字句应用及其动 因研究 [Corpus-based study of Ba-construction in the Chinese versions of Hamlet by Shakespeare]. 外语学刊 [Foreign Language Research] (1): 111–115. Hu, Kaibao 胡开宝, and Qing Tao 陶庆. 2012. 记者招待会汉英口译句法操作规范研究 [Syntactic operational norms of press conference interpreting (Chinese–English)]. 外语教学与 研究 [Foreign Language Teaching and Research] (5): 738–750, 801.

References

9

Huang, Libo 黄立波. 2008. 英汉翻译中人称代词主语的显化——基于语料库的考察 [Explicitation of personal pronoun subjects in English–Chinese translation—a corpus-based investigation]. 外语教学与研究 [Foreign Language Teaching and Research] (6): 454–459, 481. Laviosa, Sara. 1998. Core patterns of lexical use in a comparable corpus of English narrative prose. Meta 43 (4): 557–570. Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese–English): Face management and information management. Ph.D. thesis, Heriot-Watt University. Mu, Lei 穆雷, and Binhua Wang 王斌华. 2009. 国内口译研究的发展及研究走向——基于30 年期刊论文、著作和历届口译大会论文的分析 [Interpreting studies in China: a journal articles-based analytical survey]. 中国翻译 [Chinese Translator’s Journal] (4): 19–25, 94. Munday, Jeremy. 1998. A computer-assisted approach to the analysis of translation shifts. Meta 43 (4): 542–556. Munday, Jeremy. 2001. Introducing translation studies. London: Routledge. Olohan, Maeve, and Mona Baker. 2000. Reporting that in translated English. Evidence for subconscious processes of explicitation? Across Languages and Cultures 1 (2): 141–158. Pan, Feng 潘峰. 2014. 基于语料库的汉英会议口译中模糊限制语的应用研究 [A corpus-based study of the application of hedges in Chinese–English conference interpreting]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (4): 24–29. Pöchhacker, Franz. 2009. Introducing interpreting studies. Shanghai: Shanghai Foreign Language Education Press. Pérez, María Calzada. 2002. A three-level methodology for descriptive-explanatory translation studies. Target 13 (2): 203–239. Pym, Anthony, Miriam Shlesinger, and Zuzana Jettmarova (eds.). 2006. Sociocultural aspects of translating and interpreting. Amsterdam: John Benjamins. Schjoldager, Anne. 1995. An exploratory study of translational norms in simultaneous interpreting: Methodological reflections. Hermes 14: 65–87. Setton, Robin. 2009. Interpreting China, interpreting Chinese. Interpreting 11 (2): 109–117. Setton, Robin. 2011. Interpreting Chinese, interpreting China. Amsterdam: John Benjamins Publishing Company. Shlesinger, Miriam. 1989. Extending the theory of translation to interpretation: Norms as a case in point. Target 1 (2): 111–115. Shlesinger, Miriam. 1998. Corpus-based interpreting studies as an offshoot of corpus-based translation studies. Meta 43 (4): 486–493. Straniero Sergio, F, and Caterina Falbo. 2012a. Studying interpreting through corpora. An introduction. In Breaking Ground in Corpus-based Interpreting Studies, ed. F. Straniero Sergio, and C. Falbo, 9–52. Bern: Peter Lang. Straniero Sergio, Francesco, and Caterina Falbo. 2012b. Breaking ground in corpus-based interpreting studies. Bern: Peter Lang. Sun, Tingting. 2013. Interpreting China: Interpreters’ mediation of government press conferences in China. Beijing: Foreign Language Teaching and Research Press. Thompson, Geoff. 2008. Introducing functional grammar. Beijing: Foreign Language Teaching and Research Press. Toury, Gideon. 1978. The nature and role of norms in literary translation. In Literature and translation, ed. J.S. Holmes, J. Lambert, and R. van den Broeck, 83–100. Leuven: ACCO. Toury, Gideon. 1980. In search of a theory of translation. Tel Aviv: Porter Institute. Toury, Gideon. 1995. Descriptive translation studies and beyond. Amsterdam: John Benjamins Publishing Company. Wang, Binhua 王斌华. 2009. 口译规范描写及其应用——一项基于中国总理“两会”记者招待 会交传语料的研究 [Description of norms in interpreting and its application–a study based on the corpus of consecutive interpreting in Chinese premier press conferences]. Ph.D. thesis, Guangdong University of Foreign Studies. Wang, Binhua. 2012. A descriptive study of norms in interpreting: Based on the Chinese–English consecutive interpreting corpus of Chinese premier press conferences. Meta: Journal des traducteurs/Translators’ Journal 57 (1): 198–212.

10

1

Introduction

Xia, Yun 夏云. 2013. 英译汉文学语言形合特征的历时研究 [A diachronic study of hypotactic features in English–Chinese translated literary texts]. 外语教学与研究 [Foreign Language Teaching and Research] (1): 105–116, 161. Xiao, Zhonghua 肖忠华, and Guangrong Dai 戴光荣. 2010. 寻求“第三语码”——基于汉语译文 语料库的翻译共性研究 [In pursuit of the “third code”: a study of translation universals based on the ZCTC corpus of translational Chinese]. 外语教学与研究 [Foreign Language Teaching and Research] (1): 52–58, 81. Xiong, Bing 熊兵. 2015. 基于英汉双语平行语料库的翻译教学模式研究 [Exploring a translation teaching model based on an English–Chinese parallel corpus]. 外语界 [Foreign Language World] (4): 2–10. Zhao, Qizheng. 2007. Better public diplomacy to present a truer picture of China. People’s Daily Online, 30 March 2007. http://en.people.cn/200703/30/eng20070330_362496.html. Accessed 14 Sept 2015. Zhu, Chunshen 朱纯深, and Yuanyuan Mu 慕媛媛. 2013. 以文本解释力为导向的语料库翻译 教学——香港城大翻译与双语写作在线教学/自学平台的设计与试用分析 [Towards a textual accountability-driven mode of teaching and (self-) learning for translation and bilingual writing: with special reference to a CityU on-line teaching platform]. 中国翻译 [Chinese Translator’s Journal] (2): 56–62, 127.

Chapter 2

Previous Literature on Interpreting and Modality

As the present project is a corpus-based interpreting study with the aim to investigate the reconstruction of modality in GPC interpreting in China and the related sociocultural context behind the interpreting choices, this chapter will provide a selected review of corpus-based interpreting studies and studies of modality in translation that either constitute the research background or have been illuminating in terms of methodology.

2.1 2.1.1

Corpus-based Interpreting Studies An Overview of Corpus-based Interpreting Studies

Although CIS is still a “cottage industry” compared with “the relatively organized and coordinated landscape of CTS” (Setton 2011, p. 34), it proves to be a fruitful paradigm to reveal regular patterns about the interpreting product, which contributes to our understanding of the interpreting process and the shaping factors. As Cencini and Aston (2002, p. 47) wittily puts it, “Like all speech, interpreting dies on the air. In order to study it, we need to resurrect the corpse by recording and transcribing it, thereby transforming the corpse into a corpus”. In fact, CIS projects had started before Shlesinger’s (1998) landmark article “Corpus-based interpreting studies as an offshoot of corpus-based translation studies”, with the earliest attempts represented by research projects from the Paris School (Seleskovitch 1975; Lederer 1981) in the 1970s and early 1980s. For example, Seleskovitch (1975) sought evidence for her “deverabalization” hypothesis for the interpreting process in an experimental corpus of consecutive conference interpreting from English to French. Lederer (1981) explored units of meaning and illustrated the interpretive theory with authentic examples based on a corpus of simultaneous conference interpreting from German to French. After a short © Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_2

11

12

2 Previous Literature on Interpreting and Modality

stagnation period of CIS along with the popularity of “controlled experimental studies” due to the vigorous criticism directed towards the interpretive theory on “epistemological and methodological grounds” (Setton 2011, p. 38), the late 1980s and 1990s witnessed its survival with the strongest efforts made by scholars in Israel (Shlesinger 1989b, 1991, 1995) and German-speaking countries (Pöchhacker 1994; Kalina 1998). For instance, Shlesinger (1989b) analyzed the “equalizing effect” of simultaneous interpreting (SI) on shifts in the position of texts on the oral-literacy continuum based on corpora of trial interpreting and conference interpreting between Hebrew and English. Then Shlesinger (1991) reported the “normalization” tendency of court interpreters based on a corpus of simultaneous and consecutive interpretation in multilingual trials, followed by her investigation of shifts in cohesion in SI based on an experimental corpus of interpreting performed by students (Shlesinger 1995). Pöchhacker (1994) investigated the interplay between the interpreters’ performance and the macro-context based on a corpus of simultaneous conference interpreting between English, German and French, and then Kalina (1998) investigated simultaneous interpreters’ strategies and errors based on corpus data from both authentic conference interpreting and student samples. Following Shlesinger’s (1998, p. 2) visionary statement that “the compilation of bilingual and parallel corpora is indeed overdue, given the potential to use large, machine-readable corpora to arrive at global inferences about the interpreted text”, the 21st century has witnessed an unprecedented enthusiasm in CIS among interpreting scholars with the increasing availability of machine-readable interpreting corpora and the widespread use of software tools to help analyze corpus data. Generally speaking, fruitful CIS projects in the new century were conducted mainly on two broad topics: linguistic and paralinguistic features of the interpreting product, and ST-TT relations in interpreting. First, linguistic or paralinguistic features of the interpreting product were explored mainly through comparing the interpreted speeches with non-interpreted speeches, translated texts, or source speeches, so that the hypotheses for “translation universals” were tested, modality-specific features were identified, languagepair-specific features were investigated, or evidence for the interpreter’s cognitive process was found. For instance, a team of Italian scholars (Sandrelli and Bendazzoli 2005; Russo and Sandrellii 2006) tested Laviosa’s (1998) “simplification” hypothesis about translated texts in interpreting based on comparable sub-corpora from the European Parliament Interpreting Corpus, and suggested that some of the features about lexical density and variety in interpreted speeches are language-pair-specific. From a different perspective, interpreting scholars from Israel (Shlesinger 2008; Shlesinger and Ordan 2012) explored lexical features in simultaneously interpreted speeches as opposed to written translation and comparable original speeches based on comparable corpora in Hebrew, which suggested the nature of “interpretese” as more spoken than translated. Although corpus-based interpreting studies are typically product-oriented, CIS projects, especially investigations on the paralinguistic features of the interpreted texts can nevertheless reveal evidence for the cognitive process of the interpreters. Two cases in point are

2.1 Corpus-based Interpreting Studies

13

Petite (2005), who examined the patterns of “repairs” in a corpus of simultaneous conference interpreting and concluded that “interpreters’ repairs are not necessarily motivated by the detection of an error” (ibid., 46) and Bendazzoli et al. (2011), who compared the pattern of “disfluencies” in interpreted texts and their source texts and revealed features constrained by the SI process. Most recently, CIS projects have been attempted by Chinese scholars to investigate the linguistic or paralinguistic features of the interpreted texts. For instance, the CTS research team at Shanghai Jiao Tong University (Hu and Tao 2010; Li 2011; Li and Hu 2013; Pan and Hu 2013; Pan 2014; Hu and Xie 2014; Xie and Hu 2015) investigated the use of passive structures, modal auxiliary verbs, hedges, relative clauses and indefinite quantifiers in Chinese-English GPC interpreting as opposed to comparable original English speeches or translated texts based on the Chinese-English Conference Interpreting Corpus (CECIC) and attributed these marked patterns to various factors including the linguistic differences between Chinese and English, the formality level of GPCs, and the interpreting process. Based on an experimental Chinese-English simultaneous interpreting corpus, Wang and Li (2014) examined the paralinguistic features of “pauses” in interpreted speeches as compared with source speeches, and the findings “reveal not only the difficulties encountered by interpreters but also their decision-making mechanism in the interpreting process” (ibid., 140). Among these studies, the most relevant to the present study are Li and Hu (2013) and Pan (2014), which will be reviewed more closely in Sect. 2.2.2 along with other translation studies concerning modality. Second, with the help of parallel interpreting corpora, various linguistic phenomena have been examined to reveal the ST-TT relations in interpreting such as translation shifts, the rendition of specific items, or the degree of lexical similarity between the ST and TT. For example, Dam (2002) examined the distribution of form-based and meaning-based interpreting based on an experimental corpus of consecutive interpretation (from Spanish to Danish) by professional interpreters. Through a careful observation of the formal-lexical relation between the ST and TT segments, she found lexical similarity a more salient feature than lexical dissimilarity in the corpus, which suggested that “interpreters pay considerable attention to formal source text features when working in the consecutive mode” (ibid., 274) against the hypothesis from the Paris School that “interpreting generally proceeds on the basis of a non-verbal representation of the source text” (ibid., 275). Dam’s observation has important implications for the present study, as her finding about the high degree of lexical similarity between the ST and TT in professional interpreting suggests that it is valid and meaningful to investigate the linguistic operation of lexical items like modality markers in GPC interpreting. Addressing the elusive and controversial concept of “explicitation”, Gumul (2006) conducted a comprehensive study of “explicitating shifts” based on a corpus of English-Polish simultaneous interpretation by advanced interpreting students and suggested that explicitation occurs in SI mostly as a by-product of language-mediation rather than a result of a conscious strategy. Addressing the same notion of “explicitation”, Baumgarten et al. (2008) chose to examine the rendition of a proper noun “Amazônia” from Brazilian to German based on a corpus of simultaneous and

14

2 Previous Literature on Interpreting and Modality

consecutive interpreting in an authentic setting, and reached an opposite conclusion: “explicitation is not a universal feature of interpreting, but rather an option that may be more frequent in certain settings or interpreting modes” (ibid., 198). Based on the same Brazilian-German interpreting corpus as used in Baumgarten et al. (2008), Meyer (2008) compared the rendition of proper names in different modes of interpreting (consecutive vs. the simultaneous) and concluded that “a considerable number of proper names were rendered in a way that deviated from the original” but “no clear differences were found between consecutive and simultaneous interpreting” (ibid., 119). Despite the different research objectives, all these above studies suggest that it is feasible to apply the CTS methods to interpreting research to reveal regular patterns in interpreters’ behaviors as long as the transcribed data are available. However, as Straniero Sergio and Falbo (2012, p. 22) reminded us, “corpora findings may tell us how translators translate, but not why they translate the way they do.” Therefore it remains a meaningful task for CIS scholars to “go beyond mere linguistic description and look for cognitive, ethical, social, cultural and even ideological explanations” (ibid., 22) for the ST-TT relations revealed by corpus data.

2.1.2

Corpus-based Studies on the Sociocultural Aspects of Interpreting

Against the background that traditional interpreting studies are preoccupied with the cognitive process of interpreting, recent years witness a growing interest among CIS scholars (Straniero Sergio 2003; Diriker 2004; Monacelli 2005; Beaton 2007; Wang 2009; Liu 2010; Hu and Tao 2012; Sun 2013) in the sociocultural aspects of interpreting, with issues like norms, situation-specific quality, meta-discourse of interpreting, politeness, face, ideology, etc. discussed along with the description of ST-TT relations in the corpus data. As the studies (Wang 2009; Liu 2010; Hu and Tao 2012; Sun 2013) conducted by Chinese scholars are all related to GPC interpreting in China to be reviewed in Sect. 2.1.3, this section will only discuss those socially-oriented studies by non-Chinese scholars. Following Toury’s (1980) introduction of the sociology-oriented concept of “norms” into translation studies to refer to the middle-ground constraints between “rules” and “idiosyncrasies” (ibid., 51) which result in “regularities of translation behavior within a specific sociocultural situation” (Baker 2004, p. 163), the 1980s and 1990s witnessed a turning point of translation studies from the ST-oriented, prescriptive research on “equivalence” to the TT-oriented, descriptive research on “norms”. While it is still true that “most of the scholars who tried their hands with the notion of norms were first and foremost engaged in the study of literary translation” (Toury 1999, p. 13), it has attracted increasing attention from interpreting scholars since Shlesinger (1989a) first discussed the possibility of extending the theory of translational norms to interpretation. For example, Harris (1990,

2.1 Corpus-based Interpreting Studies

15

pp. 115–118) discussed the norm of “speak[ing] in the first person” (ibid., 115) and the norm of the “honest spokesman” (ibid., 118) in professional interpreting; Schjoldager (1995, p. 84) suggested the norm that “[a]n interpreter is allowed to say something which is apparently unrelated to the source-text item in question, … provided that s/he can say something which is contextually plausible”; Gile (1999, p. 99) identified the norms of “maximizing information recovery”, “maximizing the communication impact of the speech” and “minimizing recovery interference”; and Garzone (2002, p. 114) suggested the “basic norm” of “giv[ing] a complete rendition of the ST”. However, as Straniero Sergio and Falbo (2012, p. 25) pointed out, most of these norms were discussed “[w]ithin the cognitive paradigm” and their socially-situated nature was underexplored. Among the discussions on the elusive concept of “norms” in interpreting, two corpus-based interpreting studies (Straniero Sergio 2003; Diriker 2004) stand out in their emphasis on the sociocultural and situation-specific aspects of norms. Straniero Sergio (2003) conducted a corpus-based study of English-to-Italian simultaneous media interpreting for Formula-One press conferences. By comparing the ST and TT segments in the transcribed data, he identified a translational norm specific to Formula-One press conference interpreting (or media interpreting): interpreters regularly use “emergency strategies” (e.g., “summarizing”, “generalization” and “metatextual glosses and tautological repetitions”) which “usually are considered ‘last resort’” (ibid., 140). After associating this norm with the specific situational context (e.g., the “phatic” function of SI on TV) of Formula-One press conference interpreting, he further discussed the gap between the “ideal quality expectations” for SI and the “real world criteria” and concluded that “quality standards should be adjusted to concrete SI situations” (ibid., 169–171). Although the translational norm of using “emergency strategies” in media interpreting is also related to the cognitive constraints of SI, Straniero Sergio put more emphasis on the situational factors specific to media interpreting behind this norm instead of discussing a norm constraining all SI performances. Therefore, this study is a step away from the cognitive-based discussion of norms to the identification of norms for specific situations. With a more overt emphasis on the analysis of SI behavior in relation to its sociocultural context, Diriker (2004) attempted to contextualize SI through its meta-discursive representation which “does not embody just a neutral description of intrinsic professional features, but presents a selected and hierarchized set of ‘norms’” (ibid., 25). To Diriker, translational norms of SI were not a research object to be extracted from the “textual” data in the interpreting corpus, but rather existing social realities to be revealed by the “extratextual” sources (for the two terms in quotation marks, see Toury 1995, p. 65). With this view of norms, she didn’t have to make the difficult decision on whether the regular patterns in interpreting performance are norm-induced, but was able to discuss the dialectic relationship between the interpreting performance and the perceived norms and offer tentative explanations for the “convergences and divergences between what is said and what is done in simultaneous conference interpreting” (Diriker 2004, p. 5, original emphasis). She first explored the broad sociocultural context of SI through

16

2 Previous Literature on Interpreting and Modality

analyzing the meta-discourse on SI from various sources (including general reference books, codes of ethics, professional organizations, academia, Turkish media and a popular book authored by a professional interpreter), and revealed a tension between the norm of “complete identification with the speaker’s message” suggested by generalized meta-discourse and “the active participation of the interpreter in shaping the message” (ibid., 52) indicated in the more contextualized anecdotal accounts by interpreters. Then she analyzed the immediate context of her interpreting corpus data, consisting of a particular SI (between English and Turkish) event for a 2-day conference held on “metaphysics and politics” at Boğaziçi University in Istanbul. The participant observations and interviews revealed a “diversity of viewpoints with regard to the presence and performance of interpreters at a single SI assignment” (ibid., 5). As for the corpus-based textual analysis, Diriker focused on the use of first-person pronoun “I” in the TT and found a multiplicity of the “speaker-positions” in contrast to the exclusive use of the speaker’s “I” according to the norms in SI. The TT-oriented analysis revealed 58 explicit shifts from the speaker’s “I” in the delivery, which occurred mostly in 8 cases: “apologies of the speakers and/or interpreters”, “mistakes of the speakers and/or interpreters”, “overlapping/semi-verbal/inaudible interaction on the floor”, “problems with the transmission of the interpreter’s or speaker’s voice”, “ambiguous or contradictory input in the floor”, “language/culture-specific discussions or difficult word-connotations in one conference language on the floor”, “references in a non-conference language on the floor”, “accusations of misinterpretation from the floor” (ibid., 85). Finally, through juxtaposing the actual SI behavior with the meta-discourse, she made the following conclusions: the norms revealed by the meta-discourse on SI serve not as “intrinsic performance standards” for interpreters, but as “performance instructions” whose actual turnout is actively “negotiated by the interpreters” (ibid., 139); the meta-discourse which depicts “simultaneous interpreters as professionals who can unproblematically access and transparently transfer the meaning, essential content or primary information as intended by the original speakers” is “mythical” and “purposeful” in nature (ibid., 141–142) and the relationship between interpreters, ST speakers and the sociocultural contexts is “complex and dialectical” (ibid., 144). The strength of Diriker’s study lies in her incorporation of “textual” and “extratextual” analysis, her extensive discussions on the sociocultural context of SI and her classification for the situations where shifts from the speaker’s “I” occur. However, the analysis of 58 shifts in the corpus with no quantitative comparison between the occurrence of “shifts” and “non-shifts” or between the different categories of shifts offers only evidence that the uses of “I” for different speaker positions exist but doesn’t contribute to our understanding of the probability or degree of shifts or the distribution of shifts among the different categories. Another weakness is that her identification of “shifts in the speaking subject” was only based on her critical reading of the TT instead of a one-to-one comparison between the ST and TT segments, so the 58 shifts identified may not represent the whole picture. Also seeing simultaneous interpreting as a situated communicative activity, Monacelli (2005) explored interpreters’ self-regulation in terms of participation

2.1 Corpus-based Interpreting Studies

17

framework and interactional politeness (contextual shifts, changes in alignment and shifts in footing) in a corpus of simultaneous conference interpreting between French, English and Italian (transcribed from 10 conference interpreting events). To collect evidence for interpreters’ self-regulation, she conducted a comparative textual analysis of the ST and TT segments with the focus on non-obligatory translation shifts in the linguistic categories of personal reference (indicating stance), transitivity process (indicating agency and voice), mood/modality, and threats to face, which revealed the overriding trend of [+distance] (i.e., “depersonalization”) and [-direct] moves of interpreters with the major effect of mitigating ST speakers’ face-threatening acts to TT receivers (so that the TT receivers’ as well as their own faces were preserved). As a complement to the corpus findings, she added a debriefing phase to question the interpreters on their awareness and motivation of the [+distance] and [−direct] shifts. Protocol analysis showed that all interpreters recognized these moves as “strategic in nature” and the majority (6 out of 10) of them mentioned their motivation “to distance themselves from the situation and a potentially offensive text” (ibid., 172), which confirmed the nature of these shifts as self-regulatory and face-preserving. In addition, on the regularity and acquisition of these strategies, all subjects considered them as common behaviors and claimed that these strategies were “acquired from watching senior colleagues on the job”, which pointed to the “normative nature of these phenomena” (ibid., 169). The strength of this study lies in the combination of corpus textual analysis with protocol analysis, so that the overall preference for [+distance] and [−direct] shifts was unveiled as self-regulatory, face-preserving and normative in nature through the interpreters’ reflections on their own behavior instead of solely through the researcher’s intuitive judgment. The weakness is that the four linguistic categories selected for the analysis of translation shifts was not on the same level so that there was an overlap between the first three (personal reference, transitivity process, mood/modality) and the last one (threats to face). In addition, the analysis on mood/ modality was concerned only with the additions or omissions of modality markers, instead of covering modality shifts between corresponding ST-TT pairs, which might reveal different trends in terms of distance and directness. In addition to discussions on norms, politeness and face work, the sociocultural aspects of interpreting within the CIS paradigm also includes discussions on ideology, of which a typical example is Beaton (2007). Based on a small corpus of SI (German to English) for 3 European Parliament plenary debates, Beaton (ibid.) conducted a comparative ST/TT analysis of two salient features in the European Parliament discourse—lexical repetition and metaphor strings—to explore the role of SI in the European Parliament. Analysis of lexical repetition of the key concept “European Union” revealed patterns such as “lexical contraction” and “general foregrounding of the item European Union and backgrounding of European Parliament and European Community” (ibid., 292, original emphasis) and analysis of the three most frequent metaphor strings (“the enlargement process is a journey”, “accession is a race” and “the European Union (EU) is a ship”) revealed features such as “the use of non-idiomatic English language and literal translation of metaphoric concepts” and “explicitation of metaphorical concepts” (ibid., 293). All

18

2 Previous Literature on Interpreting and Modality

these analyses led to two important findings or implications: first, “EU institutional hegemony is strengthened by simultaneous interpreters”; second, “the simultaneous interpreter is an additional subjective actor in heteroglot communication” (ibid., 271). Beaton’s study well illustrates how translation shifts can reveal the active and subjective mediation of interpreters in an institutional setting, but she failed to discuss the ethnographic information about the interpreters, which might have offered explanations for their preference to strengthen the EU institutional hegemony. All the above studies illustrated the value and prospect of using parallel interpreting corpora in analyzing the sociocultural aspects of interpreting, including the exploration of situation-specific quality and norms, the discussion of the interplay between interpreting performance with the meta-discourse, the investigation of interpreters’ politeness strategies and face work through linguistic indicators, and the impact of interpreting on ideological discourse. However, all these socially-oriented CIS studies by non-Chinese scholars concentrated on SI with the “to A” direction (i.e., interpreting into the interpreter’s mother tongue), which reflects the Eurocentric tradition of conference interpreting. Interpreting practice in China where “history, demographics and present-day demands are shaping a profession somewhat different from the one that grew up in Europe fifty years ago” (Setton 2009, p. 109) is less known to the world. For instance, CI and “to B” interpreting (i.e., interpreting into a working language different from the interpreter’s mother tongue) is common practice in China. Thus it remains the responsibility of Chinese scholars to conduct descriptive and empirical interpreting studies based on authentic data to help the world understand the interpreting reality in the Chinese context. Fortunately, in recent years, a considerable number of Chinese scholars have conducted corpus-based studies on GPC interpreting (representing the top-level of professional interpreting competence in China), and their works will be reviewed in the next section.

2.1.3

Corpus-based Studies on GPC Interpreting in China

Just like all the other CIS projects, corpus-based studies on GPC interpreting in China by Chinese scholars either looked at the linguistic/para-linguistic features of the interpreting product (Li 2011; Li and Hu 2013; Pan and Hu 2013; Pan 2014; Hu and Xie 2014; Wang and Li 2014; Xie and Hu 2015) or explored the ST-TT relations (Wang 2009; Liu 2010; Hu and Tao 2012; Sun 2013). It is noteworthy that most of these studies on the ST-TT relations are socially-oriented, suggesting the Chinese interpreting scholars’ acute awareness of the socially-situated nature of interpreting. Considering their relevance to the present study, this section will focus on the review of the socially-oriented studies on ST-TT relations in GPC interpreting, including two projects on translational norms (Wang 2009; Hu and Tao 2012), and two on the GPC interpreters’ agency role in terms of participation framework and face work (Liu 2010; Sun 2013). Another two relevant projects

2.1 Corpus-based Interpreting Studies

19

(Li and Hu 2013; Pan 2014) will be reviewed in Sect. 2.2.1 due to their close relevance to the translation of modality. The doctoral thesis by Wang (2009) is probably the first corpus-based descriptive study of translational norms for interpreting in China, which was later refashioned into a short article published in Meta (Wang 2012). Although the journal article version is more updated, a lot of details and discussions in the Ph.D. thesis have been deleted for the limited space, so here I will concentrate on the review of Wang (2009) with occasional reference to Wang (2012) for the convenience of citation. To explore the interpreting norms “as the shared values and ideas among interpreters of the profession and uses of interpreting service on what the generally-accepted interpreting methods and strategies are and what the right and proper interpreting behaviors are” (ibid., 199) in the interpreting activities for China’s annual Premier Press Conferences, a sample of 5 Premier Press Conferences was drawn from an parallel corpus of 11 Premier Press Conference interpreting events (from 1998 to 2008 with Premier Zhu Rongji and Premier Wen Jiabao as the primary speakers). Three types of “actual norms” (the norms of ST-TT relations, the norms of TL communication and the norms of professional ethics) were investigated, with the first two types extracted through corpus analysis and the third one described “through an extra-textual analysis of the interpreters’ interviews and reflections”, then the “perceived norms” in the interpreting profession were described “through the analysis of meta-textual documents” before a systematic analysis of the similarities and differences between the “actual norms” and “perceived norms” (Wang 2009, p. 5). The inter-textual comparison between the ST and TT with the focus on the types and frequencies of macro-level translation shifts (expansion/addition, reduction and modification/correction,1 with the first two types divided into various subcategories) revealed four typical norms of ST-TT relations in Premier Press Conference interpreting: (1) the “norm of adequacy”; (2) the “norm of explicitation of logic cohesion and coherence”; (3) the “norm of specificity of content”; (4) the “norm of improved level of explicitness”. The intra-textual analysis of the TT revealed three typical norms of communication: (1) the “norm of optimizing communication across languages”; (2) the “norm of facilitating and promoting communication”; (3) the “norm of optimizing communication effect” (ibid., 6). The extra-textual analysis revealed the norms of professional ethics of trying to “balance their loyalty to the speaker, the commissioner, the audience and any other relevant parties”, “abiding by professional morals”, and acknowledging that “the interpreters’ role isn’t transparent and invisible”. Then the “perceived norms” based on “code of conduct/practice/ethics” documents from interpreting organizations and institutions like the International Association of Conference Interpreters were summarized, including “faithfulness”, “promot[ing] communication”, being “impartial”, “unbiased” and “non-judgmental”, acting

“Expansion”, “reduction” and “modification” were originally used in Wang (2009) to refer to the three broad types of shifts, but “addition”, “reduction” and “correction” were used in Wang (2012) to refer to the same concepts.

1

20

2 Previous Literature on Interpreting and Modality

“neutrally and objectively”, etc. (ibid., 101–108). Comparison between “actual norms” and “perceived norms” revealed partial congruence with two major differences: (1) the perceived norm of “faithfulness” is not always complied with, as Premier Press Conference interpreting tends to be “clearer in logic, more specific in content and more explicit in meaning”; (2) the perceived norm of being “non-judgmental” and “objective” is not consistent with actual practice, as Premier Press Conference interpreters are “not so invisible” evidenced in the frequent translation shifts (ibid., 7). To sum up, this study is a groundbreaking attempt to describe interpreting norms through the analysis of translation shifts, and it is illuminating to the present study with its descriptive manner and the integrated employment of inter-textual and extra-textual analysis. Particular enlightening is his findings about the differences between the “actual norms” and “perceived norms” of interpreting. Besides, there was an incidental finding relevant to the present study (Wang 2012, p. 207): “shifts of omission usually occur in the interpretation of journalists’ remarks, but not in the interpretation of the Premier’s words, which may imply that interpreters tend to be more loyal to their principal of service”, which tempts me to test if the GPC interpreters’ reconstruction choices of modality also vary between different ST speakers. Meanwhile, Wang’s (2009; 2012) study also has several weaknesses. First, the norm of “adequacy” was summarized based on the frequent occurrence of shifts of addition, including “structural addition”, “informational addition and elaboration”, “explicitation of intended meaning”, etc. (Wang 2009, p. 69), but these shifts seem to point to the TL-oriented norm of “acceptability” rather than the SL-oriented norm of “adequacy” in Toury’s (1995, p. 56) sense, thus the meaning of this term “adequacy” here needs clarification. Second, the four extracted norms in ST-TT relations seem not to be on the same level: the first is related to “initial norms” and the rest are closer to “operational norms” according to Toury’s (1995, pp. 56–58) categorization, but the author failed to make any distinction or explanation. Last, despite the broad-sense definition of “shifts” as “changes in the target text as compared with the source text” (Wang 2012, p. 200), translation shifts identified in this study were restricted to the macro-level formal shifts, but what about changes in the semantics of corresponding words or expressions in the ST and TT? These micro-level shifts also deserve attention. Another corpus-based study of interpreting norms was conducted by Hu and Tao (2012) based on CECIC 3.0, which is larger in size (it includes 30 transcribed GPC interpreting events) than the corpus used by Wang (2009). Hu and Tao (ibid.) employed two sub-corpora of CECIC 3.0: the parallel corpus of Chinese-English GPC interpreting, and the parallel corpus of Chinese-English Government Work Report translation. Different from Wang’s (2012) bottom-up approach, Hu and Tao (2012) adopted a top-down approach, i.e., formulating hypotheses about interpreting norms first, and then test them through corpus observation. The focus of this article was on the syntactic operational norms in Chinese-English GPC interpreting, so linguistic phenomena like conjunctions, compound sentences and degree adverbs were investigated in terms of the ST-TT relationship to test the hypothesized norms of “explicitation”, “implicitation”, “simplification” “complication”, and “neutrality”

2.1 Corpus-based Interpreting Studies

21

(as opposed to “strengthening” and “weakening”) (ibid., 740–741). Following Toury’s (1995, p. 82) exemplification of proceeding from both ST and TT to establish “coupled pairs” for an “exhaustive” comparative analysis, they explored the ST-TT relations from two directions in GPC interpreting as compared with Government Work Report translation: (1) starting from selected Chinese linguistic units in the ST and investigating their English translations in the TT; (2) proceeding from selected English linguistic units in the TT and then investigating their Chinese corresponding items in the ST. The comparative analysis led to the following conclusions: (1) Chinese-English GPC interpreting “is governed to varying degrees by such operational norms of explicitation, implicitation, simplification, complication, strengthening and weakening, with the impact of explicitation and complication more salient than implicitation and simplification”; (2) “the influence of the explicitation, simplification and complication norms is weaker on Chinese-English GPC interpreting than on the Chinese-English Government Work Report translation” but “the impact of the implicitation and strengthening norms on GPC interpreting is greater than on the Government Work Report translation” (ibid., 738, my translation). A particularly interesting finding relevant to the present study was that while equivalent translation (i.e., “zero-shifts”) as governed by the norm of “neutrality” appeared to be the mainstream trend, the “strengthening” and “weakening” trends were also noticeable, i.e., “GPC interpreters often strengthen or weaken the semantic information in the ST based on their understanding of the context, either to emphasize particular facts or phenomena, or to avoid extreme expressions of opinion” (ibid., 748). In addition to the description of norms (“what”), they (ibid., 738) also discussed contributing factors for the existence of these operational norms in interpreting (“why”), which include the linguistic differences between Chinese and English, time constraint of the interpreting process, and the interpreters’ active involvement in the activity. The strengths of this study lie in its fairly large corpus, the comparison between interpreting and translation, the attention to both ST features and TT features in establishing ST-TT relations, and the combination of data description and explanation. However, the comparison between the overall impact of “strengthening” and “weakening” was not discussed. The findings about the ST-TT relations based on the observation of selected linguistic phenomena offer important evidence to support or oppose existing general hypotheses, but as any single corpus-based study inevitably has a limited view of the whole picture, more research on different linguistic phenomena is needed for a comprehensive understanding of the operational norms of Chinese-English GPC interpreting. To my knowledge, so far there have been three book-length CIS projects on GPC interpreting completed by Chinese scholars. One is the doctoral thesis completed at Guangdong University of Foreign Studies (China) by Wang (2009) on interpreting norms as reviewed above. The other two are Liu (2010) and Sun (2013), both focusing on the GPC interpreters’ agency role in terms of participation framework and face work. Liu (2010) is a doctoral thesis completed at Heriot-Watt University (UK). Guided by the hypothesis that “audience design, that is the adjustment of a speaker’s output to

22

2 Previous Literature on Interpreting and Modality

suit a particular participation framework, is involved in interpreters’ performance” (ibid., 4) and based on a small corpus of three interpreted GPCs (2003–2005), she investigated translation shifts in selected linguistic parameters (including address terms and deference, personal reference, indirectness/directness, explicitation and presupposition) in three different participation frameworks (Chinese questions, English questions and Chinese replies) and analyzed these shifts as face management and information management strategies. Different patterns of translation shifts concerning face management were found between the three participation frameworks, such as “reduced deference”, “increased directness” and increased “distance and detachment” in the interpretation of Chinese questions, “increased deference” and “weakened directness” in the interpretation of English questions, and “the trend of solidarity” in the interpretation of Chinese officials’ replies, which all “suggest that the interpreter’s behaviour is related to the participation framework and audience design” (ibid., 161). As for information management, shifts of explicitness were found to occur frequently in the interpretation of Chinese questions and replies, but very rarely in the interpretation of English questions, which indicates that “the interpreter is now designing for the Chinese Ministers and she no longer needs to make adjustments for relevance because even the questions are asked in English but these questions are about Chinese policies, so these questions are compatible with the cognitive environment of the minister” (ibid., 171). Then presupposition shifts were found to occur more noticeably in the interpretation of Chinese and English questions than in the interpretation of replies, which points to the interpreters’ identification with the Chinese government officials as “backed up by the interpreter’s sociality rights as a loyal employee” (ibid., 190). In addition to confirming the initial hypothesis that “[t]he interpreters’ behaviour is influenced by the particular audience (i.e., the primary intended receiver) that they have in mind to serve”, the study also demonstrated that “the interpreters working for these press conferences do not always follow the principle of ‘faithfulness, accuracy, completeness’” and “interpreting is far from being a mechanical activity with no advocacy involved” (ibid., 194). As probably the latest book-length CIS project on GPC interpreting, Sun (2013) is a published book version of her Ph.D. thesis (Sun 2012) completed at the University of Manchester. Armed with Goffman’s social communication theory (participation framework, footing, face and face-work), she explored the interpreters’ role and positioning in GPC interpreting based on a corpus of 6 GPCs held in response to the outbreak of SARS (Severe Acute Respiratory Syndrome) in 2003. Analysis of translation shifts in personal pronouns and intensifiers revealed prominent patterns such as “the addition of the first person plural pronoun we and its variants, mainly to refer to China or the Chinese government” and “the addition of intensifiers such as really, very much and indeed, in particular in rendering Chinese remarks and questions into English”, which serve to “strengthen the presence of the Chinese government as the main player in the interaction” and “safeguard the positive image of the Chinese government by emphasizing its efforts in dealing with the crisis” (ibid., 172). Analysis of translation shifts in terms of address and hedges revealed a noticeable tendency for the interpreters to use more polite forms of address and add hedging devices in their rendition of English

2.1 Corpus-based Interpreting Studies

23

questions into Chinese, which “in effect downplay the confrontational tone of the original utterances and save ‘face’ primarily for the Chinese officials” (ibid., 167), but little evidence was found to “suggest that interpreters pay particular attention to the face of journalists” (ibid., 169). These findings suggested that GPC interpreters are “proactively engaged in the construction of interactional meaning”, and pointed to their role as “institutional aides or insiders” but not “institutional gate-keepers”, and their “distinctive position as ‘semantically neutral’ but ‘emotionally/ pragmatically partial’ (ibid., 179–180). Although the GPCs in Sun (2013) occur in a very special context (the SARS epidemic), they are similar in genre (e.g., the formal setting and the relationship between the primary participants are similar) to the GPCs during the “Two Sessions” to be investigated in the present study, so her findings are enlightening on our general understanding of GPC interpreting. By observing the ST-TT relations from different linguistic perspectives, CIS projects on GPC interpreting in China so far seem to have reached a consensus on the understanding of GPC interpreters’ performance as not always “faithful” and “neutral” and their “visible” position as “institutional aligners” of the Chinese government. However, translation shifts were mostly analyzed on the macro-level (i.e., the omission or addition of certain linguistic items/devices), and the degree and pattern of semantic shifts for critical linguistic phenomena such as modality remain underexplored. Besides the interpersonal relationship of the interpreters’ institutional alignment with the government official, other contextual factors such as the working mode, directionality and speech type are also underdiscussed.

2.2

Studies of Modality in Translation Between Chinese and English

According to Halliday and Matthiessen (2014), modality is the semantic category existing in most languages to describe different degrees/types of “possibility” and “necessity” in the expression of interpersonal meanings, and the semantic system of modality includes various dimensions such as modality type, modality value and orientation (the details of the modality system will be discussed in Sect. 3.1.2). Although the system of modality exists in most languages including Chinese and English, the translation of modality involves more room for shifts than the translation of proper nouns or technical terms, and it is interesting to investigate what kinds of modality shifts regularly occur and whether these regularities reflect differences between language systems or the translator/interpreter’s intervention in the specific sociocultural context. However, compared with the profuse literature (including many corpus-based studies) on modality in the mono-lingual context, such as modality in English (Quirk et al. 1985; Palmer 1990; Klinge 1993; Mindt 1995; Biber et al. 2000; Facchinetti et al. 2003; Millar 2009; Leech 2011) and modality in Chinese (Zhu 1982; Li and Thompson 1989; Li 2003; Zhu 2005; Hsieh 2006; Peng 2007; Xu 2008; Zhang and Peng 2016), systematic studies of modality in

24

2 Previous Literature on Interpreting and Modality

translation between Chinese and English are rather limited. Focusing on the most relevant literature to the present study, this section will first review studies on the translation of modality from English to Chinese and then review studies on the translation of modality from Chinese to English, and finally discuss some illuminating projects on the contrastive studies of modality between Chinese and English in political speeches.

2.2.1

Studies on the Translation of Modality from English to Chinese

So far discussions on the translation of modality between Chinese and English have been primarily concerned with the translation of modal verbs, especially the Chinese translation of English modal verbs. Earlier studies (Sun 1995; Xue 1999; Li 2005, 2007a) were mostly based on the researcher’s experience and examples collected with the aim to offer prescriptive suggestions on the translation of English modal verbs. For example, Sun (1995) and Li (2007a) suggested guidelines on the Chinese translation of the English modals “shall” in legal documents; Xue (1999) and Li (2005) discussed the importance of the semantic classification of the English modals before translating them into Chinese. Then corpus-based descriptive studies became the trend as evident in more recent projects on the translation of English modal verbs (Wang 2007; Zhao 2009; Luo 2011; Li 2012; Zhao and Liang 2013). For instance, Zhao (2009) investigated the treatment of will and would in English-Chinese translation based on a parallel corpus consisting of both literary and non-literary texts, Luo (2011) studied the Chinese translation of can, may, shall, should, and will in the Corpus of United Nations Assembly Resolutions, and Li (2012) explored the Chinese translation of can and could in fictional texts based on a parallel corpus of Pride and Prejudice. These three MA projects all took advantage of corpus tools and found some regularities in the Chinese translation of English modals, but their observations were limited to the ST-TT relations on the formal level (such as the use of modal verbs, modal adverbs and mood elements in Chinese to translate English modals), except that Li (2012) mentioned the possibility of translation shifts in modality value. Among the corpus-based studies on the Chinese translation of English modals, Wang (2007) and Zhao and Liang (2013) are particularly relevant to the present study, which will be reviewed in more detail. Despite the focus on a different text type and a different translation direction, Wang’s (2007) monograph based on her Ph.D. thesis (Wang 2004) is illuminating to the present study. Using the best-selling Harry Potter series and their Chinese translations as the main corpus data and the modality theory in Systemic Functional Grammar (SFG) as the analytical framework, Wang (2007) investigated the effects of the speaker’s gender and age on the modality system (including modality type, orientation and polarity) expressed by can and could in the English ST and whether these effects are equally realized in their Chinese translations. Interesting findings

2.2 Studies of Modality in Translation Between Chinese and English

25

include: (1) remarkable differences exist between the distribution of semantic types of can and could in the ST and their corresponding expressions in the TT, which challenges the traditional notion of “equivalence”; (2) the speaker’s gender and age don’t affect the semantic types of can and could in the ST, but do influence the numbers of modal expressions and semantic types in the TT: for example, there is a tendency to translate could uttered by male speakers or child speakers into obligation; (3) the speaker’s gender and age affect the orientation (explicit or implicit) of the corresponding expressions of can and could in the ST, but don’t affect their polarity (positive, direct negative, or transferred negative). All these findings pointed to the norm of functional equivalence instead of formal equivalence in the Chinese-English translation of Harry Porter (ibid., 157). This study successfully demonstrated the feasibility of incorporating corpus data, statistical tools with the SFG framework to study the phenomenon of modality in translation, but I have to disagree with her in coding the semantic type of the modal verbs strictly based on the speech functions of the clauses (ibid., 78). I would consider Halliday’s (1994, p. 91) illustration of the relationship between speech function and modality type as an idealization to facilitate our understanding of their basic relations instead of a feasible coding framework (e.g., “What should we do now?” is clearly a question, but should expresses obligation instead of probability or usuality), which probably explains why this table was deleted in Halliday and Matthiessen (2004, 2014). With the focus on the change of modal strength (i.e., modality value in Halliday’s term), Zhao and Liang (2013) investigated the translation of epistemic may and might in literary and non-literary texts based on the General EnglishChinese Parallel Corpus of Beijing Foreign Studies University. They found that the translators frequently omit the two weak modals or use stronger modal expressions in the Chinese TTs in addition to the use of equivalent modals, so that the corresponding modal strength becomes higher in the Chinese translations. After comparing the frequency of these “strengthening” shifts in literary and non-literary texts, they concluded that “the higher modal strength in translated texts is related to communicative conventions rather than genres” (ibid., 96). Particularly interesting is their discussions on the different communicative conventions between English and Chinese. Inspired by the House’s (1997) summary of the different communicative modes in English and German, Zhao and Liang (2013) argued that the communicative convention in Chinese is similar to that in German, which is content-oriented and characterized by the frequent use of strong modals, while the English communicative convention is addressee-oriented and characterized by the preferred use of weak modals, therefore the Chinese translation of English modals “may” and “might” features a tendency towards “strengthening”, to comply with the target language norm. These discussions are relevant to the present study because they lead to the hypothesis that a tendency towards “modal weakening” will emerge in Chinese-English translation to accommodate the differences in communicative conventions. However, their study only touched on the epistemic use of two low-value English modals, and thus their conclusion cannot be generalized to the translation of modality or translation of modal verbs before support from further research is available.

26

2.2.2

2 Previous Literature on Interpreting and Modality

Studies on the Translation of Modality from Chinese to English

Among the limited literature on modality in Chinese-English translation, one MA thesis (Li 2007b) and three journal articles (Fu 2013; Li and Hu 2013; Pan 2014) are particularly relevant to the present study, most of which are corpus-based projects. The MA thesis is relevant because it employed the corpus-based method and adopted the view of modality system from SFG. The other three projects are relevant because they all investigated modality in Chinese-English GPC interpreting. Drawing on a corpus consisting of 10 Chinese diplomatic speeches with their English translations, Li (2007b) conducted a comparative study of the Chinese modal verbs in the ST and the English modal verbs in the TT in terms of frequencies, modality types and modality values. She found that the ST and TT employ a similar number of modal verbs to express modality, most of Chinese modal verbs can be translated into English modal verbs of the same modality type or other supplementary modal expressions, but the distribution of modality values is significantly different: most of the Chinese modal verbs in the ST are in the high value while most of the English modal verbs in the TT are in the median value. This reveals the translators’ tendency to translate higher-value Chinese modal verbs into lower-value English modal verbs. These findings indicate that modal verbs in English and Chinese are comparable in meaning and function, but Chinese diplomatic speakers tend to use high-value Chinese modal verbs in their speeches which are often rendered into English modal verbs of lower value to meet the TT norms. It will be interesting to see if this “weakening” tendency in Chinese-English diplomatic translation will be confirmed by further studies with a more fine-grained coding framework and with investigations of various modality markers instead of modal verbs alone. Based on an earlier version of CECIC (CECIC 3.0), Li and Hu (2013) investigated the features in the use of central modal verbs in Chinese-to-English GPC interpreting in China as compared with their use in original English texts at American GPCs. Comparative analysis in terms of modality value and modality type revealed an overuse of strong modals (will, shall, should and must) and deontic modals (i.e., modals expressing obligation) and an underuse of weak modals (would, might, and could) and epistemic modals (i.e., modals expressing probability) in interpreted GPC texts in China compared with the original English GPC texts in the United States. To explore the influence of the Chinese ST on the distinctive use of modal verbs in GPC interpreting, comparative ST-TT analysis was conducted on two predominantly deontic modals: should and must, which revealed that the ST was the major trigger of the use of these two modals, but a considerable number of “explicitation” shifts and “strengthening” shifts also contributed to the overuse of these two words. Drawing on the sociocultural context of GPC interpreting, three motivating factors were summarized to account for the distinctive features in the use of modal verbs in GPC interpreting: “the tenor difference between the Chinese GPC and the US GPC texts, the implicit nature of

2.2 Studies of Modality in Translation Between Chinese and English

27

the Chinese grammar, and the interpreter’s subjectivity” (ibid., 74). This study demonstrated the value of corpus tools in studying GPC interpreting, the value of comparative analysis of modality in terms of different semantic dimensions (such as modality value and modality type), and suggested the close relationship between the interpreting practice and its sociocultural context. Due to the focus on the linguistic features of the interpreted texts and the limitation of space, this study only selectively looked at the ST-TT relations from the use of two modal verbs in the TT, but it has already pointed towards the frequent occurrence of translation shifts in the transfer of modality in GPC interpreting. What patterns will emerge if we conduct a comprehensive ST-TT analysis concerning the expression of modality? How are these patterns related to the sociocultural context of interpreting? These will be the key questions to be explored in the present study. Armed with the modality theory of SFG, Fu (2013) analyzed the use of various types of modality markers in the interpreted text of Premier Wen Jiabao’s press conference during the “Two Sessions” in 2012. He found that various modal devices including modal adjuncts (i.e., modal adverbs), expansion of the predicator (i.e., structures containing modal adjectives or passive forms) and modal metaphors (e.g., I/we believe) in addition to modal verbs are skillfully used by the interpreter to realize the interpersonal meanings in the ST, and the majority of these devices express inclination and obligation in terms of modality type and express subjective modality in terms of orientation. Although it is a case study instead of a corpus-based study in the strict sense and the researcher focused on the skillful use of modal devices instead of modality shifts, his employment of the systemic functional approach of modality and his discussions about the “importance of skillful manipulation of various modal devices” (ibid., 24) to reconstruct interpersonal meanings of the GPC texts are enlightening to the present study. Based on the same corpus as used by Li and Hu (2013), Pan (2014) investigated the use of hedges in Chinese-English GPC interpreting as compared to their use in original English (American GPC texts) and written translation (English translation of Chinese Government Work Report texts). With “hedges” defined as “lexical markers of epistemic modality that expresses uncertainty” (ibid., 25, my translation), he found that “the frequency of hedges in interpreted English is significantly lower than that in original English discourse, but markedly higher than that in translated English” (ibid., 24). Comparative analysis of ST-TT relations in both GPC interpreting and Government Work Report translation revealed a significantly stronger tendency towards adding hedges (i.e., shifts of addition) in GPC interpreting (30.1%), compared with Government Work Report translation (7.1%). Various motivational factors for these differences were then discussed, including “the functions of hedges and characteristics of interpreting process, the genre difference between interpreted texts and translated texts, and the different rhetorical conventions between Chinese and English” (ibid., 24). Although Pan (ibid.) concentrated on the use of hedges as epistemic modality markers and downtoners, his findings is relevant to the present study on modality in two aspects: First, since 30.1% of all occurrences of hedges in the interpreted texts were found to involve shifts of addition, then it is reasonable to estimate the degree of semantic shifts in

28

2 Previous Literature on Interpreting and Modality

modality (which may not may not co-occur with shifts of addition) to be even higher, which will be tested in the present study. Second, in his discussion of motivational factors, Pan (ibid., 28) found that the use of Chinese subjective markers 我想/认为/觉得/看 in the STs of Chinese GPCs was far less frequent than the use of corresponding English subjective markers I think/I believe in American GPC texts, pointing to the different styles of Chinese and English political speeches: English political speeches are more personal and consultative than Chinese ones. This is relevant to the discussion of translation shifts and norms in the present study.

2.2.3

Contrastive Studies on Modality in Chinese and English Political Speeches

Besides the above studies on modality in a direct translational context, a number of corpus-based contrastive studies on modality in Chinese and English political speeches are also insightful for us to understand the treatment of modality in GPC interpreting, as the Chinese STs of GPCs are mostly political speeches made by Chinese government leaders. Three cases in point are Wang (2008), Wang et al. (2011) and Zhang (2013). With the aim to compare the ways top political leaders in China and the US practice public communication at GPCs, Wang (2008) explored the premiers/ presidents’ strategies to handle questions, themes of their speeches, their linguistic choices to perform institutional roles, their expressions of (un) certainty, and their ways of realizing interpersonal engagement etc. based on a corpus of 4 GPC texts involving 2 Chinese premiers and 2 American presidents. Among her findings, two points concerning modality are particularly relevant to the present study: (1) US leaders appeared to be more personal than the Chinese leaders as the former use parenthetical expressions (e.g., I think and I suppose) more frequently than the latter; (2) Chinese leaders preferred to use certainty expressions (i.e., high-value modality markers) in response to questions and US leaders seemed to make a balanced use of both certainty and uncertainty expressions. These findings indicate that American officials’ speeches at GPCs are more personal, tentative and consultative in style than the Chinese officials’ speeches, which conforms to suggestions by Li and Hu (2013), Pan (2014). With the more specific objective to compare how modal operators are used to realize interpersonal functions in English and Chinese political speeches, Wang et al. (2011) investigated the frequency and modality value of modal operators in a corpus of 10 American and 10 Chinese government leaders’ speeches on college campuses. It was found that modality operators are used far more frequently in English political speeches than Chinese ones, American leaders prefer to use low-value and median-value modal operators, but Chinese leaders frequently use both high-value and low-value modal operators but rarely use median-value ones.

2.2 Studies of Modality in Translation Between Chinese and English

29

These findings again indicate that compared with Chinese government leaders, American leaders are better at utilizing modal operators especially lower-value ones to make their speeches sound consultative and tentative, thus facilitating interpersonal communication between them and their audience. Drawing more closely on the modality theory from SFG, Zhang (2013) conducted a multi-dimensional contrastive analysis of “modal verbs” (it seems that the researcher used the term “modal verbs” to refer to various modal expressions including modal adverbs and mental lexical verb structures) and mood structures in Chinese and English political speeches based on a corpus of 15 speeches by Chinese government leaders and 15 speeches by American government leaders. Several findings are relevant to the present study: (1) in terms of modality type, probability modals were found to be more frequently used in English political speeches than in Chinese and obligation modals were used more frequently in Chinese political speeches than in English; (2) in terms of modality value, Chinese leaders appeared to use median-value modals more frequently than American leaders and American leaders appeared to use both low-value and high-value modals more frequently than Chinese leaders; (3) in terms of orientation, explicit subjective and implicit objective modal expressions are used more frequently in English political speeches than in Chinese ones. It is interesting that part of her contrastive findings on modality value contradict with Wang et al. (2011). Due to the unclear definition of modal verbs, the lack of test of statistical significance and the problematic coding for implicit objective modal expressions (e.g., the implicit subjective modals can and must were coded as implicit objective, see ibid., 36), some of these findings may not be reliable, but the several remarkable differences in her presented data conform to previous findings: Chinese political leaders tend to use more obligation modals than American leaders (ibid., 16); American leaders are more ready to use low-value modals than Chinese leaders (ibid., 24); American leaders tend to use more explicit subjective modal expressions than Chinese leaders (ibid., 32).

2.3

Summary

Through a critical review of literature on corpus-based interpreting studies and studies of modality relevant to translation between Chinese and English, this chapter has shown that it is both feasible and meaningful to conduct a corpus-based study of modality in GPC interpreting with the modality theory of SFG as the analytical framework. First, the relevant corpus-based interpreting studies reveal that it is feasible to apply the CTS methods to interpreting research to explore regular patterns in interpreters’ behaviors such as translation shifts, and then seek explanations for these regularities in the sociocultural context, but CIS projects on GPC interpreting in China so far have focused on translation shifts on the macro-level (i.e., the omission or addition of certain linguistic items/devices), with the degree and pattern of semantic shifts for critical linguistic phenomena such as

30

2 Previous Literature on Interpreting and Modality

modality underexplored. Second, previous studies on modality relevant to translation between Chinese and English reveal that the one-to-one equivalence between Chinese and English modality markers in translation is a myth and the modality theory of SFG has helped scholars identify semantic shifts of modality such as “strengthening” or “weakening” in translation and interpreting, but a comprehensive description of the patterns in modality shifts in terms of modality type, modality value and orientation has not been attempted yet. Considering the necessary limitation in the scope of any single corpus-based project, the present study will focus on the regular patterns of semantic shifts in the reconstruction of modality in Chinese-English GPC interpreting, which has not received much attention so far, and then seek explanations in the sociocultural context for these interpreting choices.

References Baker, Mona. 2004. Norms. In Routledge Encyclopedia of Translation Studies, ed. Mona Baker, 163. Shanghai: Shanghai Foreign Language Education Press. Baumgarten, Nicole, Bernd Meyer, and Demet Ozcetin. 2008. Explicitness in translation and interpreting: A critical review and some empirical evidence (of an elusive concept). Across Languages and Cultures 9 (2): 177–203. Beaton, Morven. 2007. Interpreted ideologies in institutional discourse: The case of the European Parliament. The translator 13 (2): 271–296. Bendazzoli, Claudio, Annalisa Sandrelli, and Mariachiara Russo. 2011. Disfluencies in simultaneous interpreting: A corpus-based analysis. In Corpus-based translation studies: Research and applications, ed. Alet Kruger, Kim Wallmach, and Jeremy Munday, 282–306. London/New York: Continuum. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 2000. Longman grammar of spoken and written english. Beijing: Foreign Language Teaching and Research Press. Cencini, Marco, and Guy Aston. 2002. Resurrecting the corp (us|se): Towards an encoding standard for interpreting data. In Interpreting in the 21st Century, ed. Giuliana Garzone, and Maurizio Viezzi, 47–62. Amsterdam/Philadelphia: John Benjamins. Dam, Helle V. 2002. Lexical similarity versus lexical dissimilarity in consecutive interpreting: A product-oriented study of form-based versus meaning-based interpreting. In The Interpreting Studies Reader, ed. Franz Pöchhacker, and Miriam Shlesinger, 266–277. London: Routledge. Diriker, Ebru. 2004. De-/Re-Contextualizing Conference Interpreting. Amsterdam/Philadelphia: John Benjamins. Facchinetti, Roberta, Manfred G. Krug, and Frank Robert Palmer. 2003. Modality in contemporary English. Mouton de Gruyter: In Topics in English Linguistics. New York. Fu, Rongbo 符荣波. 2013. 汉英口译语篇中的情态手段及其功能分析 [Modality in C-E consecutive interpreting: A systemic functional analysis]. 北京第二外国语学院学报 [Journal of Beijing International Studies University] (12): 24–29. Garzone, Giuliana. 2002. Quality and norms in interpretation. In Interpreting in the 21st Century: Challenges and opportunities, ed. Giuliana Garzone, and Maurizio Viezzi, 107–121. Amsterdam/Philedelphia: John Benjamins. Gile, Daniel. 1999. Norms in research on conference interpreting: A response to Theo Hermans and Gideon Toury. In Translation and norms, ed. Christina Schäffner, 98–105. Clevedon: Multilingual Matters Limited.

References

31

Gumul, Ewa. 2006. Explicitation in simultaneous interpreting: A strategy or a by-product of language mediation? Across Languages and Cultures 7 (2): 171–190. Halliday, Michael A.K. 1994. An introduction to functional grammar, 2nd ed. London: Edward Arnold. Halliday, Michael A.K., and Christian Matthiessen. 2004. An introduction to functional grammar, 3rd ed. London: Edward Arnold. Halliday, Michael A.K., and Christian Matthiessen. 2014. Halliday’s introduction to functional grammar, 4th ed. Abington/New York: Routledge. Harris, Brian. 1990. Norms in interpretation. Target 2 (1): 115–119. House, Juliane. 1997. Translation quality assessment. Narr: A Model Revisited. Tübingen. Hsieh, Chia-ling 谢佳玲. 2006. 汉语情态词的语意界定: 语料库为本的研究 [The semantic categorization of Chinese modal expressions: A corpus-based analysis]. 中国语文研究 [Chinese Language Studies] (1): 45–63. Hu, Kaibao 胡开宝, and Qing Tao 陶庆. 2010. 汉英会议口译语料库的创建与应用研究 [The compilation and application of Chinese-English Conference Interpreting Corpus]. 中国翻译 [Chinese Translators Journal] (5): 49–56. Hu, Kaibao 胡开宝, and Qing Tao 陶庆. 2012. 记者招待会汉英口译句法操作规范研究 [Syntactic operational norms of press conference interpreting (Chinese-English)]. 外语教学与 研究 [Foreign Language Teaching and Research] (5):738–750, 801. Hu, Kaibao 胡开宝, and Lixin Xie 谢丽欣. 2014. 记者招待会汉英口译中英语关系从句的应用 研究 [The use of English relative clauses in Chinese-English conference interpreting]. 山东外 语教学 [Shandong Foreign Languages Teaching Journal] (4): 8–16. Kalina, Sylvia. 1998. Strategische Prozesse beim Dolmetschen. Tübingen: Gunter Narr. Klinge, Alex. 1993. The English modal auxiliaries: From lexical semantics to utterance interpretation. Journal of Linguistics 29 (2): 315–357. Laviosa, Sara. 1998. Core patterns of lexical use in a comparable corpus of English narrative prose. Meta 43 (4): 557–570. Lederer, Marianne. 1981. La Traduction Simultanée. Paris: Minard Lettres modernes. Leech, Geoffrey. 2011. The modals ARE declining–reply to Neil Millar’s Modal verbs in TIME: Frequency changes 1923–2006. International Journal of Corpus Linguistics 14: 2 (2009), 191– 220. International Journal of Corpus Linguistics 16 (4): 547–564. Li, Charles, and Sandra Thompson. 1989. Mandarin Chinese: A functional reference grammar. Berkeley: University of California Press. Li, Kexing 李克兴. 2007a. 英语法律文本中主要情态动词的作用及其翻译 [Functions of modal verbs in legal documents and their translation]. 中国翻译 [Chinese Translators Journal] (6): 54–60, 96. Li, Lijuan 李丽娟. 2012. Can 和Could 所表情态的汉译研究——基于《傲慢与偏见》平行语 料 [Study on Chinese translation of modality by can and could based on Parallel Corpus of Pride and Prejudice]. MA thesis, Dalian Maritime University. Li, Renzhi. 2003. Modality in English and Chinese: A typological perspective. Ph.D. thesis, Universitaire Instelling Antwerpen. Li, Sha’ning 李莎宁. 2007b. 汉语外交演说及其英语译文中情态动词的比较研究 [A comparative study of modal verbs in Chinese diplomatic speeches and their English versions]. MA thesis, National University of Defense Technology. Li, Xiaochuan 李小川. 2005. 情态动词的语义分析及翻译研究 [Semantic analyses and translation studies of the English modal verbs]. 长沙理工大学学报(社会科学版) [Journal of Changsha University of Science and Technology (Social Science)] (4): 117–120. Li, Xin 李鑫. 2011. 基于语料库的汉英会议口译中被动式的应用及其动因研究 [A corpus-based study of the use of passive construction and its motivation in Chinese-English conference interpretation]. 当代外语研究 [Contemporary foreign languages studies] (1): 18–23.

32

2 Previous Literature on Interpreting and Modality

Li, Xin 李鑫, and Kaibao Hu 胡开宝. 2013. 基于语料库的记者招待会汉英口译中情态动词的 应用研究 [A Corpus-based Study of modal verbs in Chinese-English government press conference interpretation]. 外语电化教学 [Computer-assisted Foreign Language Education] (3): 26–32, 74. Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese-English): Face management and information management. Ph.D. thesis, Heriot-Watt University. Luo, Wenyi 罗汶宜. 2011. 基于语料库的情态动词英汉翻译研究 [A study on the translation of modal verbs from English to Chinese]. MA thesis, Dalian Maritime University. Meyer, Bernd. 2008. Interpreting proper names: Different interventions in simultaneous and consecutive interpreting. Trans-kom 1 (1): 105–122. Millar, Neil. 2009. Modal verbs in TIME: Frequency changes 1923–2006. International journal of corpus linguistics 14 (2): 191–220. Mindt, Dieter. 1995. An empirical grammar of the english verb: Modal verbs. Berlin: Cornelsen. Monacelli, Claudia. 2005. Surviving the role: A corpus-based study of self-regulation in simultaneous interpreting as perceived through participation framework and interactional politeness. Ph.D. thesis, Heriot-Watt University. Pöchhacker, Franz. 1994. Simultandolmetschen als komplexes Handeln. Tübingen: Gunter Narr. Palmer, Frank. 1990. Modality and the English Modals. London: Longman Group Limited. Pan, Feng 潘峰. 2014. 基于语料库的汉英会议口译中模糊限制语的应用研究 [A corpus-based study of the application of hedges in Chinese-English Conference Interpreting]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (4): 24–29. Pan, Feng 潘峰, and Kaibao Hu 胡开宝. 2013. 基于语料库的汉英会议口译中 some 的应用研 究 [A corpus-based study on the use of some in Chinese-English conference interpreting]. 解放 军外国语学院学报 [Journal of PLA University of Foreign Languages] (2): 95–100. Peng, Lizhen 彭利贞. 2007. 现代汉语情态研究 [On Modality of Modern Chinese]. Beijing: China Social Sciences Press. Petite, Christelle. 2005. Evidence of repair mechanisms in simultaneous interpreting: A corpus-based analysis. Interpreting 7 (1): 27–49. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Russo, Mariachiara, and Annalisa Sandrellii. 2006. Looking for lexical patterns in a trilingual corpus of source and interpreted speeches: Extended analysis of EPIC (European Parliament Interpreting Corpus). Forum 4 (1): 221–255. Sandrelli, Annalisa, and Claudio Bendazzoli. 2005. Lexical patterns in simultaneous interpreting: A preliminary investigation of EPIC. In Corpus Linguistics Conference Series. The University of Birmingham. Schjoldager, Anne. 1995. An exploratory study of translational norms in simultaneous interpreting: Methodological reflections. Hermes 14: 65–87. Seleskovitch, Danica. 1975. Langage, langues et mémoire: étude de la prise de notes en interprétation consécutive. Paris: Minard Lettres modernes. Setton, Robin. 2009. Interpreting China, interpreting Chinese. Interpreting 11 (2): 109–117. Setton, Robin. 2011. Corpus-based interpreting studies (CIS): Overview and prospects. In Corpus-based translation studies: Research and applications, ed. Alet Kruger, Kim Wallmach, and Jeremy Munday, 33–75. London/New York: Continuum International. Shlesinger, Miriam. 1989a. Extending the theory of translation to interpretation: Norms as a case in point. Target 1 (2): 111–115. Shlesinger, Miriam. 1989b. Simultaneous interpretation as a factor in effecting shifts in the position of texts on the oral-literate continuum. MA thesis, Tel Aviv University. Shlesinger, Miriam. 1991. Interpreter latitude versus due process: simultaneous and consecutive interpretation in multilingual trials. In Empirical Research in Translation and Intercultural studies, ed. Sonja Tirkkonen-Condit, 147–155. Tübingen: Gunter Narr Verlag. Shlesinger, Miriam. 1995. Shifts in cohesion in simultaneous interpreting. The Translator 1 (2): 193–214.

References

33

Shlesinger, Miriam. 1998. Corpus-based interpreting studies as an offshoot of corpus-based translation studies. Meta 43 (4): 486–493. Shlesinger, Miriam. 2008. Towards a definition of Interpretese: an intermodal, corpus-based study. In Efforts and models in interpreting and translation research: A tribute to Daniel Gile, ed. Gyde Hansen, Andrew Chesterman, and Heidrun Gerzymisch-Arbogast. Amsterdam/ Philadelphia: John Benjamins. Shlesinger, Miriam, and Noam Ordan. 2012. More spoken or more translated? Exploring a known unknown of simultaneous interpreting. Target 24 (1): 43–60. Straniero Sergio, F. 2003. Norms and quality in media interpreting: The case of formula one press conferences. The Interpreters’ Newsletter 12: 135–174. Straniero Sergio, F., and Caterina Falbo. 2012. Studying interpreting through corpora. An introduction. In Breaking Ground in Corpus-based Interpreting Studies, ed. F. Straniero Sergio, and C. Falbo, 9–52. Bern: Peter Lang. Sun, Tingting. 2012. Interpreters’ mediation of Government Press Conferences in China: Participation framework, footing and face Work. Ph.D. thesis, The University of Manchester. Sun, Tingting. 2013. Interpreting China: interpreters’ mediation of government press conferences in China. Beijing: Foreign Language Teaching and Research Press. Sun, Ximin 孙喜民. 1995. 浅谈 shall 在法律、合同文件中的译法 [A preliminary analysis of the translation of shall in legal documents]. 中国翻译 [Chinese Translators Journal] (4): 59–60. Toury, Gideon. 1980. In search of a theory of translation. Tel Aviv: Porter Institute. Toury, Gideon. 1995. Descriptive translation studies and beyond. Amsterdam/Philadelphia: John Benjamins Publishing Company. Toury, Gideon. 1999. A handful of paragraphs on ‘translation’ and ‘norms’. In Translation and norms, ed. Christina Schäffner, 9–31. Clevedon: Multilingual Matters Limited. Wang, Binhua 王斌华. 2009. 口译规范描写及其应用——一项基于中国总理“两会”记者招待 会交传语料的研究 [Description of norms in interpreting and its application–a study based on the corpus of consecutive interpreting in Chinese premier press conferences]. Ph.D. thesis, Guangdong University of Foreign Studies. Wang, Binhua. 2012. A descriptive study of norms in interpreting: Based on the Chinese-English consecutive interpreting corpus of chinese premier press conferences. Meta 57 (1): 198–212. Wang, Binhua, and Tao Li. 2014. An empirical study of pauses in Chinese-English simultaneous interpreting. Perspectives 23 (1): 124–142. Wang, Hesi 王和私, Pi’an Yin 尹丕安, and Furong Wang 王芙蓉. 2011. 中英文政治演说的情 态对比研究 [A comparative study on modality in Chinese and English political speeches]. 西 安外国语大学学报 [Journal of Xi’an International Studies University] (2): 38–42. Wang, Peng 王鹏. 2004. 《哈利波特》与其汉语翻译 [Harry Potter and its Chinese translation]. Ph.D. thesis, Sun Yat-sen University. Wang, Peng 王鹏. 2007. 《哈利波特》与其汉语翻译 [Harry Potter and its Chinese translation]. Chongqing: Chongqing University Press. Wang, Wangni 王望妮. 2008. 中美政府首脑记者招待会语言对比研究 [Evaluating the way government practice public communication: A comparative approach]. Ph.D. thesis, Shanghai International Studies University. Xie, Lixin 谢丽欣, and Kaibao Hu 胡开宝. 2015. 记者招待会汉英口译中不定量词的应用研究 [The use of English indefinite quantifiers in Chinese-English conference interpreting]. 外语电 化教学 [Computer-assisted Foreign Language Education] (1): 17–22. Xu, Jingning 徐晶凝. 2008. 现代汉语话语情态研究 [On Discourse Modality of Modern Chinese]. Beijing: Kunlun Publishing House. Xue, Muxuan 薛慕煊. 1999. 英语情态动词的正确理解和翻译 [Correct understanding and translation of English modal verbs]. 上海科技翻译 [Shanghai Journal of Translators for Science and Technology] (1): 4–6. Zhang, Guanfang 张冠芳. 2013. 系统功能语言学视角下的英汉政治演讲的情态对比研究 [A contrastive analysis on modality in English and Chinese political speeches from the perspective of Systemic Functional Linguistics]. MA thesis, Ningbo University.

34

2 Previous Literature on Interpreting and Modality

Zhang, Hanbing 张寒冰, and Lizhen Peng 彭利贞. 2016. 现代汉语情态与指称的同现限制研究 [Co-occurence restriction of modality and reference in Mandarin Chinese]. 语言文字应用 [Applied Linguistics] (2): 143. Zhao, Qiurong 赵秋荣, and Maocheng Liang 梁茂成. 2013. 认识型情态动词 may 和 might 汉译强 度变化研究 [The modal strength of epistemic verbs may and might in English-Chinese Translation]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (6): 96–99, 108. Zhao, Yujuan 赵宇娟. 2009. 英语情态动词的汉译: Will 与 Would [Realizations of modal auxiliaries in E-C translation: with will and would in focus]. MA thesis, Qufu Normal University. Zhu, Dexi 朱德熙. 1982. 语法讲义 [The Lecture Notes of Grammar]. Beijing: The Commercial Press. Zhu, Guanming 朱冠明. 2005. 情态与汉语情态动词 [Modality and Chinese modal auxiliaries]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (2): 17–21.

Chapter 3

Theoretical Framework and Methodology

3.1

Theoretical Framework

As the current study attempts to describe the modality shifts in GPC interpreting based on the corpus data and then seek explanations for the linguistic regularities in the sociocultural context, there are three major sources of theoretical insights leading to my descriptive-explanatory analysis: the broad framework of DTS from Toury (1995), the systemic functional view of modality from Halliday and Matthiessen (2014), and the integrated sociocultural context analysis inspired by Diriker (2004), Alexieva (1997) and Pöchhacker (2009).

3.1.1

Toury’s Framework of DTS

Although Toury (1978, 1980, 1995) himself hasn’t used any electronic corpus in his various case studies on literary translation, his “discovery procedures” for “translational norms”, comparative model of the ST-TT “coupled pair”, and “target-oriented” approaches under the umbrella of DTS prove to be highly insightful for and compatible with corpus-based translation and interpreting studies. In fact, it was the integration of Descriptive Translation Studies and corpus linguistics that gave rise to the paradigm of Corpus-based Translation Studies (Laviosa 2004; Hu 2012). The most important contribution of Toury’s DTS framework to translation and interpreting studies lies in that his approach “offers a way of escape from the tradition of prescriptive studies” (Chesterman 1999, p. 90). Instead of prescribing what is good translation and what kind of equivalence should be achieved, translation scholars are encouraged to step back and observe translation as a “norm-governed activity” (Toury 1995, p. 56) in its sociocultural and historical context. Borrowing the definition from sociology, Toury (ibid., 51) explains norms © Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_3

35

36

3 Theoretical Framework and Methodology

as “the translation of general values or ideas shared by a certain community—as to what is right and wrong, adequate and inadequate—into performance instructions appropriate for and applicable to particular situations, specifying what is prescribed and forbidden as well as what is tolerated and permitted in a certain behavioural dimension”. Therefore, “norms” do exert prescriptive pressure on translators, but “it is the norms that do the prescribing, not the scholars” (Chesterman 1999, p. 90). To offer a systematic methodology for DTS, (Toury 1995, pp. 38–39; 102) proposed a series of “discovery procedures” with the aim of reconstructing translational norms: (1) Situate the TT within the target system, and mainly address the question of acceptability (usu. focus on one linguistic phenomenon at a time, such as conjoint phrases). (2) Map the TT onto ST and identify translation relationships (esp. translation shifts) between coupled pairs of ST and TT segments. (3) Attempt generalizations (explanatory hypotheses) about the underlying concept of translation. (4) Extend the corpus, repeat (1) and (2) with additional texts of similar type and strive for higher-level generalizations about translational norms (for a certain period, or school of translators, etc.). The first step signals a clear “target-orientedness” of Toury’s DTS approach, which emphasizes translation product as a valid object of study in its own right. But this “target-orientedness” only means TTs and target cultures are “where its observations start”, and “neither source text nor transfer operations and transferred features, nor even translation relationships, would have been excluded from a target-oriented program of DTS” (ibid., 36). Thus the second step centers on the comparison between ST-TT coupled pairs for translation shifts. One example for this ST-TT comparison is given by Toury concerning the translation of metaphor. As Toury (1995, pp. 82–85) summarizes, the coupled-pair relationships relevant to “any study of metaphor-and-translation” include six types: (1) metaphor into ‘same’ metaphor; (2) metaphor into ‘different’ metaphor; (3) metaphor into non-metaphor; (4) metaphor into 0; (5) non-metaphor into metaphor; (6) 0 into metaphor. However, only the first three types of relationships, or at best the first four, are included by TS scholars adopting the traditional source-oriented approach. Therefore, this comparative model of ST-TT coupled pair that involves proceeding from the TT pole as well as the ST pole is particularly enlightening to my research on modality in GPC interpreting. In order to have a comprehensive picture of how modality is reconstructed, I will trace modality markers from ST to TT as well as the other way around. The 3rd and 4th step shows that it is not sufficient to study translational norms through observing just one pair of texts. Toury emphasizes the importance of extending the “corpus” from one pair to additional pairs of texts, but his way of using “corpus” is simply through the case study of one pair of texts after another to

3.1 Theoretical Framework

37

test the explanatory hypotheses formulated by previous studies. Without the assistance of the electronic corpus in the modern sense, it will be tedious and arduous to compare numerous pairs of texts even for one linguistic phenomenon before we can reach any conclusion about translational norms. This may explain why Toury (1978, 1980, 1995) only offered case studies involving very limited textual data despite his acknowledgement of the importance of a large corpus. Fortunately, this dilemma can now be solved with the availability of the translational corpora in the modern sense, as we can at least focus on one linguistic phenomenon at one time and extract parallel concordances from a fairly large corpus conveniently for comparative analysis. Toury’s descriptive stance and comparative model for translation shifts proceeding from both ST and TT poles offer important guidance for the current study. But there are two points where I have to deviate from him. First, although Toury’s the target-orientedness doesn’t exclude the observation of the ST, his positioning of translations as only “facts of target cultures” (Toury 1995, p. 29) prevents him from recognizing the important role of the initiator, commissioner, or monitoring institutions that may exist in the source culture, as in the case of GPC interpreting in China. This study will adopt a broader understanding of target-orientedness, which means giving equal status to ST and TT and acknowledging TT as an actual translation product despite its deviation from ST. Second, Toury’s concept of “translational norms”, though constructive, is still too broad and ambiguous, covering a series of sub-notions as “operational norms”, “initial norm”, and “preliminary norms” (ibid., 56–58). On one hand, Toury (1999, p. 15) states the difference between “norms” and “regularities of behavior”, in that “whatever regularities are observed, they themselves are not the norms” and they are “only external evidence” of norms which are “still to be extracted”. On the other hand, in Toury’s case studies (1978, 1980, 1995) he fails to draw a clear line between the two concepts or give instructions on how to extract one from the other, which may lead to other scholars’ (see Baker 2004, p. 163) assumption that the regularities of translation behavior themselves are what he calls “translational norms”. In addition, the exploration of “norms” seems to be a descriptive end according to Toury’s (1995, p. 38) “discovery procedures”, but he also emphasizes that norms are only “explanatory hypotheses” (ibid., 59) rather than social realities. How can the same concept work as a descriptive end and an explanatory tool which is not even given the status of entity? To avoid this ambiguity, I will adopt the definition of norms by Bartsch (1987, p. xiv) as the “social reality of correctness notions” which can be directly extracted from “extratextual” (Toury 1995, p. 65) sources. Drawing on Toury’s descriptive stance on actual translations in specific sociocultural contexts but trying to draw a clearer line between description and explanation, I will look for only “translation shifts” and “regularities” in my description of textual data and leave the concept of norms to the explanation stage along with other contextual factors.

38

3.1.2

3 Theoretical Framework and Methodology

The Semantic System of Modality

As mentioned in Sect. 1.1.3, I choose to focus on the grammatical-semantic category of modality in this research because of its important interpersonal function in the political discourse of GPC and the realizational choices it involves in GPC interpreting. For an in-depth analysis of the semantic shift in the reconstruction of modality, I turn to the modality system as depicted in the Systemic Functional Grammar (SFG) proposed by M.A.K. Halliday (see Halliday 1985, 1994; Halliday and Matthiessen 2004, 2014). Although Chinese and English belong to two different families, they do present remarkable similarities in the system of modality, as noticed by Halliday himself (Halliday and McDonald 2004) as well as Chinese scholars (Zhu 1996; Hu 1999; Peng 2000; Wei 2008; Liu 2009). I will present the English modality system based on the latest version of SFG (Halliday and Matthiessen 2014) and discuss its application to Chinese based on the above-mentioned works as well as my observation of the corpus data. In SFG, the role of modality system is to “to construe the region of uncertainty that lies between ‘yes’ and ‘no’” (ibid., 176). As Fig. 3.1 illustrates, the semantics of

probability modalization

MODALITY TYPE

usuality obligation

modulation

inclination subjective ORIENTATION

objective

modality

explicit implicit VALUE

median outer

high low

POLARITY

positive negative

direct transferred

Fig. 3.1 The system of modality in Halliday and Matthiessen (2014, p. 182)

3.1 Theoretical Framework

39

modality covers four dimensions: modality type, orientation, modality value and polarity. The relationship between modality and negation has been widely acknowledged as a highly complex issue that deserves separate treatment in modality studies (de Haan 1997; van der Auwera 2001; Palmer 1995, 2003; Usoniene and Šoliene 2010). Considering the fact that negative forms of modality are rather limited in my corpus data, I will confine my analysis to only the positive forms for the present study, thus the dimension of polarity will not be discussed here. First, modality is divided into two broad types—modalization and modulation— according to the speech functions it typically realizes in a clause. Speech functions here refer to whether language is used to exchange information or exchange goods and services. The former case is called proposition, which is functioning as a statement (giving information) or question (demanding information). The latter is called proposal, which is functioning as offer (giving goods and services) or command (demanding goods and services). Modalization typically expresses the intermediate possibilities between “it is so” and “it isn’t so” in a proposition, and it covers two sub-types of modality: probability (possibly/probably/certainly) and usuality (sometimes/usually/always). Modulation typically expresses the intermediate possibilities between “do it” and “don’t do it” in a proposal and it also covers two sub-types of modality: obligation (allowed to/supposed to/required to, as in a command) and inclination (willing to/anxious to/determined to, as in an offer). Probability and usuality can be expressed by a finite modal operator (i.e., modal verb: that will be John; he’ll sit there all day), a modal adjunct (i.e., modal adverb: that’s probably John; he usually sits there all day), or both together (that’ll probably be John; he’ll usually sit there all day). Obligation and inclination can be expressed by a modal verb (you should know that; I’ll help them) or by an expansion of the predicator, typically a passive construction (you’re supposed to know that) or an adjective (I’m anxious to help them) (Halliday and Matthiessen 2014, pp. 177–178). So in the end there are four modality types distinguished: probability, usuality, obligation and inclination. Among them three types—probability, obligation and inclination—correspond respectively to Palmer’s (1990) epistemic, deontic and dynamic modality, and the type of usuality is added, which is often realized by modal adverbs (usually, always, often, seldom) instead of modal verbs. By extending the semantics of modality to include usuality and extending the realizations of modality beyond modal verbs, SFG offers a broader view of modality than traditional grammar. By relating modality types to speech functions, SFG helps us understand the important interpersonal role played by modal expressions in our communication. As Thompson summarizes (2008, p. 67), in exchanging information, “the modality relates to how valid the information is being presented as in terms of probability or usuality”; while in exchanging goods and services, “modality relates to how confident the speaker can appear to be in the eventual success of the exchange”. However, we have to notice that the relationship between speech functions, modality types and realizational forms described here is a prototypical one, rather than a comprehensive picture. For example, the use of inclinational will or be going to with a third-person subject often functions as a proposition instead of proposal as the speaker is not offering goods and services but

40

3 Theoretical Framework and Methodology

giving information about his/her plan. Thus it will be misleading to identify modality types only based on the speech functions and grammatical forms of modality. Probably this is why the table on this simplified relationship available in the earlier versions of the Introduction to Functional Grammar (Halliday 1985, p. 86; 1994, p. 91) disappeared in later versions (Halliday and Matthiessen 2004, 2014). Therefore, to identify the semantic type of a modality marker, we can use this typical relationship between speech function and modality type as a reference, but the final say is still in the actual linguistic context around the modal expressions. According to Halliday and McDonald (2004, p. 343) this four-type classification of modality is fully applicable into Chinese, although some scholars may deny the existence of usuality in Chinese modality because it is not realized by any Chinese modal verb (Zhu 1996, p. 7). Acknowledging non-modal-verb forms like modal adverbs as modality markers, I sympathize with Halliday and McDonald (2004) and recognize usuality as one semantic type of modality typically marked by modal adverbs (e.g., always, 一直 (always)) for both English and Chinese. With the help of contextual information, the above system of modality types can be readily used to analyze most of modal expressions in the corpus, except for the “ability/ potentiality” sense of can or be able to, which is considered by Halliday and and Matthiessen (2014, p. 696) as “on the fringe of the modality system”. As the pure “ability”-sense use of modality markers in both English and Chinese rarely occurs in the corpus, I will discard these rare cases of “pure ability” markers (e.g., can, be able to, 能/能够 (can), 会 (can)) and code other cases as “probability”, “obligation” or “inclination” (which is often mixed with the “ability” sense) depending on the linguistic context. Table 3.1 shows examples for these four types of modality extracted from my corpus. The second dimension in the modality system is orientation, which distinguishes two pairs of opposing features in the presenting form of the modal expression: subjective vs. objective, and implicit vs. explicit. Depending on the orientation, there are four different ways of presenting modality: implicit subjective (John may be right; John should go), explicit subjective (I think John is right; I want John to go), implicit objective (John probably knows; John’s supposed to go) and explicit objective (It’s possible that Johns knows; It’s expected that John goes). However, not every type of modality can be realized by all the four different orientations. For example, “there are no systematic forms for making the subjective orientation explicit in the case of usuality or inclination” (Halliday and Matthiessen 2014, p. 692). Note that in essence all modal expressions are subjective, because they all express the speaker’s opinion or attitude, but the subjective evaluation of the speaker can be either explicit or disguised in some ways to sound as if they are the statements of objective facts, which shows “the extent to which the speaker openly accepts responsibility for the subjective assessments being expressed” (Thompson 2008, p. 70). This dimension is interesting to explore in translation and interpreting, as the translator/interpreter may render an explicit subjective modal expression into an apparently objective one, or vice versa, for ideological reasons. Although these objectivized forms of modality are not as readily available in Chinese as in English, I do find examples in the corpus for the different orientations of Chinese modality

3.1 Theoretical Framework

41

Table 3.1 Modality types in English and Chinese Modality type

Realization

Example in the English TT

Example in the Chinese ST

Probability

Modal verb

Will, can, would, may, could Be going to, be able to I/we think, I/we believe Certainly

会 (Can/will), 能/能够 (can), 可能 (may) 一定 (Must/definitely)

Possible

可能的 (Possible)

Always

一直 (always)

Should, must

要 (should/need to), 可以 (can/may), 应该 (should) 需要 (Need to) (+verb), 必须 (must/have to) 要 (will)

Semi-modal

Usuality Obligation

Inclination

Mental verb + clause Modal adverb Modal adjective Modal adverb Modal verb Semi-modal

Need to

Modal verb

Will, would, could Be going to, want to

Semi-modal Modal adverb Mental verb Modal idiom

我想 (I think), 我/我们认为 (we think), 我/我们相信 (I/we believe) 将 (will)

将 (Will) 想 (Want to) (+verb) Would like to

expressions (c.f. Wei 2008). For the present study, I will see the orientation of modality as a continuum from the “explicit subjective” to “explicit objective” and add another category “polar objective” at the objective end because a statement without any modality marker is in the most extreme form of “fact” rather than “opinion”. Some typical realizations in English and Chinese for the different orientation are illustrated in Table 3.2 with examples from the corpus. Because there is no clear-cut grammatical difference between Chinese modal verb and modal adverb (for example, opinions are divided on whether 一定 (must/definitely) is a modal verb or modal adverb; see Peng 2007) as that in English, in actual analysis, no difference will be distinguished in terms of orientation between modal verbs and modal adverbs for Chinese modality. The third dimension is modality value, referring to the strength attached to the modal judgment, and it is presented in three scales—high, median or low—in Halliday and Matthiessen (2014, p. 694). These labels will be useful for us to investigate the degree of the speaker’s commitment, i.e., “the degree to which the speaker commits herself to the validity of what she is saying” (Thompson 2008, p. 69). But as Thompson (ibid.) reminds us, “this is an idealization”, and “the three

It is +modal/evaluative adjective +clause

Modal noun Proposition without modality marker

Polar objective

Be+past participle/adjective to

Explicit objective

Implicit objective

Modal adverb

1st-person subject +mental lexical verb +clause 1st-person subject+be +modal adjective +clause Prepositional phrase with 1st-person possessive pronoun Modal/Semi-modal

Explicit subjective

Implicit subjective

Realization in English

Orientation

Well, to answer this question, it may take a bit more time We also need to actively advance reform in the political system But certainly there also exist some differences between China and the United States And in 2009, China is expected to become the largest greenhouse gas emitter in the world We are willing to continue maintaining contacts with Taiwan authorities… …Given the huge debt level of the United States, it is likely that the US dollar will further depreciate… It is important for leaders of the BRICS countries to step up their exchanges and dialogue. There is still possibility for a political solution. This year, we have three major tasks

In my view, our relations with the United States up to now are very good

I’m sure there’re a lot of things to say…

I think this is a very important question.

Example

Proposition without modality marker

Modal noun

Modal adverb

Modal/semi-modal

1st-person subject +mental lexical verb +clause

Realization in Chinese

Table 3.2 Orientation scale from modal subjectivity to polar objectivity in English and Chinese

我有一个对 台政策的问题。

使用政治外交手段解决问题的可 能性 (possibility) 仍然是存在的。

解决民生问题, 要 (should/need to) 首先着眼于 困难群体。 各项准备工作 加紧进行, 年内一 定 (must) 开工。 贫富差距的扩大在一定的历史时 期里面, 也许 (perhaps/maybe) 是 不可避免的。

我认为 (I think) 这个政策还是积 极可行的。 我相信 (I believe) 这是我们相互 沟通的一个大好的机会。 我希望 (I hope) 双方对这一点都 看得很清楚。

Example

42 3 Theoretical Framework and Methodology

3.1 Theoretical Framework

43

labels represent areas on a scale rather than absolute categories, with more delicate distinctions possible”. Based on Halliday and Matthiessen (ibid., 145; 694; 696) and actual data in the corpus, a more delicate modality value scale is proposed in the present study, with examples of high-frequency English and Chinese modal expressions for 6 different values summarized in Table 3.3. Some modal forms such as will, 要 and 愿意 are fuzzier than others in modality value, so I assign a range of values to them for more flexible coding. Besides these 6 values, an “extra high” value is added here to indicate our agreement with Halliday and Matthiessen (2014, p. 177) that “even a high value modal (‘certainly’, ‘always’) is less determinate than a polar form”. There is actually a complex interplay between negation and modality values, but since the present study is confined to only the positive forms of modal expressions, the discussion of this complexity (e.g., the value of can’t is higher than can) will be avoided. In the present study, I will focus on the high-frequency modality markers in the Chinese ST and the English TT and explore the degree/types of translation shifts between the “coupled pairs” mainly in terms of the three dimensions—modality type, orientation and modality value—in Chinese-English GPC interpreting. Note that in actual analysis for modality shifts, various modality markers not covered in Tables 3.1, 3.2 and 3.3 will be involved, so sometimes the coding has to be established on an ad hoc basis, but these tables will be used as benchmarks and I will also consult dictionaries and bilingual colleagues whenever there is any uncertainty. Inter-coder reliability will also be tested on a data sample before the coding of the whole corpus. The identification of high-frequency modality markers in my corpus and the coding scheme for modality shifts will be discussed in Sect. 3.2.

3.1.3

The Sociocultural Context of Interpreting

The close relationship between language and context has long been acknowledged by linguists and thinkers such as Malinowski (1923), Austin (1962), Hymes (1964), Halliday (1978), Martin (1999) and van Dijk (2001). However, despite the extensive discussions, “there is strictly speaking no theory of what exactly a ‘context’ is” (van Dijk 1998, p. 211, original emphasis). Generally speaking, there are two major strands of thought on context: the cognitive-oriented and sociology-oriented perspectives. Compared with the cognitive view of context as “mental constructs of participants” (van Dijk 2001, p. 18), the sociological view of context as “a series of pre-existing entities and relations in the real world” (Baker 2006, p. 323) is more relevant to the present study, although I might as well “take into consideration cognitive factors of communicating parties” (Monacelli 2005, p. 35) in the accounts of the sociocultural context. Among the sociology-oriented theories of context, the systemic functional linguistic views of context have been widely applied to translation studies (Bell 1991;

0 Extra high

Positive form without modality marker

Positive form without modality marker

Modality value

Realization in English

Realization in Chinese

必须 (Must, have to) 一定 (must, definitely) 我/我们相信 (I/we believe)

Must, always, I/we believe

1 High

Table 3.3 Modality value scale in English and Chinese 3 Medium

Will, be going to Need to Should, want to, I/we think, I/we hope 要 (Wil, need to, should, want to) 将 (will, be going to) 会 (will, be going to) 需要 (Need 应该 (Should) to) 我/我们想 (I/we think) 我/我们认为 (I/we think) 我/我们希望 (I/we hope)

2 Medium-high

May, can, be able to, possible

5 Low

可以 (Can, maybe able to) 可能 (May, can, maybe) 能/能够 (Can, be able to)

愿意 (Be willing to, be ready to, would like to)

Would

4 Medium-low

Could

6 Extra low

44 3 Theoretical Framework and Methodology

3.1 Theoretical Framework

45

House 2006; Zhang 2001; Xu 2003). Following the anthropologist Malinowski’s (1923) ideas about language in relation to context of culture and context of situation, Halliday (1978) developed the register theory by subcategorizing context of situation into three variables, i.e., field (what is going on), tenor (the way you relate to other participants) and mode (the channel you choose to communicate), and Martin (1992) further theorized context of culture into a system of genres, with genre defined as a “staged, goal-oriented social process realized through register”(ibid., 505). Although inspirational to the present study, these systemic functional linguistic theories of context designed for monolingual analysis prove to be more applicable to translation studies than interpreting studies. Because the TT of written translation is often produced at a different setting from the ST, translation can be conveniently viewed as a “recontextualization” process (House 2006, p. 338) and scholars (e.g., Zhang 2001; Xu 2003) can naturally view the register and genre of the ST and the TT separately and compare the “first order context” and the “second order context” (Halliday 1992, p. 25) to account for the specific translating choices. However, different from written translation, the TT of interpreting is produced in the same temporal and spatial setting as the ST and the primary participants are all present in the communication, thus it is not so convenient and meaningful to discuss the context of the ST and TT separately as to analyze the relevant contextual factors (e.g., the power relationship between participants, the temporal and spatial setting, etc.) of the interpreting event in an integrated manner. Fortunately, integrated models for analyzing the sociocultural context of interpreting have been suggested by interpreting scholars, among which the context analysis attempted by Diriker (2004) is particularly illuminating to the present study. In her analysis of the interplay between SI behavior and its sociocultural context, Diriker (2004) explored both the broad context and the immediate context of the interpreting events. First, the broad context was approached through the meta-discursive representation of SI, which “does not embody just a neutral description of intrinsic professional features, but presents a selected and hierarchized set of ‘norms’” (ibid., 25). Her analysis of the meta-discourse covered various sources related to “actors inside and outside the profession” (ibid., 26), including general reference books, codes of ethics, professional organizations, academia, Turkish media and a popular book authored by a professional interpreter. These sources were later categorized into two levels because of discrepancies between them. The general and de-contextualized discourse on SI such as general reference books, media, the discourse of professional associations, ethical codes and the discourse from academia was found to depict simultaneous interpreters as “competent professionals who can identify with the speakers, replace them in the eyes of the audience, and unproblematically access and transfer the original meaning fluently and completely” (ibid., 48). By contrast, the specific and contextualized meta-discourse on SI as represented by the anecdotal accounts of the interpreters on their personal interpreting experiences in real-life situations was found to recognize “the active involvement of the interpreters in shaping the message” (ibid., 49).

46

3 Theoretical Framework and Methodology

Second, the immediate context of SI was analyzed mainly based on “ethnographic material (field observations and interviews)” and “transcribed recordings of actual SI performance” (ibid., 17) with the focus on a specific conference interpreting event (“International Philosophy Colloquium—Martin Heidegger and Hannah Arendt: Metaphysics and Politics”) involving two simultaneous interpreters. To be specific, the analysis of the immediate context covers the topic, date and venue of the conference, conference languages, interpreters, organizer and the recruitment process, technical matters, position of the booths, the researcher’s position, participants of the conference, and preparations of the interpreters, interviews with the organizer, the speakers, the interpreters and the users of SI. All these specific contextual factors together with the meta-discourse served as important background for analyzing the interpreters’ linguistic behaviors later and finally led to Diriker’s (ibid., 145) conclusion about the “complex and dialectical relationship between interpreters, speakers and socio-cultural contexts”. I sympathize with Diriker’s (ibid.) idea that both the broad context (the meta-discourse) and the immediate context are important clues to understand interpreting as a socially situated practice. However, different from her aim to explore conference interpreting as a profession, my attempt is to investigate the more specific type of GPC interpreting in the Chinese context, therefore I will borrow her idea of the meta-discourse as the indicator of interpreting norms, but focus more specifically on the meta-discourse on diplomatic interpreting in China. GPC interpreting is the top-level diplomatic interpreting activity in China performed by in-house interpreters from the Department of Translation and Interpreting affiliated to the Ministry of Foreign Affairs, but its code of practice is not available to outsiders, so the published articles written by institutional insiders and my personal interview with a former Director of the Department on how to do diplomatic interpreting will be my major sources of meta-discourse to extract the institutional norms of GPC interpreting. Also different from Diriker’s (ibid.) focus on one particular interpreting event (lasting two days), my aim is to look for regular patterns in a larger corpus covering 21 interpreting events between 1989 and 2014, therefore her analysis of the immediate context of one interpreting event will be replaced with the analysis of the typological features of GPC interpreting shared by all the interpreting events in my corpus. In fact, the importance of typological analysis in contextualizing interpreting was acknowledged by Diriker (ibid., 12) herself: “Other lines of research, such as those on conference typologies and user expectation surveys, have also contributed to our understanding of the social context(s) in SI, even if their direct focus was not on exploring actual SI behavior.” Recognizing the complexity of interpreting activities, interpreting scholars such as Alexieva (1997) and Pöchhacker (2009) have been advocating the “multi-parameter” approach of typological analysis for interpreting events, which call attention to the various contextual/situational factors affecting the interpreting performance. Viewing interpreter-mediated events as instances of intercultural communication with different degrees of “culture-specificity”, Alexieva (1997)

3.1 Theoretical Framework

47

proposed to categorize such events in terms of various parameters under two broad headings: “mode of delivery” and “elements of the communicative situation”. “Mode of delivery” involves “the use, or non-use, of ancillary equipment” (such as the microphone), “the specificity of the setting”, “the nature of the contact” (direct or indirect), “distance between participants” as well as the distinction between simultaneous interpreting and consecutive interpreting. “Elements of the communicative situation” include “the primary participants (Speaker and Addressee), the secondary participants (Interpreter, Organizer, Moderator), the topic discussed and the way it relates to the communicative context, the type of texts used in the communication, the spatial and temporal specificities of the communication, and the purpose of communication or goals pursued by the participants” (ibid., 157). Although the present study does not share the aim to categorize the numerous types of interpreting events, all these parameters are useful in outlining the relevant situational factors shaping the performance of the Chinese-English GPC interpreting under investigation. This “multi-parameter” approach of typological analysis was echoed by Pöchhacker (2009), who included five parameters in discussing a “systematic inventory of types and subtypes of interpreting”: “language modality, working mode, directionality, technology and professional status” (ibid., 17). Among these five parameters, “working mode” corresponds to Alexieva’s (1997) “mode of delivery”. “Language modality” (concerning the difference between spoken language interpreting and sign language interpreting), “technology” (concerning the technical equipment used in SI) and “professional status” (concerning the distinction between professional interpreting and lay interpreting or natural interpreting) are not useful in contextualizing GPC interpreting. The most important parameter that complements Alexieva’s (ibid.) inventory is “directionality”, which distinguishes the “western tradition” of “B-to-A” interpreting from the practice of “A-to-B” interpreting in the Chinese GPC context (Pöchhacker 2009, p. 21). Therefore, drawing on the two inventories, I will include the following parameters in the typological analysis of GPC interpreting: participants, goal of communication, topics and text type, spatial and temporal setting, working mode and directionality.

3.1.4

The Working Model for Descriptive-Explanatory Analysis

To sum up, inspired by the three interpreting scholars (Diriker 2004; Alexieva 1997; Pöchhacker 2009), I will approach the sociocultural context of GPC interpreting from two perspectives: the institutional norms of GPC interpreting as indicated in the meta-discourse on diplomatic interpreting, and the multi-parameter typological features of GPC interpreting. So far I have discussed the three major sources of theoretical insights for the present study: the descriptive and comparative model for translation shifts from Toury (1995), the semantic system of modality

48

3 Theoretical Framework and Methodology

Fig. 3.2 The working model of descriptive-explanatory analysis

from Halliday and Matthiessen’s (2014), and the integrated models of sociocultural context analysis for interpreting from Diriker (2004), Alexieva (1997) and Pöchhacker (2009). All these works inspire me to propose the following working model (see Fig. 3.2) for descriptive-explanatory analysis on modality in interpreting, which will guide the present study. The description stage involves conducting corpus-based comparative textual analysis and exploring modality shifts (including the semantic shifts in modality type, modality value, and orientation) within the parallel sentence segments from two directions: the use of English modality markers and the translation of Chinese modality markers. The explanation stage involves exploring the relationship between the interpreting choices concerning modality and the sociocultural context, which includes the institutional norms of GPC interpreting and its typological features (participants, goals of communication, topics and text type, spatial and temporal setting, working mode and directionality).

3.2

Data and Method

To address the research questions discussed in Sect. 1.2, this project will integrate corpus-based quantitative analysis of modality shifts with qualitative analysis of the interpersonal effects of modality shifts and qualitative analysis of the sociocultural context of interpreting.

3.2 Data and Method

3.2.1

49

The Corpus

The interpreting data in this study are drawn from the latest version of the Chinese-English Conference Interpreting Corpus (CECIC 4.0) constructed by a CIS research team at Shanghai Jiao Tong University. As a team member of the CIS project, I have participated in all the construction procedures including data collection and updating, transcribing, text alignment, etc. CECIC 4.0 consists of three sub-corpora with a total size of 1,229,564 tokens (for corpus size, one single Chinese character or one English word is counted as one token): the parallel corpus of Chinese-English GPC Interpreting (Sub-corpus A), the parallel corpus of Chinese-English Government Work Report translation (Sub-corpus B) and the mono-lingual corpus of Original English (Sub-corpus C). Due to the aim of investigating modality shifts in GPC interpreting, only Sub-corpus A will be used in the present study. The raw data for Sub-corpus A are collected by our CTS research team from the live broadcast of major press conferences (with the Premier or Foreign Minister as the primary speaker) during the annual “Two Sessions” (i.e., the plenary sessions of NPC and CPPCC; see Sect. 1.1.2 for more details) from 1989 to 2014, which lead to 30 transcribed texts in total. For practical reasons, GPCs during the 25 years are not evenly represented, with more texts from recent years included than earlier years. For a more balanced representation of data, I draw 21 parallel texts from Sub-corpus A as the interpreting corpus for the present study, including only one text for each year. Note that the interpreting data for the GPCs in the year 1992, 1994, 1995, 1996 and 1997 are not included here, because the cassette tapes used to record these events were unfortunately lost or broken before their conversion to mp3 files. An overview of the final GPC interpreting corpus used for this study is presented in Table 3.4. As Table 3.4 shows, the raw data of the corpus are 21 audio files recorded during the live broadcast of the GPC events, with a total duration of 2217 min. In transcribing the recordings into written texts, the research team did not rely on the available transcriptions on line which are often approximate representations of the speeches, but resorted to our own ears to produce a verbatim orthographic transcription, taking down repetitions, hesitations, revisions as well as fluent speeches to ensure that the textual corpus consists of faithful representations of the original speeches. Annotations were offered for paralinguistic features such as pauses, interruptions, repetitions and revisions, although they will not be used in the present study. After transcription and annotation, the Chinese STs and English TTs were aligned at the sentence level, therefore the parallel corpus is ready to be operated on concordance software such as ParaConc. Note that the original recordings include consecutive interpretation for both directions (Chinese to English and English to Chinese) by professional interpreters from the Ministry of Foreign Affairs, but English questions in the ST are rather limited compared with answers and questions in Chinese. For the research purpose of the CIS project, all the English-to-Chinese interpretations were excluded before alignment during the corpus compilation. The

50

3 Theoretical Framework and Methodology

Table 3.4 An overview of the GPC interpreting corpus No.

Year

Primary speaker

ST (tokens)

TT (tokens)

ST + TT (tokens)

Interpreter

Duration (minutes)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Total

1989 1990 1991 1993 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

PM PM PM FM PM PM PM PM PM FM PM FM FM PM FM PM FM FM PM PM PM

9526 7019 9364 3657 6402 6408 6524 9924 6155 4590 7199 6907 9209 6564 10,140 8181 9096 9744 9311 8200 8878 162,998

7675 5358 7876 2784 5296 5170 6984 8269 4789 2852 6056 5403 7326 6624 6290 7360 6570 6703 7707 6362 6832 130,286

17,201 12,377 17,240 6441 11,698 11,578 13,508 18,193 10,944 7442 13,255 12,310 16,535 13,188 16,430 15,541 15,666 16,447 17,018 14,562 15,710 293,284

I1 and I2 I3 I4 and I3 I5 I6 I6 I7 I8 I8 I9 I10 I11 I12 I13 I14 I13 I15 I15 I15 I16 I15 16

125 87 123 49 83 85 99 123 74 60 110 92 104 115 115 144 114 116 182 107 110 2217

Li Peng Li Peng Li Peng Qian Qichen Zhu Rongji Zhu Rongji Zhu Rongji Zhu Rongji Zhu Rongji Tang Jiaxuan Wen Jiabao Li Zhaoxing Li Zhaoxing Wen Jiabao Yang Jiechi Wen Jiabao Yang Jiechi Yang Jiechi Wen Jiabao Li Keqiang Li Keqiang

final corpus used here is a Chinese-English parallel corpus, covering 21 interpreting events between 1989 and 2014 and involving 16 interpreters. The total size is 293, 284 tokens, including 162,998 Chinese characters and 130,286 English words. As these press conferences during the “Two Sessions” are the most influential among all Chinese GPCs with the interpretation offered by the in-house interpreters from the Ministry of Foreign Affairs who “represent the top-level of the translating/ interpreting profession in China” (Ren 2004, p. 61), the corpus is considered representative of diplomatic-political interpreting in China.

3.2.2

Identifying High-Frequency Modality Markers

Because modality markers can appear in different grammatical forms and they are not so much a closed category as modal verbs, it will be impractical to conduct a corpus-based study of all modality markers in the corpus. Instead of exhausting all

3.2 Data and Method

51

modality markers in the corpus, I choose to extract the most frequently used ones as a starting point. My identification of high-frequency modality markers is based on three criteria: (1) they appear in the top 350 wordlist (the English and Chinese wordlists; see Appendices A and B); (2) they express modal meanings as discussed in Sect. 3.1.2; (3) they appear mostly in positive forms; (4) the number of valid cases for analysis is larger than 30. With the help of Wordsmith 6.0, 15 English modality markers and 15 Chinese modality markers are identified (see Table 3.5), which appear in the wordlist head (top 350) of the English TTs and the Chinese STs respectively. Note that concordance reading is necessary for the identification of most of these forms to make sure they are used mostly to mark modality and used mostly in the positive form in the data. In cases that certain lexical items appear mostly in fixed structures to mark modality, the structures are identified as modality markers instead of the single words. For example, the top 350 wordlist for the English TTs reveals need, going, able, and want as potential markers of modality, but a careful reading of the concordance lines shows that they appear mostly in the modal structures need to + verb, be going to + verb and be able to + verb and want to + verb, thus I extract need to, be going to, be able to and want to as high-frequency semi-modals used in the English TTs. The term “semi-modal” is borrowed from Biber et al. (2000, p. 73) to refer to modal forms that are not listed in the nine core modals (can, could, may, might, must, shall, should, will, would) but “are on the borderline between auxiliaries and lexical verbs” or “are related in meaning to the modal auxiliaries” (ibid.). Similarly, the Chinese modality markers 一定, 必须, and 需要 (+verb) that cannot be classified into modal verbs without controversy are also considered as semi-modals. The forms think, believe, hope, 想, 认为, 相信, and 希 望 appear predominantly with a first-person subject followed by a clause when they convey modal meanings, so I confine my analysis of this group to the “1st-person subject +mental lexical verb +clause” structure which is categorized as “interpersonal metaphor” in SFG (Halliday and Matthiessen 2014, p. 687). According to their predominant grammatical properties, these high-frequency modality markers are grouped into central modals, semi-modals, “1st-person subject +mental lexical verb +clause” structures, and modal adverbs, as shown in Table 3.5.

Table 3.5 High-frequency modality markers in the corpus Language

English

Chinese

Modality marker Central modal

Will, can, would, should, must, may, could 要, 会, 可以, 能/ 能够, 应该, 愿意, 可能

Semi-modal

1st-person subject +mental lexical verb +clause

Modal adverb

Be able to, be going to, need to, want to 一定, 必须, 需要 (+verb)

I/we think, I/we believe, I/we hope

Always

我/我们想, 我/我们认 为, 我/我们相信, 我/ 我们希望



52

3 Theoretical Framework and Methodology

Fig. 3.3 A snapshot of the parallel concordance view of 应该 with ParaConc

These high-frequency modality markers are used as “search words” in the parallel corpus to establish “coupled pairs” between the Chinese STs and the English TTs, so that the ST-TT relationship concerning modality can be further explored. Figure 3.3 presents a snapshot of the parallel concordance view of 应该 with ParaConc., which can be imported into Microsoft Office Excel forms for semantic coding.

3.2.3

Coding Modality Shifts

As the present study aims to investigate translation shifts in the semantics of modality instead of grammatical forms, my coding scheme is primarily concerned with semantic shifts in modality type, modality value and orientation. In practice, I have to first determine whether the investigated modality marker has a corresponding form in its Chinese ST or English TT, and then compare the semantics of modality between the “coupled pair” for shifts. It is interesting to see whether a semantic shift is caused by the addition or omission of the investigated modality marker (i.e., macro-shift1), or because there is a semantic difference in the corresponding modal expressions in the ST and TT. Therefore, my coding scheme involves analyzing modality shifts at two levels: semantic shifts of modality at the general level and specific shifts of modality within the three semantic dimensions of modality, hence two-level variables.

For the convenience of discussion, “macro-shift” is used to refer to the “addition” or “omission” of the modality marker in the interpreting. However, those cases of macro-shifts that do not cause semantic shifts are not counted as modality shifts in this study.

1

3.2 Data and Method

53

The first-level variable “SEMANTIC SHIFT” (SST) is coded as SST-C (semantic shift that occurs when there is a corresponding expression in the ST or TT for the investigated modality marker), SST-ZC (semantic shift that occurs when there is no corresponding expression in the ST or TT for the investigated modality marker) or ZSST (zero semantic shift). Semantic shifts of modality are considered to occur when there is at least one identifiable change in any of the three dimensions of modality (modality type, modality value and orientation) between the ST-TT “coupled pair” involving a modality marker. When there is a corresponding expression in the ST/TT for the investigated modality marker, semantic shifts are identified by comparing the two corresponding forms within the linguistic context of the parallel sentences. When the investigated modality marker has no corresponding expression in the ST (but there is a correspondence between the parallel clauses), semantic shifts are analyzed by comparing the two corresponding sentences to see if the omission or addition of the modality marker on the clause level affects the semantics of modality of the whole sentence. Another possible value for the variable “SEMANTIC SHIFT” (SST) is “NA” (not applicable), which refers to cases when it is hard to compare the ST-TT “coupled pair” for any of the three semantic dimensions of modality (i.e., when shifts of all the three semantic dimensions are coded as “NA”) because the whole clause containing the modality marker is omitted or added in the translation. Theoretically speaking, semantic shifts of modality involve both obligatory shifts (shifts that are unavoidable because of linguistic differences between the two languages) and optional shifts (shifts that can be avoided despite the linguistic differences) (for the two terms, see Bakker et al. 2004, p. 228), but obligatory shifts are found to be rather rare in my data, so I choose to code all the obligatory shifts as ZSST,2 which ensures that all the semantic shifts identified are optional shifts motivated by sociocultural factors. To obtain detailed information of semantic shifts in the three dimensions of modality, my coding scheme includes three second-level variables: SHIFT of MODALITY TYPE (SMT), SHIFT of ORIENTATION (SO), and SHIFT of MODALITY VALUE (SV). As discussed in Sect. 3.1.2, Chinese and English modality markers can express four types of modality: probability (P), usuality (U), obligation (O), and inclination (I). So shifts of modality type occur when the corresponding modality markers in the ST and TT do not express the same type of modality. Depending on the specific cases, the variable SMT is coded as “O-P” (shift from obligation to probability), “O-I” (shift from obligation to inclination), etc. For example, the shift of “O-I” occurs in the translation of “应该鼓励更多的人 租房” into “We will continue to encourage people to rent homes” because the obligation marker 应该 (should) is rendered into an inclination marker will. In addition, the variable SMT is coded as “E” (equivalence) for cases where the corresponding modality markers in the ST and TT do express the same type of modality, and coded as “NA” (not applicable) when there is no corresponding form

Accordingly, all the three variables SMT, SO and SV are coded as “E” in the cases of obligatory shifts.

2

54

3 Theoretical Framework and Methodology

in the ST/TT for the investigated modality marker or when it is hard to compare the modality types of the ST-TT “coupled pair”. The second specific variable SHIFT of ORIENTATION (SO) is coded based on our discussion of “subjective vs. objective” orientation of modal expressions in Sect. 3.1.2. Because modality markers and their corresponding forms in the ST/TT may express different orientations of modality ranging from “explicit subjective” to “polar objective” (see Table 3.2), two kinds of shifts may occur in terms of orientation: “objectivization” and “subjectivization”. “Objectivization” occurs when a relatively subjective modality marker is translated into a more objective one, or when the modality marker in the ST is dropped in the TT without any compensation of modal expressions in the co-text. “Subjectivization” occurs when a relatively objective modality marker is rendered into a more subjective one, or when a polar statement without modality markers is rendered into a modalized statement. Shifts of orientation are also considered to occur when the use of “I/we+think/believe/ hope” or the translation of “我/我们+想/认为/相信/希望” involves a shift in the subject between the 1st-person singular and the 1st-person plural, because the phrase with a singular subject clearly marks the speaker’s personal opinion, while it marks the collective opinions of a group (usu. the government) with a plural subject. For example, “objectivization” occurs when 我相信 (I believe) is rendered into we believe, while “subjectivization” occurs when 我们相信 (we believe) is rendered into I believe. Accordingly, these two types of shifts are coded as “SOO” (objectivization) and “SOS” (subjectivization) under the variable SO. In addition, the variable SO will be coded as “E” (equivalence) when there is no difference between the ST-TT “coupled pair” in terms of orientation, or “NA” (not applicable), when the whole clause containing the modality marker is omitted or added in the translation. In line with the third semantic dimension of modality discussed in Sect. 3.1.2, the third specific variable SHIFT of MODALITY VALUE (SV) concerns the possible difference in modality value between the investigated modality marker and its corresponding expressions in the ST/TT. According to the proposed scale of modality value summarized in Table 3.3, shifts of modality value include two major types: “strengthening” and “weakening”. “Strengthening” occurs when the a lower-value modality marker is rendered into a higher-value one or when an existing Chinese modality marker is dropped in the English interpretation, and “weakening” occurs when a higher-value modality marker is rendered a lower-value one or when a polar statement is rendered into a modalized one. These two specific types of shifts in modality value are coded as “SV+” (strengthening) or “SV−” (weakening) respectively under the variable SV. In addition, like in the case of SO, the variable SV is coded as “E” (equivalence) when there is no difference between the ST-TT “coupled pair” in terms of modality value, or “NA” (not applicable), when the whole clause containing the modality marker is omitted or added in the translation. The typical examples for the different modality types, modality values and orientations of English and Chinese modality markers summarized in Tables 3.1, 3.2 and 3.3 (see Sect. 3.1.2) are used as benchmarks for shift analysis,

3.2 Data and Method

55

although actual coding often has to be established on an ad hoc basis depending on the linguistic context. To ensure the inter-coder reliability, a colleague (with similar bilingual background and research interest) and I have coded a random sample (500 translation cases from each direction) of the corpus data independently, and then discussed the discrepancies to reach a consensus before the actual coding of the whole interpreting data. In addition to the information about modality shifts, the ST speaker and the time of event as contextual clues are also coded for each parallel sentence segments extracted from the corpus. The time variable is coded as the specific year when the GPC event occurred. The ST speaker is coded as “PMT” (Premier or Foreign Minister), “JNL” (journalist), “PTQ” (Premier or Foreign Minister quoting others), “JLQ” (journalist quoting others) or “CP” (Chairperson). The corresponding form (CF) of the investigated modality marker is coded as “Z” (zero correspondence) or the specific form for the convenience of further analysis. The coding scheme is summarized in Table 3.6.

3.2.4

Describing Modality Shifts in the Corpus

As mentioned in Sect. 3.1.1, I will analyze modality shifts in the interpreting corpus from two directions (details will be presented in Chaps. 4 and 5). First, I will start with the high-frequency English modality markers in the TT and see how they creep into the English TT from the Chinese ST. With the help of ParaConc, all the parallel concordance lines containing the high-frequency English modality markers in the corpus will be extracted into an excel form for semantic coding (for samples of the excel forms and coding results, see Appendix C). Based on the coding of modality shifts and the ST speaker (government official or journalist), I will explore whether the identity of ST speaker affects the degree and major types of modality shifts in the use of English modality markers (Research Question 1b), as well as the overall patterns of modality shifts revealed by the use of English modality markers (Research Question 1a). Statistical analysis will be further conducted with the help of the “pivot table” function of Microsoft Excel Worksheet and the calculator “LLX2” developed by Beijing Foreign Studies University. Following a description of overall patterns of modality shifts in the use of English modality markers, the major types of modality shifts will be further discussed with examples of specific English modality markers to reveal the interpersonal effects of these shifts (Research Question 1c). Then I will proceed from the Chinese ST and trace the high-frequency Chinese modality markers into English. Parallel concordance lines containing the Chinese modality markers will be extracted to see how Chinese modality is rendered into the English TT. By repeating the same coding procedure and statistical analysis, I will be able to explore Research Question 1 from a different direction. Following a description of overall patterns of modality shifts in the translation of Chinese modality markers (Research Question 1a) and a comparative analysis of modality

56

3 Theoretical Framework and Methodology

Table 3.6 The coding scheme of modality shifts Variable

Code

Description

SEMANTIC SHIFT (SST)

SST-C

Semantic shift that occurs when there is a corresponding expression in the ST or TT for the investigated modality marker Semantic shift that occurs when there is no corresponding expression in the ST or TT for the investigated modality marker Zero semantic shift Not applicable Shift from obligation to inclination Shift from obligation to probability Shift from inclination to obligation Equivalence (no shift of modality type) Not applicable Objectivization Subjectivization Equivalence (no shift of orientation) Not applicable Strengthening Weakening Equivalence (no shift of modality value) Not applicable Premier or Foreign Minister Journalist Premier or Foreign Minister quoting others Journalist quoting others Chairperson Zero correspondence The corresponding form of the investigated modality marker in the translation case is “should” …

SST-ZC

SHIFT of MODALITY TYPE (SMT)a

SHIFT of ORIENTATION (SO)

SHIFT of MODALITY VALUE (SV)

ST speaker

Corresponding form (CF)

ZSST NA O-I O-P I-O E NA SOO SOS E NA SV+ SV− E NA PMT JNL PTQ JLQ CP Z Should

… 1989 1990 … a To save space, only the three most frequently occurring shifts of modality type are described and illustrated here Year

shifts with different ST speakers (Research Question 1b), the major types of modality shifts will be further discussed with examples of specific Chinese modality markers to reveal the interpersonal effects of these shifts (Research Question 1c). Data from the two directions will then be compared so as to summarize the findings on modality shifts in GPC interpreting.

3.2 Data and Method

3.2.5

57

Exploring the Relationship Between Modality Reconstruction and Its Sociocultural Context

For data explanation, I will explore the relationship between the regular patterns of modality shifts revealed in the descriptive analysis and the sociocultural context of GPC interpreting (Research Question 2). Two factors of the sociocultural context will be analyzed: the institutional norms (Research Question 2a), and the typological features of GPC interpreting (Research Question 2b). The institutional norms will be extracted from the meta-discourse on GPC interpreting, i.e., published articles from institutional insiders and a personal interview with the former director of the Department of Translation and Interpreting, Ministry of Foreign Affairs. The typological features will be discussed according to the following parameters: participants, goal of communication, topics and text type, spatial and temporal setting, working mode and directionality.

References Alexieva, Bistra. 1997. A typology of interpreter-mediated events. The Translator 3 (2): 153–174. Austin, John. 1962. How to do things with words. Oxford: Oxford University Press. Baker, Mona. 2004. Norms. In Routledge Encyclopedia of Translation Studies, ed. Mona Baker, 163. Shanghai: Shanghai Foreign Language Education Press. Baker, Mona. 2006. Contextualization in translator- and interpreter-mediated events. Journal of Pragmatics 38 (3): 321–337. Bakker, Matthijs, Cees Koster, and Kitty van Leuven-zwart. 2004. Shifts of translation. In Routledge Encyclopedia of Translation Studies, ed. Mona Baker, 226–231. Shanghai: Shanghai Foreign Language Education Press. Bartsch, Renate. 1987. Norms of Language. London: Longman. Bell, Roger. 1991. Translation and Translating: Theory and Practice. London: Longman. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 2000. Longman Grammar of Spoken and Written English. Beijing: Foreign Language Teaching and Research Press. Chesterman, Andrew. 1999. Description, explanation, prediction: A response to Gideon Toury and Theo Hermans. In Translation and Norms, ed. Christina Schäffner. Clevedon: Multilingual Matters Limited. de Haan, Ferdinand. 1997. The Interaction of Modality and Negation: A Typological Study. New York: Garland. Diriker, Ebru. 2004. De-/Re-Contextualizing Conference Interpreting. Amsterdam/Philadelphia: John Benjamins. Halliday, Michael A.K. 1978. Language as Social Semiotic: the Social Interpretation of Language and Meaning. London: Edward Arnold. Halliday, Michael A.K. 1985. An Introduction to Functional Grammar, 1st ed. London: Edward Arnold. Halliday, Michael A. K. 1992. Language theory and translation practice. Rivista Internazionale di Technica Della Traduzione (0):15–25. Halliday, Michael A.K. 1994. An Introduction to Functional Grammar, 2nd ed. London: Edward Arnold.

58

3 Theoretical Framework and Methodology

Halliday, Michael A.K., and Christian Matthiessen. 2004. An Introduction to Functional Grammar, 3rd ed. London: Edward Arnold. Halliday, Michael A.K., and Christian Matthiessen. 2014. Halliday’s Introduction to Functional Grammar, 4th ed. Abington/New York: Routledge. Halliday, Michael A.K., and Edward McDonald. 2004. Metafunctional profile of the grammar of Chinese. In Language Typology: A Functional Perspective, ed. Alice Caffarel, James R. Martin, and Christian Matthiessen, 305–396. Amsterdam: Benjamins. House, Juliane. 2006. Text and context in translation. Journal of Pragmatics 38 (3): 338. Hu, Zhuanglin 胡壮麟. 1999. 系统功能语法与汉语研究 [Systemic functional grammar in the study of Chinese]. In 语法研究入门 [An Introduction to the Study of Chinese Grammar], ed. Qingzhu Ma 马庆株, 261–262. Beijing: The Commercial Press. Hu, Kaibao 胡开宝. 2012. 语料库翻译学: 内涵与意义 [Corpus-based Translation Studies: connotations and implications]. 外国语 [Journal of Foreign Languages] (5): 59–70. Hymes, Dell. 1964. Toward ethnographies of communicative events. In Language and Social Context, ed. Pier Paolo Giglioli, 21–44. Harmondsworth: Penguin. Laviosa, Sara. 2004. Corpus-based translation studies: where does it come from? Where is it going? Language Matters 35(1): 6–27. Liu, Yanqin 刘燕芹. 2009. 功能视角下英汉情态表达语对比研究 [A comparative study of modality in English and Chinese from a functional perspective]. 西安外国语大学学报 [Journal of Xi’an International Studies University] (2): 25–28, 33. Malinowski, Bronislaw. 1923. The problem of meaning in primitive languages. In The Meaning of Meaning, ed. Charles K. Ogden, and Ivor A. Richards, 296–336. New York: Harcourt Brace. Martin, James R. 1992. English Text: System and Structure. Amsterdam: Benjamins. Martin, James R. 1999. Modelling context. In Text and Context in Functional Linguistics, ed. Mohsen Ghadessy, 25–61. Amsterdam: John Benjamins. Monacelli, Claudia. 2005. Surviving the role: a corpus-based study of self-regulation in simultaneous interpreting as perceived through participation framework and interactional politeness. Ph.D. thesis, Heriot-Watt University. Pöchhacker, Franz. 2009. Introducing Interpreting Studies. Shanghai: Shanghai Foreign Language Education Press. Palmer, Frank. 1990. Modality and the English Modals. London: Longman Group Limit. Palmer, Frank. 1995. Negation and the modals of possibility and necessity. In Modality in Grammar and Discourse, ed. Joan Bybee, and Suzanne Fleischman, 453–473. Amsterdam/ Philadelphia: John Benjamins. Palmer, Frank. 2003. Modality in English: Theoretical, descriptive and typological issues. In Modality in Contemporary English, ed. Roberta Facchinetti, Manfred Krug, and Frank Palmer, 1–20. New York: Mouton de Gruyter. Peng 彭利贞, Lizhen. 2007. 现代汉语情态研究 [Modality of Modern Mandarin]. Beijing: China Social Sciences Press. Peng 彭宣维, Xuanwei. 2000. 英汉语篇综合对比 [A Comprehensive Comparison Between English and Chinese Texts]. Shanghai: Shanghai Foreign Language Education Press. Ren, Xiaoping 任小萍. 2004. 外交部高级翻译培训 [The training of senior translators/ interpreters at the Ministry of Foreign Affairs]. 中国翻译 [Chinese Translators Journal] (1): 61–62. Thompson, Geoff. 2008. Introducing Functional Grammar. Beijing: Foreign Language Teaching and Research Press. Toury, Gideon. 1978. The nature and role of norms in literary translation. In Literature and Translation, eds. James S Holmes, José Lambert, and Raymond van den Broeck, 83–100. Leuven: ACCO. Toury, Gideon. 1980. In Search of a Theory of Translation. Tel Aviv: Porter Institute. Toury, Gideon. 1995. Descriptive Translation Studies and Beyond. Amsterdam/Philadelphia: John Benjamins Publishing Company. Toury, Gideon. 1999. A handful of paragraphs on ‘translation’ and ‘norms’. In Translation and Norms, ed. Christina Schäffner, 9–31. Clevedon: Multilingual Matters Limited.

References

59

Usoniene, Aurelia, and Audrone Šoliene. 2010. Choice of strategies in realizations of epistemic possibility in English and Lithuanian: A corpus-based study. International Journal of Corpus Linguistics 15 (2): 291–316. van der Auwera, Johan. 2001. On the typology of negative modals. In Perspectives on Negation and Polarity Items, ed. Jack Hoeksema, Hotze Rullmann, Víctor Sánchez-Valencia, and Ton van der Woulden, 21–48. Amsterdam/Philadelphia: John Benjamins. van Dijk, Teun A. 1998. Ideology: A Multidisciplinary Approach. London: Sage. van Dijk, Teun A. 2001. Discourse, ideology and context. Folia Linguistica 35 (1): 11–40. Wei, Zaijiang 魏在江. 2008. 基于功能的英汉语情态隐喻对比研究 [Metaphors of modality in English and Chinese: A contrastive study]. 现代外语 [Modern Foreign Languages] (3): 263– 272, 329. Xu, Jun 徐珺. 2003. 功能语法用于《儒林外史》汉英语篇的研究:情景语境观 [A functional study of Ru Lin Wai Shi and its English translation in a context-of-situation perspective]. 现代 外语 [Modern Foreign Languages] (2): 128–134, 160. Zhang, Meifang 张美芳. 2001. 从语篇分析的角度看翻译中的对等 [Aspects of translation from a discoursal perspective]. 现代外语 [Modern Foreign Languages] (1): 78–84. Zhu, Yongsheng. 1996. Modality and modulation in Chinese. In Meaning and Form: Systemic Functional Interpretations, ed. Margaret Berry, Christopher Butler, Robin Fawcett, and Guowen Huang, 183–210. Norwood: Ablex Publishing Corporation.

Chapter 4

Modality Shifts in the Use of English Modality Markers

4.1

General Features of Modality Shifts

As discussed in Sect. 3.2.2, 15 English modality markers are identified as high-frequency modality markers to be investigated in this study. While analyzing the parallel concordance lines with these 15 forms generated with ParaConc, I have deleted negative patterns (e.g., will not, will never, may not), noisy homonyms (e.g., may as in “last May”, must as in “one must, three completions and five reforms”), and modal forms that fall out of the four types of modality in Fig. 3.1 (e.g., be able to in its pure-ability sense as in “if the people are able to run the affairs of a village well”, should in its hypothetical conditional sense as in “Should one view the actions taken by a country to defend its core interests and dignity as being tough, …”). When two modality markers are used together to form a structure which cannot be analyzed separately in terms of translation, this structure is analyzed under one modality marker only. For example, cases involving “I/we hope…will” as a fixed structure is analyzed under “I/we hope” and deleted under “will”. Finally a total number of 3013 parallel concordance lines are generated for the present study, with the distribution of valid modality markers presented in Table 4.1. As the table shows, central modals are the primary forms of modality markers in the English TTs, but the other forms also occur with a considerably high frequency so it is necessary to take them all into account in exploring the reconstruction of modality. In this section, I will present my findings about the general features of modality shifts, first addressing the overall patterns revealed by the 3013 parallel cases, and then discussing the degree/types of modality shifts occurring with different ST speakers. More detailed discussion of the interpersonal effects of modality shifts will be offered in Sect. 4.2 with examples from the corpus.

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_4

61

62

4 Modality Shifts in the Use of English Modality Markers

Table 4.1 High-frequency EMMs extracted from the corpus EMMs in the TT Central modal (2112)

Semi-modal (427)

1st-person subject + mental lexical verb + clause (412)

Modal adverb (62) Total

4.1.1

Occurrence Will Would Can Should Must May Could Need to Be going to Want to Be able to I/we think I/we believe I/we hope Always

1021 325 316 249 99 62 40 165 110 103 49 169 154 89 62 3013

Overall Degree and Distribution of Modality Shifts

Following the coding scheme presented in Table 3.6, all the 3013 parallel concordance lines involving the use of high-frequency English modality markers were coded manually for semantic shifts of modality, with both the first-level variable SEMANTIC SHIFT (SST) and the second-level variables—SHIFT of MODALITY TYPE (SMT), SHIFT of ORIENTATION (SO), and SHIFT of MODALITY VALUE (SV)—attended to. The inter-coder agreement on the coding of the sample data (500 concordance lines) is 86.6%, which means the coding process is sufficiently reliable for this study. For convenient reference to the coding scheme discussed in Sect. 3.2.3, the original codes involved for various categories of modality shifts will be indicated in parentheses in the tables of this chapter. First, let’s look at the overall degree of modality shifts revealed by the use of English modality markers. As Table 4.2 shows, of all the 3013 cases of English modality markers used in the GPC interpreting corpus, 41.02% involve semantic shifts of modality and 57.75% witness no semantic shifts of modality. The LL (log-likelihood) test (LL = 85.768; p < 0.01) confirms the pattern that compared with the option of modality shifts (SST-C/ZC), equivalent translation (ZSST) is the preferred choice in the use of English modality markers. However, due to the availability of Chinese equivalent expressions for these English modality markers, the degree of modality shifts (41.02%) is still fairly high, and it is important to explore how and why semantic shifts of modality occur in these cases. Although this study focuses on

4.1 General Features of Modality Shifts Table 4.2 Semantic shifts of modality revealed by the use of EMMs

63 F.a

Semantic shifts (SST-C/ZC) Zero semantic shifts (ZSST) Not applicable (NA) Total a “F.” is used to refer to “frequency”

%

1236 41.02 1740 57.75 37 1.23 3013 100.00 in the tables of this book

modality shifts at the semantic level, it is also interesting to see how many of these semantic shifts are caused by macro-shifts. Breaking semantic shifts of modality into shifts occurring with corresponding forms in the ST of the investigated English modality markers (SST-C) and shifts occurring without corresponding forms in the ST of the investigated English modality markers (SST-ZC) (i.e., shifts caused by macro-shifts of addition), we can see that over half of the semantic shifts of modality occur (52.59%) with macro-shifts of addition (see Table 4.3). By contrast, the occurrence of semantic shifts (47.41%) with corresponding forms in the ST of the investigated English modality markers is lower, although no significant difference exists between the two groups (LL = 3.315, p = 0.069). It is safe to say that the frequent addition of English modality markers is an important contributor of modality shifts in the use of English modality markers. Next, let’s look at the distribution of modality shifts in the three semantic dimensions (modality type, orientation and modality value). As Table 4.4 reveals, 89.72% of modality shifts in the use of English modality markers involve shifts of modality value, 57.36% involve shifts of orientation and only 7.04% involve shifts of modality type. Thus, shifts of modality value are the most common, shifts of orientation are fairly common and shifts of modality type are the least common. Note that the sum of the frequencies of the three types of shifts exceeds the total number of cases involving modality shifts, because it is possible for the three types of shifts to co-occur in the same parallel case (see Appendix C for examples). Table 4.3 Modality shifts occurring with or without corresponding forms of the EMMs in the ST

Table 4.4 Modality shifts in three semantic dimensions in the use of EMMs

F. Modality shifts with corresponding forms in the ST (SST-C) Modality shifts without corresponding forms in the ST (SST-ZC) Total

Shifts of modality type (O-I/O-P/…) Shifts of orientation (SOO/SOS) Shifts of modality value (SV+/SV−) Total cases involving modality shifts

%

586

47.41

650

52.59

1236

100.00

F.

%

87 709 1109 1236

7.04 57.36 89.72

64

4 Modality Shifts in the Use of English Modality Markers

Within each semantic dimension of modality, there are different types of shifts (e.g., shifts from probability to obligation, “strengthening”, “subjectivization”, etc.) and their relative proportion also deserves our attention. The distribution of specific shifts within the three dimensions is presented in Tables 4.5, 4.6 and 4.8. As Table 4.5 shows, among the limited cases of shifts in modality type, shifts from obligation to inclination take up the largest proportion (47.13%), followed by shifts from inclination to obligation (18.39%) and shifts from obligation to probability (16.09%). It seems that interpreters have a relatively strong tendency to replace markers of obligation with markers of inclination or probability (63.22%) among all the optional changes between modality types (i.e., the tendency of “de-obligation”). A significant difference (LL = 6.153, p < 0.05) is found to exist between the frequency of “de-obligation” (55 cases) and the frequency of other shifts in modality type (32 cases). As all shifts of modality type occur with corresponding forms in the ST of the investigated English modality markers, it is not necessary to explore the contribution of macro-shifts here. As for shifts of orientation, a clear preference for “subjectivization” instead of “objectivization” is indicated in Table 4.6: compared with only 28 cases (3.95%) of “objectivization”, there are 681 cases (96.05%) of “subjectivization”. The significant difference between the two is confirmed through the LL test (LL = 747.031, p < 0.01). Different from shifts of modality type, shifts of orientation often co-occur with macro-shifts, so it is necessary to explore to what extent the specific shifts of “subjectivization” and “objectivization” are caused by the addition of English modality markers. As revealed by Table 4.7, most of the shifts of “subjectiviation” (622 cases, 91.34%) result from the addition of English modality markers. By contrast, all shifts of “objectivization” occur with corresponding forms in the ST of the English modality markers. But because the total cases of Table 4.5 Shifts of modality type in the use of EMMs

Shifts of modality type

F.

%

Inclination to obligation (I-O) Inclination to probability (I-P) Obligation to inclination (O-I) Obligation to probability (O-P) Probability to inclination (P-I) Probability to obligation (P-O) Probability to usuality (P-U) Usuality to obligation (U-O) Total

16 7 41 14 5 2 1 1 87

18.39 8.05 47.13 16.09 5.75 2.30 1.15 1.15 100.00

Table 4.6 Shifts of orientation in the use of EMMs

Shifts of orientation

F.

%

Objectivization (SOO) Subjectivization (SOS) Total

28 681 709

3.95 96.05 100.00

4.1 General Features of Modality Shifts

65

Table 4.7 Shifts of orientation occurring with or without corresponding forms in the ST of the EMMs Subjectivization (SOS) F. % With corresponding forms in the ST (SST-C) Without corresponding forms in the ST (SST-ZC) Total

Objectivization (SOO) F. %

59

8.66

28

100.00

622

91.34

0

0.00

681

100.00

28

100.00

“objectivization” are rather limited, there is still a significantly stronger pattern of “subjectivization” than “objectivization” (F1 = 59, F2 = 28; LL = 11.292, p < 0.01) considering only the cases with corresponding forms in the ST of the English modality markers. All these data suggest that interpreters at Chinese GPCs tend to foreground the speaker’s subjective stance towards the message communicated, to the effect that the Chinese officials’ positive image of “taking responsibility for what they express” is presented. Of course, this effect might be offset by an opposite pattern of “objectivization” caused by omitting the Chinese modality markers in interpreting, which will be discussed in Chap. 5. Table 4.8 suggests a much stronger tendency towards “weakening” (76.10%) than “strengthening” (23.90%) within shifts of modality value and the significant difference is confirmed by the LL test (LL = 317.786, p < 0.01). Due to the prevalence of shifts in modality value (1109 cases), we can assume this predominant distribution of “weakening” shifts to have an important “weakening” effect on the interpreting product. In other words, the modal commitment of the ST speakers tends to be weakened by interpreters so that the English TT will sound more consultative and less assertive than the Chinese TT. Examining the relationship between these two types of shifts in modality value and macro-shifts (Table 4.9), we can see that the overall predominance of “weakening” is caused by the frequent addition of English modality markers, as most of the “weakening” shifts (623 cases, 73.82%) occur without corresponding forms in the ST of the English modality markers. If only the cases with corresponding forms in the ST of the English modality markers are considered, the frequency of “strengthening” appear to be significantly higher than “weakening” (F1 = 265, F2 = 221; LL = 3.989, p < 0.05). Therefore, we can conclude that interpreters tend to employ the addition of modality markers in the TT rather than replace a strong modality marker in the ST with a weak one in the TT to achieve the overall “weakening” effect of modality.

Table 4.8 Shifts of modality value in the use of EMMs

Shifts of modality value

F.

%

Weakening (SV−) Strengthening (SV+) Total

844 265 1109

76.10 23.90 100.0

66

4 Modality Shifts in the Use of English Modality Markers

Table 4.9 Shifts of modality value occurring with or without corresponding forms in the ST of the EMMs

With corresponding forms in the ST (SST-C) Without corresponding forms in the ST (SST-ZC) Total

4.1.2

Strengthening (SV+) F. %

Weakening (SV−) F. %

265 0 265

221 623 844

100.00 0.00 100.00

26.18 73.82 100.00

Modality Shifts with Different ST Speakers

So far my analysis has responded to the first part of Research Question 1 (1a) by overviewing the regular patterns in modality shifts revealed by the use of English modality markers in Chinese-English GPC interpreting without distinguishing the identities of the ST speakers. As the overwhelming majority of the ST segments are uttered by Chinese officials, we can neglect the possibility of journalists as ST speakers for general discussions of modality shifts. However, it is necessary in our next step of analysis to make this distinction to address the second part of Research Question 1 (1b): Does the identity of ST speaker affect the degree or major types of modality shifts in the use of English modality markers in Chinese–English GPC interpreting? While coding the speaker’s identity, I have made distinctions between premiers/ministers (PMT), journalists (JNL), chairpersons (CP), journalists quoting others (JLQ) and premiers/ministers quoting others (PTQ). Because the last three types of ST speakers have very limited occurrence in our interpreting data (CP: 14; JLQ: 28; PTQ: 13), I will only compare the degree/types of shifts occurring in the translation of speech segments by the major two groups of speakers: premiers/ministers (PMT) and journalists (JNL). Previous studies on GPC interpreting (Wang 2009, 2012; Liu 2010) suggest that interpreters tend to discriminate journalists from premiers/ministers in making interpreting choices because of their alignment with the latter. Will this difference be reflected in the degree or major types of modality shifts while translating speeches of premiers/ministers and journalists? As Table 4.10 shows, there is no significant difference between ST speeches by premiers/ministers and journalists in terms of the relative frequency of modality Table 4.10 Modality shifts in the use of EMMs for translating speeches by different speakers

Modality shifts (SST-C/ZC) Occurrence of EMMs

Premier/Foreign minister (PMT)

Journalist (JNL)

LL

Sig.

1070

145

0.083

0.773

2597

361

4.1 General Features of Modality Shifts

67

Table 4.11 Specific types of modality shifts in the use of EMMs for translating speeches by different speakers Premier/Foreign minister (PMT) Shifts of modality type in total Shifts of modality De-obligation type in detail (O-I/P) Other Shifts of modality value in total Shifts of modality Weakening value in detail (SV−) Strengthening (SV+) Shifts of orientation in total Shifts of orientation Subjectivization in detail (SOS) Objectivization (SOO) Occurrence of EMMs

Journalist (JNL)

LL

Sig.

76 47

10 7

0.027 −0.029

0.869 0.866

29 965 717

3 125 111

0.257 0.561 −1.086

0.612 0.454 0.297

248

14

14.194

0.000

598 574

96 92

−1.656 −1.553

0.198 0.213

24

4

−0.108

0.742

2597

361

shifts in general (LL = 0.083, p = 0.773). As for more specific types of modality shifts, Table 4.11 reveals the similar preference for “de-obligation”, “weakening” and “subjectivization” in translating speeches by premiers/ministers and journalists, but the shift of “strengthening” occurs with a significantly higher probability in translating speeches of premiers/ministers than journalists (LL = 14.194, p < 0.01). Thus I conclude that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the use of English modality markers in Chinese-English GPC interpreting but does affect the degree of one specific type of semantic shift, i.e., the shift of “strengthening”. As “strengthening” involves the use of a higher-value modality marker in the English TT to translate a lower-value modality marker in the Chinese ST or the dropping of an existing modality marker in the ST, it probably has the primary effect of emphasizing the ST speaker’s attitude towards some crucial political issues, so it appears with a higher probability in translating Chinese officials’ answers than journalist’s questions. To sum up, the analysis of concordance lines with English modality markers in the TT reveals the following overall patterns of modality shifts in Chinese-English GPC interpreting (Research Question 1a): (1) Comparatively speaking, semantic shifts of modality occur less frequently than zero shifts, which reveals the preference for “equivalent translation” in the reconstruction of modality in ChineseEnglish GPC interpreting; (2) As equivalent modal expressions for these English modality markers are available in Chinese, the degree of modality shifts (1236 out of 3013) is still pretty high, to which the frequent addition of English modality markers is an important contributor; (3) Among the three dimensions of semantic

68

4 Modality Shifts in the Use of English Modality Markers

shifts, shifts of modality value are the most common, shifts of orientation are fairly common and shifts of modality type are the least common. (4) The distribution within the three types of semantic shifts demonstrates clear preferences for “de-obligation”, “subjectivization” and “weakening” on the part of the interpreters, but the predominance of “weakening” is caused by the frequent addition of English modality markers in the translation. As for Research Question 1b, the statistical analysis shows that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the use of English modality markers in Chinese-English GPC interpreting but does affect the degree of one specific type of semantic shift, i.e., the shift of “strengthening”, which features more prominently in translating speeches by premiers/ministers than journalists. In the next section, I will analyze the interpersonal effects of the major types of modality shifts revealed by the use of English modality markers with examples from the corpus (Research Question 1c).

4.2

Interpersonal Effects of the Major Types of Modality Shifts

Since the analysis of modality shifts in Sect. 4.1 based on the ST-TT comparison involving the use of 15 English modality markers has already revealed the overall degree and major types of modality shifts, now I will examine how the major types of modality shifts are distributed in the use of these 15 modality markers and then focus on the major contributors for the analysis of interpersonal effects. In addition to the three predominant types of modality shifts within the three semantic dimensions (“de-obligation”, “subjectivization” and “weakening”), “strengthening” is also included here as a major type of modality shifts because its raw frequency is comparatively high and its frequency is significantly affected by the identity of the ST speaker. They will be discussed in the order of their frequency: “weakening” (F = 844), “subjectivization” (F = 681), “strengthening” (F = 265) and “de-obligation” (F = 55). The distribution of the four major types of modality shifts in the use of the 15 English modality markers is presented in Table 4.12. It appears that “weakening” occurs mostly in the use of the central modals would (27.13%), will (17.30%), can (14.57%) and the mental verb structure I/we think (10.19%); “subjectivization” mainly occurs in the use of the central modals will (19.97%), would (17.91%), can (14.10%) and the mental verb structure I/we think (13.22%); “strengthening” mostly occurs in the use of the central modals will (34.34%), must (15.85%) and the verbal structure I/we believe (29.81%); “de-obligation” mainly occurs in the use of the central modals will (56.36%), would (20.00%) and the semi-modal be going to (10.91%). Now let’s discuss the interpersonal effects of these four major types of modality shifts through case studies of these major contributors.

Will Would Can Should Must May Could Need to Be going to Want to Be able to I/we think I/we Believe I/we hope Always Total

146 229 123 38 3 28 28 16 15 35 16 86 42 10 29 844

F.

17.30 27.13 14.57 4.50 0.36 3.32 3.32 1.90 1.78 4.15 1.90 10.19 4.98 1.18 3.44 100.00

%

Weakening (SV−)

136 122 96 33 4 20 10 12 16 35 15 90 51 14 27 681

19.97 17.91 14.10 4.85 0.59 2.94 1.47 1.76 2.35 5.14 2.20 13.22 7.49 2.06 3.96 100.00

Subjectivization (SOS) F. %

Table 4.12 Occurrence of major types of modality shifts with each EMM

91 18 4 6 42 0 0 7 3 1 0 11 79 0 3 265

F. 34.34 6.79 1.51 2.26 15.85 0.00 0.00 2.64 1.13 0.38 0.00 4.15 29.81 0.00 1.13 100.00

%

Strengthening (SV+)

31 11 2 0 0 1 0 0 6 4 0 0 0 0 0 55

56.36 20.00 3.64 0.00 0.00 1.82 0.00 0.00 10.91 7.27 0.00 0.00 0.00 0.00 0.00 100.00

De-obligation (O-I/P) F. %

4.2 Interpersonal Effects of the Major Types of Modality Shifts 69

70

4 Modality Shifts in the Use of English Modality Markers

4.2.1

Weakening and Subjectivization

Since the major contributors of “weakening” and “subjectivization” are the same four modality markers (will, would, can and I/we think), I will discuss these two types of shifts together through case studies of the four modality markers to explore their interpersonal effects. As the predominant types of semantic shifts in terms of modality value and orientation, “weakening” (844 cases) and “subjectivization” (681 cases) both occur frequently in the use of English modality markers. Most of “weakening” and “subjectivization” shifts occur with the addition of the English modality markers in the translation, as illustrated through the following case studies. Will Traditional grammar used to view will as a future tense marker, but I side with modern linguists such as Palmer (1990), Biber et al. (2000) and Huddleston and Pullum (2002) in the claim that there is no formal future tense in English and will is a marker of modality rather than tense. Although will can indicate future time among numerous other ways, it usually carries modal meanings of prediction or volition instead of conveying the meaning of “plain futurity”. Therefore, will is treated in this study as a medium-high-value or medium-value modal verb expressing either probability (equivalent to 会/将/将会 in Chinese) or inclination (equivalent to 要 in Chinese). Among the 1021 translation cases involving the use of will, 146 cases witness “weakening” shifts and 136 cases witness “subjectivization” shifts, accounting respectively for 17.30 and 19.97% of all the “weakening” and “subjectivization” shifts in the use of English modality markers in GPC interpreting. Most of the “weakening” and “subjectivization” shifts involving will co-occur in the same translation case with the addition of will in translating speeches of the Chinese officials and journalists. With these “weakening”/“subjectivization” shifts, the basic speech function of “statement” or “question” remains unchanged, but the Chinese government’s subjectivity is foregrounded and the speaker’s statement or question sounds more moderate and polite. For example: (1) ST: 这对维护半岛的和平, 乃至东北亚的和平与稳定都有重要意义, 也符合各方的 共同利益。 [LT: This has not only great significance for the maintenance of peace and stability on the Korean peninsula and Northeast Asia as a whole, but also serves the interests of all parties.] TT: This is not only of great significance for the maintenance of peace and stability on the Korean peninsula and Northeast Asia as a whole, but will also serve the interests of all parties. (2) ST: 那么如何确保反腐败的成效, 特别是如何有效地遏制一些行政领域权钱交易 的现象? [LT: Then how do you ensure the effectiveness of the anti-corruption measures, and particularly how do you effectively curb the trading of administrative power for money in the administrative area?]

4.2 Interpersonal Effects of the Major Types of Modality Shifts

71

TT: And my question is as follows. That, er, what will you do to ensure that our anti-corruption measures can effectively play out and what can you do to ensure that we can avoid the trading of administrative power for money, and particularly in the administrative area?

The ST of Example 1 is a speech segment of Foreign Minister Li Zhaoxing addressing the nuclear issue on the Korean peninsula in 2006. Following the suggestion that “the relevant parties” (ROK, DPRK, China, the US, Russia and Japan) demonstrate flexibility for consultation and dialogue and “strive to follow up on the Joint Statement”, Mr. Li here is stating the significance of these efforts. The polar statement “也符合各方的共同利益” could have been translated literally into “but also serves the interest of all parties”, but with the addition of the modal verb will, the assertiveness is weakened and the objective statement is turned into an subjective assessment. As the previous suggestion to “the relevant parties” and this statement about “the interest of all parties” both constitute face-threatening acts to the foreign journalists representing the relevant countries apart from China, the modalized TT sounds more polite and acceptable to the English-speaking audience than the literal translation. The ST of Example 2 is a question asked by a Chinese journalist on the anti-corruption issue to Premier Wen Jiabao in 2007. After mentioning the general public’s concern about the government’s anti-corruption measures taken recently, the journalist asks the above question in a concise way with the subject unidentified and no modality marker used. With only the subject you added, the literal translation “then how do you ensure…” would have conveyed the message faithfully, but it would sound abrupt and impolite. With the modal structure “what will you do to ensure…” instead, the modal strength of the clause is weakened and the addressee’s subjectivity is foregrounded, so the question sounds more proper and polite in the TT, although the primary listener of the TT is not the Premier but the foreign journalists. Would The modal verb would appears 325 times in the corpus and most of them (318 cases) are used as a tentative form of will instead of the past-tense form (only seven cases, involving no modality shift). Thus it is considered in this study as a medium-low value modal verb marking inclination (equivalent to 愿意 in Chinese) or probability (no equivalent form in Chinese). As Table 4.12 reveals, the use of would contributes to 27.13% (229 cases) of the “weakening” shifts and 17.91% (122 cases) of the “subjectivization” shifts in GPC interpreting. Most of the “weakening” shifts co-occur with the “subjectivization” shifts resulting from the addition of would in translating speeches of the Chinese officials and journalists (Examples 3 and 4). But there are also a considerable number of translation cases involving only the “weakening” shifts with would used in the TT to replace a higher-value modality marker in the ST (Examples 5 and 6). Due to the “weakening” shift either coupled with “subjectivization” or not, the speaker’s utterance

72

4 Modality Shifts in the Use of English Modality Markers

sounds more tentative and polite, but the basic speech function of the clause remains unchanged. (3) ST: 中国有13多亿人口, 实现现代化还有很长的路要走, 需要持久的世界和平 环境。

[LT: There are more than 1.3 billion people in China and there is a long way to go before we realize modernization. So that requires an international environment of lasting peace.] TT: And there’re more than 1. 3 billion people in this country. So we are on a long journey towards modernization. So that would require an international environment of lasting peace. (4) ST: 您, 呃怎么评价, 呃现在的日中关系?

[LT: How do you evaluate the current Japan-China relations?] TT: How would you characterize the current state of affairs in the Japanese-Chinese relations? (5) ST: 因此, 我留给下届政府的不只是债务, 而是25,000亿的优质的资产, 在将来中 国的经济发展中将长期发挥巨大的经济效益和社会效益。

[LT: So what I leave behind to the next government are not only debts, but also 2.5 trillion RMB yuan of excellent assets, which will generate great economic and social benefits lasting through China’s future economic development.] TT: So what I will leave behind to the next government is much more than just debts. Actually the 2.5 trillion RMB yuan will be excellent assets. And these excellent assets would generate very significant economic and social benefits for China’s future economic growth. (6) ST: 我想问的题目是, 财政风险是否可控?

[LT: The question I want to ask is: are the fiscal risks within control?] TT: So I would like to ask whether the fiscal risk, risks… are the fiscal risks within control? Example 3 involves a speech segment of Premier Li Keqiang in 2013 responding to the foreign media’s discussions on whether a stronger China will seek hegemony. Premier Li here states China’s need for peace before a direct answer “we will not seek hegemony”. The ST clause “需要持久的世界和平环境” is an objective statement about the required condition for China’s achievement of modernization, which is closely equivalent to “that requires an international environment of lasting peace”. But the interpreter has chosen to add the modal verb would, to the effect that Premier Li’s subjectivity in making this assessment is foregrounded and the message sounds more tentative and audience-friendly. The ST of Example 4 is a question addressed by a Japanese journalist to Premier Zhu Rongji in 2001 on the relations between Japan and China. After mentioning Japan’s recent amendments to the history textbook, here the journalist uses an unmodalized Chinese structure “您怎么评价…” to seek Premier Zhu’s view on the current Sino-Japanese relations. The question sounds polite enough in the Chinese ST with the honorific form of 2nd-person pronoun 您, but its close equivalent structure in English “how do you comment on…” would sound not so polite with no honorific form of you available. With the addition of the modal

4.2 Interpersonal Effects of the Major Types of Modality Shifts

73

would, the premier’s subjectivity is highlighted and the expected commitment is weakened, so the question in the TT sounds more polite and less face-threatening. The ST of Example 5 is a speech segment of Premier Zhu Rongji in 2002 responding to the question on the impact of China’s fiscal deficit on the next government. After justifying for the fiscal deficit, Premier Zhu uses the probability marker 将 to state his prediction about the benefit of the assets he is leaving behind to the next government. It could have been translated into its close equivalent will, but the interpreter opts for the tentative form would instead, to the effect that Premier Zhu sounds more humble in the TT. Example 6 illustrates a typical semantic weakening occurring in journalists’ questions. To avoid being too abrupt, the journalist here uses the semi-modal 想 to express his/her inclination to raise a question to Premier Wen Jiabao before giving the content of the question, which is a common discursive structure of questionasking in GPCs. The interpreter could have translated 想 into its close equivalent want to, but the lower-value modal phrase would like to is chosen here to make the journalist’s request sound more polite, so that the proper relationship between the journalist and the premier is reconstructed. Can As discussed in Sect. 3.1.2, the rare cases of can to express “pure ability” have been discarded in my data. So the valid concordance lines (318) of can include its use as a low-value marker of “probability”, “obligation” or “inclination”, equivalent to 能/能够 or 可以 in Chinese. As shown in Table 4.12, the use of can contributes to 14.57% (123 cases) of the “weakening” shifts and 14.10% (96 cases) of the “subjectivization” shifts in GPC interpreting. Similar to the case of will, most of the “weakening” and “subjectivization” shifts involving can co-occur in the same translation case with the addition of can in translating speeches of the Chinese officials and journalists. With these “weakening”/“subjectivization” shifts, the basic speech function of the clause hardly changes, but the speaker or addressee’s subjectivity is highlighted and the speaker’s statement or question sounds more moderate and consultative. For example: (7) ST: (我相信, 只要全体中华儿女共同努力, 祖国统一和民族振兴的大业一定能够 实现。) 这是整个中国人的骄傲。

[LT: …This is the pride of all Chinese people.] TT: (I truly believe that with joint efforts, of all compatriots of the Chinese nation, we will achieve the reunification and revitalization of the country.) And that is something that all Chinese can take pride in.

(8) ST: 究竟中英双方现在具体的分歧在哪里, 中方是否担心英国会在97年之前把香 港的储备花光?

[LT: What on earth are the differences between China and the UK? Is the Chinese government worried that the British side will deplete the treasury coffer of Hong Kong before 1997?]

74

4 Modality Shifts in the Use of English Modality Markers

TT: So can you tell us specifically the differences between the two sides, and also does it mean that the Chinese government is afraid that the British side will deplete the coffer, the treasury coffer of Hong Kong, before 1997? The ST of Example 7 is a speech segment of Premier Wen Jiabao in 2012 addressing the Taiwan question. Premier Wen used a polar statement “这是整个中 国人的骄傲” to expresses the pride in China’s reunification and revitalization on behalf of the all Chinese people in order to highlight our collective stance on the Taiwan Question. If translated literally into “This is the pride of all Chinese people” or “That is something that all Chinese people take pride in”, it might sound too assertive to western ears. By adding the probability marker can, the interpreter weakens the assertion and subjectivizes the as-a-matter-of-fact statement, making the TT sound more consultative and convincing. In Example 8, the ST is a question posed by a Hong Kong journalist in 1991 to Premier Li Peng concerning the difficulty for a Sino-British agreement on the construction of a new airport in Hong Kong. The journalist’s original question in Chinese is very direct and face-threatening, closely equivalent to “What on earth are the differences between…?” Although the English TT is not meant for the premier to hear, the interpreter still adds a weaker and more subjective modal structure “can you tell us” to make the question sound more polite and proper in this GPC setting. I/we think With a high occurrence of 169 times, the mental verb structure I/we think (mostly I think, equivalent to 我想 or 我认为) is used typically in the TTs as a medium-value subjective marker of probability. As Table 4.12 reveals, I/we think is also a major contributor to “subjectivization” (13.22%, 86 cases) and “weakening” (10.19%, 90 cases) shifts in GPC interpreting. Similar to the case of will and can, most of the “weakening” and “subjectivization” shifts involving the use of I/we think co-occur in the same translation case with the addition of I think (the use of we think involving these two shifts is rare). But different from the case of will and can, these shifts in the use of I think occur exclusively in the translation of Chinese officials’ speeches (with only one exceptional case for the speech of the chairperson). With the addition of I think, the basic speech function of statement remains unchanged, but the speaker’s subjectivity is explicitly foregrounded and the statement sounds less assertive and more friendly than the literal translation. For example: (9) ST: 人民币能否成为世界货币, 这个问题不应该由我来回答, 而且现在还言之过 早。

[LT: The question as to whether RMB can become a world currency should not be answered by me and it is still too early now to discuss it.] TT: The question whether the RMB will become a world currency, in my view, is not a question that should be answered by me and I think it is too early now to discuss this question.

4.2 Interpersonal Effects of the Major Types of Modality Shifts

75

Example 9 presents Premier Zhu Rongji’s response to a foreign journalist’s question on the “possibility for Chinese RMB to become one of the major currencies in the world” in 1999. After a modalized negative statement on his position to answer this question “这个问题不应该由我来回答”, Premier Zhu uses a polar structure “现在还言之过早” to turn down the request to discuss the question, implying that this was not a proper question to be asked at that moment. With the subjective marker I think added in the TT clause, the polar statement is turned into a subjective and tentative opinion, so the TT sounds more friendly and less face-threatening to the western audience.

4.2.2

Strengthening

Compared with “weakening”, “strengthening” occurs with a much lower frequency in the use of English modality markers in GPC interpreting. But different from “weakening” which occurs mainly with the addition of English modality markers in the clause/sentence translation, all “strengthening” shifts occur with corresponding forms available in the ST for the English modality markers in the TT. Thus “strengthening” takes the upper hand if we only consider the use of English modality marker in the TT to replace a corresponding form in the ST (see Table 4.9). Besides, the raw frequency of “strengthening” is pretty high (265 cases) and its relative frequency is significantly affected by the identity of the ST speaker, thus it is important to discuss the interpersonal effect of “strengthening”. As the shift of “strengthening” mainly occur in the use of the modals will and must and the mental verb structure I/we believe, I will examine the translation cases involving the three modality markers to explore the interpersonal effects of “strengthening”. Will As shown in Table 4.12, the use of will contributes to 34.34% (91 cases) of all “strengthening” shifts in modality. Out of the 91 cases, 77 cases occur in the translation of premiers/foreign ministers’ speeches, 12 cases occur in the translation of journalists’ speeches. Comparing these frequencies against the relative occurrence of English modality markers used to translate different speakers (F1 = 2597; F2 = 361), we can see that the probability of “strengthening” shifts in the use of will is not significantly affected by the identity of ST speaker (LL = −0.132; p = 0.716). In other words, the use of will contributes evenly to “strengthening” shifts in translating speeches of Chinese officials and journalists. First, the “strengthening” shifts in translating speeches of Chinese officials consist in the use of will in the TT to replace ST modality markers such as 能/能够, 愿意, 可以, etc. The basic speech function of “suggestion”, “statement” or “offer” in the relevant clause/sentence remains unchanged, but the speaker sounds more powerful, confident or sincere in the TT than in the literal translation. For example:

76

4 Modality Shifts in the Use of English Modality Markers (10) ST: 我们希望, 新, 美国的新政府能为改善中美关系做出自己的努力。

[LT: We hope that the new US Administration can make their efforts to improve Sino-American relations.] TT: We hope that the new US Administration will make efforts for the improvement of Sino-American relations. (11) ST: 你不要着急, 这个问题很快就可以解决。

[LT: Don’t be anxious. This problem can be solved very soon.] TT: Just be a little bit patient because this question will be resolved very soon. (12) ST: 中方呢愿意同有关国家、组织共同努力, 按照联合国气候变化框架公约及 其京都议定书, 按照巴厘路线图, 按照共同但有区别的责任的原则, 同各国一起 积极地推进应对气候变化的进程, 使得今年在应对气候变化方面各国的努力能 够取得更大的成果。

[LT: China is willing to work hard with other countries and organizations in meeting the climate challenge according to the United Nations Framework Convention on Climate Change and its Kyoto Protocol, the Bali Roadmap and the principle of common but differentiated responsibilities, so that efforts of all countries in the world in meeting the climate challenge can yield greater results this year.] TT: China will work actively with other countries in the world in meeting the climate challenge according to the United Nations Framework Convention on Climate Change and its Kyoto Protocol, the Bali Roadmap and the principle of common but differentiated responsibilities, so that efforts of all countries in the world in meeting the climate challenge will yield more positive results this year. Example 10 displays a typical “strengthening” shift from the lower-value 能/能 够 (equivalent to can) to the higher-value will in translating clauses led by 我/我们 希望 (I/we hope). The ST here is a speech segment of Foreign Minister Qian Qichen in 1993 responding to some American officials’ remark that “the renewal of its MFN status to China will hinge upon progress China can make in human rights, trade and non-proliferation issues” quoted by the journalist. After stating China’s disagreement on attaching any political condition to MFN (Most Favored Nation) treatment, Mr. Qian uses the structure 我们希望…能 (we hope…can) to suggest the new US administration “make efforts for the improvement of Sino-American relations”, implying that they should stop attaching political condition to establishing reciprocal trade relations with China. With 能 translated into will, the TT places more pressure on the American government and demonstrates a more forceful position of China than the literal translation. The ST in Example 11 is a speech segment of Premier Zhu Rongji in 2000 responding to a series of questions posed by a foreign journalist on the Chinese government’s attitude towards the issue of the People’s Liberation Army operating and collecting revenue from commercial mobile phone networks. After admitting that several pilot programs concerning CDMA (Code Division Multiple Access) have been conducted jointly by China Telecom and some entities under direct jurisdiction of the PLA, and stating that “later the Military Commission made the

4.2 Interpersonal Effects of the Major Types of Modality Shifts

77

decision that all the PLA should cut off all the business links with the business and commercial entities that used to be run under their jurisdiction”, Premier Zhu utters the above sentence to state his prediction about solving the problem soon. The low-value modal 可以 is used in the ST as a hedging device to make the statement sound not too assertive, but Premier Zhu’s confidence is implied in the co-text, so the translation of 可以 into the higher-value modal will more clearly demonstrates Premier Zhu’s confidence in properly solving the problem. Example 12 illustrates a typical “strengthening” shift in the use of will to translate the inclination marker 愿意 (equivalent to be willing to or would like to) in clauses that function as an offer of the Chinese government. The ST here is a speech segment of Foreign Minister Yang Jiechi in 2010 addressing the issue of “the climate negotiations after the Copenhagen Conference”. Mr. Yang uses “愿意 ” to state China’s inclination to work hard and cooperate with other countries in meeting the climate challenge, which functions as an offer to the international audience. With the use of a higher-value inclination marker will to replace 愿意 in the interpretation, the Chinese government’s offer sounds more sincere in the TT than in its literal translation. Second, the “strengthening” shifts in translating questions of journalists consist in the use of will in the TT to replace ST modality markers such as 能/能够, 可以, etc. These modality markers are used to realize the speech function of “question”, requesting the addressee’s (i.e., the Chinese officials representing the Chinese government) opinion. With the “strengthening” shift in modality, the speech function remains unchanged, but a more confident and powerful position of the addressee or the Chinese government is presupposed in the TT than in its literal translation. For example: (13) ST: 请问中国认为什么时候能够重开六方会谈呢?

[LT: May I ask, when can the six-party talks be resumed in China’s view?] TT: In China’s view, when will the six-party talks reopen, be reopened? The ST of Example 13 is a question raised by a South Korean journalist in 2005 to Foreign Minister Li Zhaoxing concerning the nuclear issue on the Korean Peninsula. The journalist uses the low-value modality marker 能够 (equivalent to can or be able to) with the subjective marker 中国认为 (China thinks) to seek the Chinese government’s prediction on the time for reopening the six-party talks. With the translation of 能够 into a higher-value modal will, a more confident and powerful position of the Chinese government on the nuclear issue on the Korean Peninsula is presupposed in the TT, so that China’s influential position on this issue is better conveyed to the English-speaking audience. Must The 99 valid concordance lines of must extracted from the corpus witness the its typical use as a high-value obligation modal closely equivalent to 必须 or 一定要 in Chinese. As revealed by Table 4.12, the use of must contributes to 15.85% (42 cases) of all “strengthening” shifts in modality. Different from the case of will,

78

4 Modality Shifts in the Use of English Modality Markers

almost all “strengthening” shifts involving the use of “must” occur in translating the speeches of Chinese officials (41 out of 42) with only one exceptional case in translating a journalist’s quotation of the premier. Apparently, the probability of “strengthening” shifts in the use of must is significantly affected by the identity of the ST speaker. The “strengthening” shifts with must in translating speeches of Chinese officials consist in the use of the obligation or inclination marker must in the TT to replace ST modality markers such as 要 and 应该. The basic speech function of “statement”, “offer” or “command/suggestion” in the relevant clause/ sentence remains unchanged, but the speaker sounds more assertive, forceful or determined in the TT than in the literal translation. For example: (14) ST: 将心比心啊, 政府工作人员啊应该以百姓之心为心。

[LT: We should try to feel for others. Government employees should put themselves in the positions of the people and feel their hearts.] TT: Just imagine what if such situation happens on ourselves. All government employees must always put such needs, special needs of our people on the top of their minds. (15) ST: 我们对藏族同胞要1 采取平等和尊重的态度, 并且不断地改善我们的工作。

[LT: We should treat our Tibetan compatriots as equals and with respect and improve our work continuously.] TT: We must treat all our Tibetan compatriots as equals and with respect and make continuous improvement in our work in this area. (16) ST: 日方要以严肃认真的态度处理好历史问题。

[LT: The Japanese side should handle well the historical issue with a serious and earnest attitude.] TT: The Japanese side must adopt a serious and prudent attitude to handle well the historical issue. Example 14 illustrates the interpretation of Premier Li Keqiang’s speech segment in 2014 on improving people’s livelihood in China. After mentioning “the occurrence of such instances in which people become homeless or people give up on seeking medical treatment because they cannot afford it”, Premier Li uses the medium-value obligation modal 应该 (equivalent to should) to state government employees’ obligation (including himself) to place people’s needs “on the top of their minds”. With the higher-value modal must used to replace 应该, the speaker sounds more assertive and sincere in emphasizing the government employees’ obligation to take care of people’s needs, so that the positive image of a highly responsible government is reconstructed in the TT.

Note that in some Chinese-English dictionaries, 要 is explained as an ambiguous item equivalent to either will, should or must, but a difference in modality value is detected between 要 and must in this study as the compound form 一定要 equivalent to must has an apparently higher value than 要 used alone. 1

4.2 Interpersonal Effects of the Major Types of Modality Shifts

79

Example 15 involves Premier Wen Jiabao’s speech segment in 2012 responding to the recent incidents of self-demolition on the part of Tibetan monks. After expressing deep sorrow for the innocent monks and the Chinese government’s firm position against the political forces trying to separate Tibet from China, Premier Wen here uses the ambiguous modal 要 (closely equivalent to will or should) to express the central government’s inclination/obligation to treat “all our Tibetan compatriots as equals and with respect” to ensure its stability and development. The interpreter translates 要 into the higher-value modal must, so that the government sounds more sincere and determined in its offer to respect and befriend Tibetan compatriots in the TT. The ST of Example 16 is a speech segment of Foreign Minister Li Zhaoxing in 2006 responding to a question on Sino-Japanese relations raised by a Japanese journalist. After emphasizing the fact that the current difficulty in the political relationship between China and Japan is mainly caused by some Japanese leaders still “paying homage to Class-A war criminals who have launched and commanded that war of aggression”, Mr. Li here uses the obligation modal 要 (closely equivalent to should) to press the Japanese leaders to deal with the historical issue with more prudence. With 要 translated into the higher-value modal must, the speech function of “command/suggestion” remains unchanged, but more pressure is placed on the Japanese government in the TT. Considering the Chinese government’s consistent firm position on this issue, I would see this “strengthening” shift as a proper reconstruction of modality to convey the intended message of the speaker to the TT audience. I/we believe Besides the two central modals, the mental verb structure I/we believe is also a major contributor (29.81%, 79 cases) of “strengthening” shifts in GPC interpreting. As another frequently used subjective modality marker, I/we believe occurs 154 times in the corpus, including 114 cases of I believe and 40 cases of we believe. Equivalent to 我/我们相信 in Chinese, I/we believe is used typically in GPC texts as a high-value subjective marker of probability. Similar to the case of must, almost all “strengthening” shifts involving the use of I/we believe occur in translating the speeches of Chinese officials (78 out of 79) with only one exceptional case in translating a journalist’s speech. The LL test (LL = 13.785; p < 0.01) reveals that the probability of “strengthening” shifts in the use of I/we believe is significantly affected by the identity of the ST speaker. These “strengthening” shifts in translating speeches of Chinese premiers/foreign ministers consist in the use of the high-value subjective marker I/we believe in the TT to replace lower-value subjective markers in the ST such as 我/我们认为 and 我/我们想. The basic speech function of “statement” in the relevant clause/sentence remains unchanged, but the speaker sounds more assertive and confident in the TT than in the literal translation. For example:

80

4 Modality Shifts in the Use of English Modality Markers (17) ST: (卢旺达总统卡加梅先生说, 中国的投资和贷款带来了非洲所需要的东西。 ) 我想这些评价是公道的。

[LT: I think these remarks are fair.] TT: (President Kagame of Rwanda remarked that China’s investment and loans have brought what Africa needs.) I believe these are fair statements. (18) ST: 我们之所以提出和平统一、一国两制的方针, 就是我们认为, 这样的方针符 合祖国大陆和台湾人民的现实利益和长远利益。

[LT: We put forward the policies of “peaceful reunification” and “one country, two systems” because we think that these policies accord with the practical and long-term interests of the Chinese people in the mainland and in Taiwan.] TT: The reason why we put forward the policies of “peaceful reunification” and “one country, two systems” is because we believe that this policy is in the immediate and long-term interests of all the Chinese people on the mainland and on Taiwan.

The ST in Example 17 is a speech segment of Foreign Minister Yang Jiechi in 2010 responding to some western media’s accusation of China of going to Africa only for energy and resources. To correct this misunderstanding, Mr. Yang here quotes President Kagame of Rwanda’s positive remark that “China’s investment and loans have brought what Africa needs” and states that “这些评价是公道的” (“these are fair statements”). In a humble manner, Mr. Yang uses 我想 (I think) to hedge the assertion that “这些评价是公道的”. However, the interpreter renders it into the higher-value subjective marker I believe, so Mr. Yang sounds more assertive and his real intention of stressing his agreement with President Kagame’s remark is conveyed more directly. The ST in Example 18 is a speech segment of Premier Wen Jiabao in 2004 responding to the issue of the election and referendum to be held in Taiwan raised by a Taiwanese journalist. Emphasizing the central government’s consistent principles of “peaceful reunification” and “one country, two systems”, Premier Wen uses the medium-value subjective marker 我们认为 (closely equivalent to we think) to state the government’s opinion that these policies are beneficial to all the Chinese people across the Taiwan straits. With 我们认为 replaced by the higher-value subjective marker we believe, the speaker representing the Chinese government sounds more assertive and confident in the benefit of the policies of “peaceful reunification” and “one country, two systems” to Chinese people across the straits, thus the firm stance of the Chinese government on the Taiwan question is clearly conveyed to the international audience.

4.2.3

De-obligation

Generally speaking, shifts of modality type occur rather infrequently compared with shifts of modality value and shifts of orientation, but the tendency of

4.2 Interpersonal Effects of the Major Types of Modality Shifts

81

“de-obligation” deserves our attention as the frequency of “de-obligational” shifts (including shifts from obligation to inclination and shifts from obligation to probability) is significantly higher than the frequency of other shifts in modality type (see Sect. 4.1.1). Since “de-obligational” shifts mainly occur in the use of will, would and be going to, I will examine the translation cases involving the three modality markers closely to explore the interpersonal effects of these shifts. Will As Table 4.12 reveals, the use of will contributes to 56.36% (31 cases) of all “de-obligational” shifts in the use of English modality markers. Among them there are 20 cases of shifts from obligation to inclination (O-I), and 11 cases of shifts from obligation to probability (O-P), both consisting in the use of will to replace obligation markers in the ST such as 应该, 应当, 需要, 要, etc. First, most of the translation cases involving the modality shift of “O-I” have 我 们/we (referring to the Chinese government) or 政府 (the government) as the modal subject (sometimes implied), so the obligation markers in the STs are typically used to state the obligation of the Chinese government. With the obligation markers replaced with the inclination marker will, the speech function of “statement” (giving information) is changed into “offer” (giving goods-&-services). For example: (19) ST: 从方向上看, 应该鼓励更多的人租房。

[LT2: In terms of the direction, we should encourage more people to rent homes.] TT: We will continue to encourage people to rent homes. (20) ST: 当然, 我们也需要啊推进公租房等保障房建设, 让新就业的年轻人和长期进 城务工的人员等啊, 来缓解住房困难, 并且呢实行公平分配。

[LT: Of course, we also need to speed up the building of governmentsubsidized housing, such as public rental housing, to ease the housing difficulty for those newly-employed young people and rural migrant workers who have lived in the city for a long time, and ensure that such housing units will be distributed in a fair manner.] TT: At the same time, we will also build more government-subsidized housing, such as public rental housing to ease the housing difficulty for those young people who have just got employed and rural migrant workers, and ensure that such housing units will be distributed equitably. The ST of Example 19 is a speech segment of Premier Wen Jiabao in 2012 addressing a journalist’s inquiry about the target of the recent round of housing market regulation and whether it would be “abandoned halfway” under the economic pressure. Followed by two strong statements “房价还远远没有回到合理价 位” (“the home price in China is far from coming back to a reasonable level”) and

2

LT is short for literal translation, which will be offered for each translation case for readers to compare.

82

4 Modality Shifts in the Use of English Modality Markers

“调控不能放松” (“we must not slack our efforts in regulating the housing sector”), this statement “应该鼓励更多的人租房” is meant to convey the Chinese government’s attitude towards whether it is necessary for everyone in China to purchase a house. On the surface, it is a statement of the government’s obligation to “encourage people to rent homes”, but what it implies is the government’s actual attitude of “encouraging people to rent homes”, which would help bring down the housing price. By translating the obligation modal 应该 into the inclination modal will, the interpreter turns the statement of obligation with an implied performative force into an offer that the government is making to the audience. Compared with the literal transition “we should encourage more people to rent homes”, the TT above is more relevant to the journalist’s question and therefore it would sound more sincere and responsive to the English-speaking audience. Example 20 is the interpreting case of Premier Li Keqiang’s speech segment in 2014 on the same housing issue. In response to the journalist’s inquiry about “new ideas and measures of the government taken on this issue”, Premier Li uses the obligation marker 需要 to state the government’s obligation to “build more government-subsidized housing”, which also implies the government’s inclination to do so. The literal translation with its equivalent form need to would have conveyed the ST message adequately into the TT, because the English obligation marker need to can also imply the government’s inclination. But the interpreter renders it into the inclination marker will, so the statement of obligation becomes a clear offer, which again sounds more responsive and more appealing to the audience. Second, most of the translation cases involving the modality shift of “O-P” occur either with “zero-subject” clauses, or with implicit passive-voice clauses in the ST. With the obligation marker in the ST rendered into the probability marker will in the TT, the statement of the obligation is turned into a prediction or estimation. For example: (21) ST: 我估计明年可能还需要再发1500亿。

[LT: I estimate that we may need to issue another 150 billion RMB yuan of treasury bonds next year.] TT: And I expect that in the next year, maybe another 150 billion RMB yuan of treasury bonds will be issued. (22) ST: 比方说, 我们的个人的所得税制, 最多的应该交45%的税, 也就是他的收入的 一半应该交给国家。

[LT: For instance, according to our personal income tax policy, the highest tax rate should be 45%. In other words, half of his income should be submitted to the country.] TT: For instance, the highest tax rate for personal income tax can be as high as 45%. In other words, nearly half of the personal income will be turned over as tax. Example 21 involves a speech segment of Premier Zhu Rongji in 2001 addressing the question about China’s financial deficit and the possible fiscal risk of issuing treasury bonds. After reassuring the audience that “all the deficit has been

4.2 Interpersonal Effects of the Major Types of Modality Shifts

83

used to support the infrastructure projects” and there is no real risk of issuing treasury bonds, Premier Zhu here uses the obligation marker 需要 to express the government’s internal obligation to issue another 150 billion yuan of treasury bonds next year to “support the ongoing projects and also to undertake new projects in the development of China’s western region”. But with the use of 我估计 in the main clause, the whole statement is meant primarily to convey the speaker’s prediction about the figure of next year’s treasure bonds instead of the government’s obligation to do so. Therefore, with the obligation marker 需要 rendered into the probability marker will, the speaker’s main point of prediction is more clearly conveyed, which is more appealing to the audience than the statement of obligation. Example 22 involves a speech segment of Premier Zhu Rongji in 2001 responding to the question about the measures that the government will take to reduce the income gap between different groups of people. After making a promise to “resort to taxation as an approach to narrow the income gap”, Premier Zhu uses 应该 to state the obligation of 他 (he, referring to the richest group) to pay a high income tax (45%) to the country according to the current personal income tax policy in China. Then he admits that “the policy in taxation has not been adequately implemented” in the following sentence. With the obligation modal 应该 translated into the probability modal will, the statement of obligation is turned into an a statement of estimation about the effect of this income tax policy if it is fully implemented. Compared with the literal translation, this interpretation more clearly conveys Premier Zhu’s confidence in easing the income gap problem through taxation. Would Among all the 55 “de-obligational” shifts in the use of English modality markers, 11 cases (20%) involve the use of would, which are mostly (10 out of 11) shifts from obligation to inclination (O-I). These shifts consist in the use of would (mostly with a first-person subject and often in the phrase would like to) in the TT to replace obligation markers in the ST such as 需要, 必须, 要, 建议, etc. With these “de-obligational” shifts, the basic speech function remains unchanged, but more delicate interpersonal effect is brought on: either the “offer” is realized more directly and explicitly, or the “command” becomes more indirect and polite. For example: (23) ST: 其实, 人们没有读懂中国所采取的一揽子计划的全部内涵, 我需要借这个机 会再扼要地向大家介绍一下。

[LT: In fact, people haven’t fully understood the essence of the package plan adopted by China. I need to take this opportunity to give you another brief introduction.] TT: In fact, I don’t think people were really appreciating what we mean in the package plan, which is very important. I would like to take this opportunity to give you an outline of this plan. (24) ST: 所以为了不浪费大家的时间, 我建议你再去读读我的报告。

[LT: So in order to not to waste everyone’s time, I suggest you read my report on the work of the government again.]

84

4 Modality Shifts in the Use of English Modality Markers

TT: So in order to not to waste the time of others, I would like to refer you again to my report on the work of the government. Example 23 is an instance of “de-obligationization” which results in turning an indirect “offer” into a direct one. Addressing the question about the Chinese government’s stimulus package plan to tackle the international financial crisis, Premier Wen Jiabao uses the semi-modal 需要 in the ST to express his strong internal obligation to clarify the meaning of the “package plan” to the media in response to “some rumors and some misunderstanding”, which is followed by his introduction of the four crucial aspects of the plan. On the surface, the ST clause “我需要借这 个机会再扼要地向大家介绍一下” functions as a statement of the speaker’s obligation. But in effect, the preposition functions as an indirect “offer”, for it initiates the act of introducing the plan. Of course, 需要 could have been translated into the English equivalent need to, but the interpreter opts to use the inclinational modal phrase would like to to express this “offer” more directly and explicitly (with the modality value reduced at the same time), which sounds more appealing and comprehensible to the English-speaking audience. Example 24 illustrates the translation shift of “de-obligationization” which results in turning a direct “command” into an indirect one. This ST is a speech segment uttered by Premier Li Peng in 1990 in response to a foreign journalist’s question on the Chinese government’s plan to maintain economic, political and social stability. After mentioning that the journalist’s question has already been covered in the Government Work Report released at the just-concluded session of the National People’s Congress, here Premier Li uses the lexical verb 建议 to advise the journalist to read the report again, implying that he would not answer the question here. Apparently, this direct speech act of suggestion (one type of command) coupled with the refusal to answer the journalist’s question would seriously threaten the journalist’s face if translated literally. With the obligation marker replaced by the inclinational phrase would like to followed by refer you to…, the speech function of “suggestion”/“command” is realized in a milder and less direct way, thus the Premier sounds more polite and accommodating to the foreign addressee in the TT. Be going to The semi-modal be going to occurs 110 times in the corpus and it is used as a medium-value or medium-high-value probability/inclination marker equivalent to 将, 要 or 准备. Out of the 55 cases of “de-obligational” shifts in the use of English modality markers, 6 (10.91%) involve the use of be going to, which are all shifts from obligation to inclination (O-I). In these translation cases, be going to is used in the TT to replace obligation markers in the ST such as 应该, 该, 要, etc. Compared with “de-obligational” shifts involving the use of will and would which occur mostly in the speeches of premiers and foreign ministers (PMT), “de-obligational” shifts in the use of be going to occur mostly in the speeches of journalists (JNL). With these modality shifts, the speech function of “question” remains unchanged, but the focus of the question is shifted from the obligation of the Chinese

4.2 Interpersonal Effects of the Major Types of Modality Shifts

85

government to their plan for future actions, which more closely relates to the questioner’s intention and the audience’s interest. For example: (25) ST: (请问总理, 现在世界各国都在谈论建立国际新秩序, 你认为应该建立什么样 的国际新秩序?) 我国政府应该为此做出哪些努力?

[LT: What efforts should our government make to establish a new world order?] TT: (Premier Li, now all the countries in the world are talking about the establishment of a new world order. So what kind of new world order do you think that should be established?) And what efforts is the Chinese government going to make to the establishment of that kind of new world order? (26) ST: (我认为呃中国现在的社会阶层两极分化比较厉害, 所以呢有一些中国人的 专家也认为, 要改善这个现象, 培育和扩大那个中层阶层, 就是中产阶层是…是 必要的, 你同不同意这样的看法?) 如果同意的话呢, 应该采取什么样的措施来培 育那个这样的那个中产阶层?

[LT: If you agree, what measures should the Chinese government take to cultivate the so-called middle-class?] TT: (Now in the Chinese society, there has been a problem of polarization between the high-income people and low-income people. Some Chinese experts have also called for the improvement of this situation. And they think it is necessary to cultivate and expand so-called middle-class.) Do you agree to such a view? And what measures are you going to adopt if you agree to their view? The ST of Example 25 is a question raised by a Chinese journalist to Premier Li Peng in 1991. Against the background of the international discussion about “the establishment of a new world order”, the journalist uses the obligation marker 应该 to demand Premier Li’s opinion on what efforts the Chinese government should make to establish a new world order. What implies is the journalist’s interest in what the Chinese government is planning to do, which is shared by the other journalists in the audience. With the obligation marker in the ST replaced by the inclination marker be going to, the intentional meaning of the question-asker is conveyed more clearly than the literal translation to the English-speaking audience, and Premier Li’s following statement of “what should/must be done” implying what the Chinese government offers to do would be easily comprehensible to the English-speaking audience. The ST of Example 26 is a question raised by a foreign journalist to Premier Zhu Rongji in 2002. Quoting some experts’ proposal of cultivating and expanding the middle-class to ease the problem of polarization between the rich and poor in China, the journalist asks Premier Zhu about how to cultivate this middle-class if he agrees to this view. Although the obligation marker 应该 is used to pose the question, the journalist’s major concern is not the Chinese government’s obligation, but their plan or future action. Therefore, similar to Example 7, the obligation marker 应该 is replaced by the inclination marker be going to and the intentional meaning of the question-asker is better conveyed to the other English-speaking

86

4 Modality Shifts in the Use of English Modality Markers

journalists. Besides, without the subjective marker 你/您认为 used in the co-text, the ST question with 应该 sounds not very polite because the foreign journalist could be understood as exerting obligational pressure on the Chinese government. With the modality shift, the question in the TT sounds more polite and the proper power relationship between the journalist and the addressee is reconstructed.

4.3

Summary

In this chapter, I have discussed not only the general patterns of modality shifts in the use of high-frequency English modality markers in Chinese-English GPC interpreting but also explored the interpersonal effects of the major types of modality shifts through case studies of the major contributors. Analysis of the semantic shifts of modality through the use of English modality markers reveals that modality shifts occur less frequently than zero shifts in Chinese–English GPC interpreting, but the degree of modality shifts is still high, for which an important reason is the frequent addition of English modality markers in the interpreting. The distribution of modality shifts in the three dimensions follows the pattern that shifts of modality value are the most common, shifts of orientation are fairly common and shifts of modality type the least common. Within the three types of semantic shifts, “de-obligation”, “subjectivization” and “weakening” shifts are predominant respectively. A comparative analysis of modality shifts in translation cases with different ST speakers shows that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the use of English modality markers in Chinese-English GPC interpreting but does affect the degree of the “strengthening” shift, which features more prominently in translating speeches of Chinese officials than journalists (mainly in the use of must and I/we believe). Note that so far all these discussions are only based on one side of the story: the use of the English modality markers. Whether modality shifts on the translation of Chinese modality markers will reveal similar or different patterns, it is yet to be explored in Chap. 5. The interpersonal effects of the major types of modality shifts are revealed through the case studies of their major contributors: (1) The shifts of “weakening” and “subjectivization” mostly occurring with the addition of will, would, can and I think have the effect of foregrounding the Chinese government’s subjectivity and making the speaker (the Chinese official or the journalist) sound more friendly and polite. (2) The “strengthening” shifts mainly involving the use of will, must and I/we believe have the effect of presenting the Chinese government as more confident, sincere and powerful on important political issues. Even in the case of will used in translating journalists’ speeches, “strengthening” shifts also have the primary effect of presenting the Chinese government as more confident and powerful, because will involving “strengthening” shifts is mainly used in a journalist’s question requesting the addressee’s (i.e., the Chinese officials representing the

4.3 Summary

87

Chinese government) opinion. It appears that the effect of the ST speaker’s identity on the degree of “strengthening” shifts is mainly reflected in the use of must and I believe, because these two modality markers are typically used to express the opinion or attitude of the speaker rather than that of the addressee, and the interpreter is not as motivated to strengthen journalists’ opinions/attitudes as to strengthen Chinese officials’. (3) The “de-obligation” shifts mainly occurring in the use of will, would and be going to either render the Chinese officials’ speeches more sincere and responsive, or make the journalists’ questions more explicit and appealing to the audience. Despite the frequent shifts in the semantics of modality, the basic interpersonal speech function—statement/question/offer/command (suggestion is taken as one specific type of command) remains unchanged, except for the “de-obligational” shifts in terms of modality type. In a word, the interpreters frequently allow modality shifts instead of always obeying the rule of “equivalence” in their use of English modality markers mainly for achieving a better communicative effect. In the next chapter, I will investigate modality shifts from another direction, the translation of Chinese modality markers, and also discuss the interpersonal effects of the major types of modality shifts, followed by a summary of the regular patterns of the overall reconstruction of modality in Chinese-English GPC interpreting.

References Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 2000. Longman grammar of spoken and written English. Beijing: Foreign Language Teaching and Research Press. Huddleston, Rodney, and Geoffrey K. Pullum. 2002. The Cambridge grammar of the English language. Cambridge: Cambridge University Press. Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese–English): Face management and information management. Ph.D. Thesis, Heriot-Watt University. Palmer, Frank. 1990. Modality and the English modals. London: Longman Group Limited. Wang, Binhua 王斌华. 2009. 口译规范描写及其应用——一项基于中国总理 “两会” 记者招 待会交传语料的研究 [Description of norms in interpreting and its application—a study based on the corpus of consecutive interpreting in Chinese premier press conferences]. Ph.D. Thesis, Guangdong University of Foreign Studies. Wang, Binhua. 2012. A descriptive study of norms in interpreting: Based on the Chinese–English consecutive interpreting corpus of Chinese premier press conferences. Meta 57 (1): 198–212.

Chapter 5

Modality Shifts in the Translation of Chinese Modality Markers

5.1

General Features of Modality Shifts

As discussed in Sect. 3.2.2, 15 Chinese modal forms are identified as high-frequency Chinese modality markers for the present study. Similar to the study of English modality markers, negative patterns (e.g., 不能, 不会) and noisy homonyms (e.g., 将 as a preposition, 需要 as a lexical verb followed by a nominal object) are discarded. In addition, compared with English, the Chinese high-frequency modality markers more often occur in pairs and function as a modal phrase, so these cases are analyzed under one modality marker and discarded in the other to avoid repetition. For example, 一定要, 一定会 and 一定能/能够 are analyzed as a phrase under 一定; 将会 and 会要under 会; 可能会 and 可能要 under 可能; 必须要under 必须; 应该要 under 应该. Finally, a total number of 2144 parallel concordance lines are generated for the present study, with the distribution of Chinese modality markers presented in Table 5.1.

5.1.1

Overall Degree and Distribution of Modality Shifts

Following the same coding scheme (see Table 3.6) for analyzing modality shifts occurring in the use of high-frequency English modality markers, all the 2144 parallel concordance lines involving the translation of high-frequency Chinese modality markers are coded manually for semantic shifts of modality, with both the first-level variable “SEMANTIC SHIFT” (SST) and the second-level variables— SHIFT of MODALITY TYPE (SMT), SHIFT of ORIENTATION (SO), and SHIFT of MODALITY VALUE (SV)—attended to. The inter-coder agreement on the coding of the sample data in this direction (500 concordance lines) is 85.7%, also suggesting a sufficiently high level of inter-coder reliability of the semantic coding. For convenient reference to the coding scheme, the original codes involved © Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_5

89

90

5 Modality Shifts in the Translation of Chinese Modality Markers

Table 5.1 High-frequency CMMs extracted from the corpus CMMs in the ST

Occurrence

Central modal (1466)

Semi-modal(134)

1st-person subject + mental lexical verb + clause (353)

Modal adverb (191) Total

要 会 能/能够 可以 应该 愿意 可能 一定 必须 需要 我/我们想 我/我们认为 我/我们相信 我/我们希望 将

548 251 194 197 144 83 49 53 44 37 158 84 50 61 191 2144

for various categories of modality shifts will also be indicated in parentheses in the tables of this chapter. First, let’s look at the overall degree of modality shifts revealed by the translation of Chinese modality markers. As Table 5.2 shows, of all the 2144 cases of Chinese modality markers used in the STs of our GPC corpus, 39.79% are translated with semantic shifts and 58.21% involve no semantic shifts of modality. Similar to what’s found in Chap. 4 about the use of English modality markers, the LL test (LL = 74.706; p < 0.01) confirms the pattern that compared with the option of modality shifts (SST-C/ZC), equivalent translation (ZSST) is the preferred choice in the translation of Chinese modality markers. However, as the English equivalent modal expressions for these Chinese modality markers are available, the degree of modality shifts (39.79%) is still high, and it is important to explore how and why semantic shifts of modality occur in these cases. Before exploring the distribution of modality shifts in the three semantic dimensions, let’s explore to what extent these semantic shifts of modality in the translation of Chinese modality markers are caused by macro-shifts (“omission” in this case). Table 5.2 Semantic shifts of modality revealed by the translation of CMMs

Semantic shifts (SST-C/ZC) Zero semantic shifts (ZSST) Not applicable (NA) Total

F.

%

853 1248 43 2144

39.79 58.21 2.01 100.00

5.1 General Features of Modality Shifts

91

Table 5.3 Modality shifts occurring with or without corresponding forms of the CMMs in the TT Modality shifts with corresponding forms in the TT (SST-C) Modality shifts without corresponding forms in the TT (SST-ZC) Total

F.

%

535 318 853

62.72 37.28 100.00

Breaking modality shifts in the translation of Chinese modality markers into shifts occurring with corresponding forms in the TT of the investigated Chinese modality markers (SST-C) and shifts occurring without corresponding forms in the TT of the investigated Chinese modality markers (SST-ZC), we can see that (Table 5.3) an overwhelming majority of modality shifts (62.72%) occur with corresponding forms in the TT, and only a minority (37.28%) occur with the omission of Chinese modality markers in the translation (LL = 55.815, p < 0.01). Comparison between the data in Tables 4.3 and 5.3 reveals that the contribution of the addition of English modality markers to modality shifts is significantly greater than that of the omission of Chinese modality markers (LL = 26.179; p < 0.01). In addition, comparing the relative frequency of macro-shifts (i.e., those coded as “Z” under the variable CF, including those causing no semantic shifts1) against the total occurrence of English/Chinese modality markers, we can also see that the probability of adding English modality markers is significantly higher than that of omitting Chinese modality markers in GPC interpreting (LL = 165.698; p < 0.01). This means that GPC interpreters prefer to add English modality markers than omit Chinese modality markers in Chinese-English interpreting in terms of macro-shifts. Next, let’s look at the distribution of modality shifts revealed by the translation of Chinese modality markers among the three semantic dimensions (modality type, orientation and modality value). As Table 5.4 reveals, 88.5% of modality shifts in the translation of Chinese modality markers involve shifts of modality value, 50.76% involve shifts of orientation and only 4.92% involve shifts of modality type (for the same reason as discussed in Sect. 4.1.1, the sum of the three types of shifts exceeds the total number of translation cases involving modality shifts). Thus the distribution of modality shifts among the three semantic dimensions in the translation of Chinese modality markers follows the same pattern as that revealed by the use of English modality markers: shifts of modality value are the most common, shifts of orientation fairly common and shifts of modality type the least common. Since there are different types of shifts within each semantic dimension of modality, it is necessary to examine their relative proportion in order to know which specific type of shifts is predominant in each dimension. The distribution of specific shifts within the three dimensions is presented in Tables 5.5, 5.6, 5.7 and 5.8.

1

If only macro-shifts causing semantic shifts are considered, a significant difference will also be found between the frequencies of “addition” and “omission” (LL = 31.129; p < 0.01).

92

5 Modality Shifts in the Translation of Chinese Modality Markers

Table 5.4 Modality shifts in three semantic dimensions in the translation of CMMs

Shifts of modality type (O-I/O-P/…) Shifts of orientation (SOO/SOS) Shifts of modality value (SV+/SV−) Total cases of modality shifts

F.

%

42 433 755 853

4.92 50.76 88.51

Table 5.5 Shifts of modality type in the translation of CMMs

Shifts of modality type

F.

%

Inclination to obligation (I-O) Inclination to probability (I-P) Obligation to inclination (O-I) Obligation to probability (O-P) Probability to inclination (P-I) Probability to obligation (P-O) Total

5 3 22 8 2 2 42

11.90 7.14 52.38 19.05 4.76 4.76 100.00

Table 5.6 Shifts of orientation in the translation of CMMs

Shifts of orientation

F.

%

Objectivization (SOO) Subjectivization (SOS) Total

405 28 433

93.53 6.47 100.00

Table 5.7 Shifts of orientation occurring with or without corresponding forms in the TT of the CMMs

Subjectivization (SOS) F. % With corresponding forms in the TT (SST-C) Without corresponding forms in the TT (SST-ZC) Total

Table 5.8 Shifts of modality value in the translation of CMMs

Objectivization (SOO) F. %

28

100.00%

99

24.44%

0

0.00%

306

75.56%

28

100.00%

405

100.00%

Shifts of modality value

F.

%

Weakening (SV−) Strengthening (SV+) Total

149 606 755

19.74 80.26 100.00

Despite the limited frequency of shifts in modality type, Table 5.5 reveals a tendency of “de-obligation” in the translation of Chinese modality markers, as shifts from obligation to inclination (52.38%) and shifts from obligation to probability (19.05%) together account for 71.43% of all shifts in modality type. The LL test

5.1 General Features of Modality Shifts

93

reveals a significant difference (LL = 7.970, p < 0.01) between the frequency of “de-obligation” (30 cases) and the frequency of other shifts in modality type (12 cases). This is consistent with the pattern for shifts of modality type in the use of English modality markers (see Table 4.5). Thus it is safe to conclude that interpreters tend to replace markers of obligation with markers of inclination or probability among all the optional changes between modality types in both the use of English modality markers and the translation of Chinese modality markers, although shifts of modality type are not common in general. Contrary to the predominance of “subjectivization” in the use of English modality markers, Table 5.6 suggests a much clearer tendency towards “objectivization” (93.53%) in the translation of Chinese modality markers than “subjectivization” (6.47%). The difference between the two is found to be highly significant (LL = 392.759; p < 0.01). A close examination of these two specific shifts of orientation (see Table 5.7) reveals that most of the shifts of “objectivization” (306 cases, 75.56%) result from the macro-shift of omitting the Chinese modality marker in the translation. By contrast, all shifts of “subjectivization” (28 cases, 100%) occur with corresponding forms in the TT of the Chinese modality markers. Considering only the cases with corresponding forms of the Chinese modality markers in the TT, there is also a significantly stronger pattern of “objectivization” (99 cases) than “subjectivization” (28 cases) (LL = 42.073; p < 0.01). All these data show that GPC interpreters tend to objectivize the existing Chinese modality markers in Chinese-English translation either through the dropping of the Chinese modality marker or through replacing the Chinese modality marker with a more objective form of English modality marker, probably to the effect that the ST speaker’s subjective position towards the message delivered becomes less explicit. But looking at the translation of Chinese modality markers together with the use of English modality markers (see Tables 4.6, 4.7 in Sect. 4.1.1), we can see that the tendencies of “objectivization” and “subjectivization” offset each other to a high extent, but “subjectivization” (mainly caused by the addition of English modality markers) takes the upper hand in the overall reconstruction of modality. As Table 5.8 reveals, the distributional pattern for shifts of modality value in the translation of Chinese modality markers is also opposite to that in the use of English modality markers, with a clear preference for “strengthening” (80.26%) over “weakening” (19.74%). A close examination of data (see Table 5.9) reveals that while all “weakening” shifts occur with corresponding forms in the TT, up to 51.32% (311 out of 606) of the “strengthening” shifts are caused by the omission of Chinese modality markers in the interpreting. For the rest cases where there is a corresponding English form in the TT for the investigated Chinese modality marker, “strengthening” and “weakening” cases are closer in number (295:149), but “strengthening” still clearly takes an upper hand (LL = 48.914; p < 0.01). This means that Chinese modality markers tend to undergo a “strengthening” shift when they are translated into English in GPC interpreting, either through the omission of the Chinese modality marker in translation, or through the use of higher-value modality markers in the TT. Bringing the data revealed by the translation of

94

5 Modality Shifts in the Translation of Chinese Modality Markers

Table 5.9 Shifts of modality value occurring with or without corresponding forms in the TT of the CMMs With corresponding forms in the TT (SST-C) Without corresponding forms in the TT (SST-ZC) Total

Strengthening (SV+) F. %

Weakening (SV−) F. %

295

48.68

149

100.00

311

51.32

0

0.00

606

100.00

149

10.00

Chinese modality markers together with those revealed by use of English modality markers (see Sect. 4.1.1), we can reach the following conclusion: generally speaking, the shift of “weakening” prevails over “strengthening” in the reconstruction of modality, making the speaker sound more tentative and polite, but it is mainly caused by the frequent addition of English modality markers in interpreting. If only the cases with corresponding forms in the ST/TT of the investigated modality marker are considered, then “strengthening” occurs more often, probably to the effect that the ST speaker sounds more confident and authoritative. Detailed discussion of the interpersonal effects of these major types of modality shifts (“de-obligation”, “objectivization” and “strengthening”) in the translation of Chinese modality markers will be offered in Sect. 5.2. It is interesting to note that the overall tendency towards “subjectivization” in the reconstruction of modality in Chinese-English GPC interpreting (most noticeably reflected in the addition of I/we think) is in contrast with the overall trend of “de-personalisation” in simultaneous conference interpreting between French, English and Italian found by Monacelli (2005), although her analysis of shifts in mood and modality revealed [-direct] as the preferred moves of interpreters, similar in nature to the interpreters’ preference towards “weakening” found in the present study. The difference suggests that the sociocultural context of interpreting does play a significant role in shaping interpreters’ choices, which will be the focus of discussion in Chap. 6.

5.1.2

Modality Shifts with Different ST Speakers

As mentioned in Sect. 4.1.2, the overwhelming majority of the ST speech segments are uttered by Chinese officials, so it is reasonable to neglect the possibility of other people as ST speakers for general discussions of modality shifts. However, to explore whether the identity of ST speaker affects the degree or major types of modality shifts in the translation of Chinese modality markers in Chinese-English GPC interpreting, we need to compare the frequency of modality shifts occurring

5.1 General Features of Modality Shifts

95

Table 5.10 Modality shifts in the translation of CMMs occurring with different ST speakers Premier/Foreign minister (PMT) Modality shifts (SST-C/ZC) Occurrence of CMMs

Journalist (JNL)

775

74

1909

214

LL

Sig.

1.818

0.178

Table 5.11 Specific types of modality shifts in the translation of CMMs occurring with different ST speakers Premier/foreign minister (PMT) Shifts of modality type in total Shifts of De-obligation modality type in (O-I/P) detail Other Shifts of modality value in total Shifts of Strengthening modality value (SV +) in detail Weakening (SV−) Shifts of orientation in total Shifts of Objectivization orientation in (SOO) detail Subjectivization (SOS) Occurrence of CMMs

Journalist (JNL)

LL

Sig.

37 26

5 4

−0.147 −0.321

0.702 0.571

11 687 557

1 65 46

0.043 1.790 4.328

0.836 0.181 0.037

130

18

−0.669

0.414

390 366

41 37

0.156 0.369

0.693 0.543

24

4

−0.490

0.484

1909

214

with the two major groups of ST speakers, i.e., Chinese premiers/foreign ministers (PMT) and journalists (JNL). Similar to the findings in the comparison of modality shifts in the use of English modality markers, the LL tests (see Tables 5.10, 5.11) reveal no significant difference in the overall degree of modality shifts in the translation of Chinese modality markers occurring with different ST speakers, but the shift of “strengthening” does occur with a significantly higher probability in translating speeches of premiers/ministers than journalists (LL = 4.328; p < 0.05). As for the distribution of modality shifts within the three semantic dimensions, the same preference for “de-obligation”, “objectivization” and “strengthening” is shared by the translation cases involving the speeches of Chinese officials and journalists. Thus we can conclude that the identity of ST speaker does not affect the general degree or predominant types of modality shifts in the reconstruction of modality in Chinese-English GPC interpreting, but it does affect the degree of one specific type of semantic shift, i.e., the shift of “strengthening”, which features more prominently in translating speeches of premiers/ministers than those of journalists. This lends

96

5 Modality Shifts in the Translation of Chinese Modality Markers

partial support to previous suggestions (Wang 2009, 2012; Liu 2010) that GPC interpreters tend to discriminate journalists from government leaders in making interpreting choices. To sum up, the analysis of concordance lines with Chinese modality markers in the ST reveals the following overall patterns of modality shifts in Chinese-English GPC interpreting (Research Question 1a): (1) Comparatively speaking, semantic shifts of modality occur less frequently than zero shifts, which reveals the preference for “equivalent translation” in the reconstruction of modality in Chinese-English GPC interpreting; (2) With the availability of equivalent English expressions for these Chinese modality markers considered, the degree of modality shifts (853 out of 2144) is still pretty high; (3) Among the three dimensions of semantic shifts, shifts of modality value are the most common, shifts of orientation are fairly common and shifts of modality type are the least common. (4) The distribution within the three types of semantic shifts demonstrates clear preferences for “de-obligation”, “objectivization” and “strengthening”. As for Research Question 1b, the statistical analysis shows that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the translation of Chinese modality markers in GPC interpreting but does affect the degree of one specific type of semantic shift, i.e., the shift of “strengthening”, which occurs more often in translating speeches by premiers/ministers than journalists. Before explaining why these patterns emerge in the translating Chinese modality markers in GPC interpreting, we also need to take a close look at the translation cases involving modality shifts and analyze their interpersonal effects in the interpreter-mediated communication, which will be the focus of the next section.

5.2

Interpersonal Effects of the Major Types of Modality Shifts

So far my analysis of modality shifts based on the ST-TT comparison involving the translation of 15 Chinese modality markers has revealed the overall degree and major types of modality shifts. Different from what is revealed by the use of English modality markers, modality shifts in the three semantic dimensions in the translation of Chinese modality markers turn out to be dominated by “de-obligation”, “objectivization” and “strengthening”. Now I will examine how the three major types of modality shifts are distributed in the translation of these 15 modality markers and then focus on the major contributors for the analysis of interpersonal effects (Research Question 1c). As in Sect. 4.2, they will also be discussed in the order of their frequency: “strengthening” (F = 606), “objectivization” (F = 405), and “de-obligation” (F = 30). As Table 5.12 reveals, “strengthening” mostly occurs in the translation of the central modals 要 (19.97%), 可以 (15.18%), 能/能够 (16.50%) and the mental verb structure 我/我们想 (16.01%); “objectivization” also mainly occurs in the

5.2 Interpersonal Effects of the Major Types of Modality Shifts

97

Table 5.12 Occurrence of major types of modality shifts with each CMM

要 会 能/能够 可以 应该 愿意 可能 一定 必须 需要 我/我们想 我/我们认为 我/我们相信 我/我们希望 将 Total

Strengthening (SV +) F. %

Objectivization (SOO) F. %

De-obligation (O-I/P) F. %

121 13 100 92 46 21 23 1 2 8 97 58 7 3 14 606

80 15 43 65 35 10 11 3 5 9 77 23 8 6 15 405

5 0 0 5 10 0 0 0 2 7 1 0 0 0 0 30

19.97 2.15 16.50 15.18 7.59 3.47 3.80 0.17 0.33 1.32 16.01 9.57 1.16 0.50 2.31 100.00

19.75 3.70 10.62 16.05 8.64 2.47 2.72 0.74 1.23 2.22 19.01 5.68 1.98 1.48 3.70 100.00

16.67 0.00 0.00 16.67 33.33 0.00 0.00 0.00 6.67 23.33 3.33 0.00 0.00 0.00 0.00 100.00

translation of the central modals 要 (19.75%), 可以 (16.05%), 能/能够 (10.62%) and the mental verb structure 我/我们想 (19.01%); “de-obligation” mainly occurs in the translation of the central modals 应该 (33.33%), 要 (16.67%), 可以 (16.67%) and the semi-modal 需要 (23.33%). Now the translation cases involving these major contributors will be examined closely to explore the interpersonal effects of these modality shifts.

5.2.1

Strengthening and Objectivization

As shown in Table 5.12, the major contributors of “strengthening” and “objectivization” are the same four modality markers (要, 能/能够, 可以, and 我/我们想), therefore I will discuss these two types of shifts together through case studies of the four modality markers to explore their interpersonal effects. The relative frequency of “strengthening” shifts with different ST speakers (PMT vs. JNL) occurring in the translation of each modality marker will also be discussed. As the predominant types of semantic shifts in terms of modality value and orientation, “strengthening” (606 cases) and “objectivization” (405 cases) both occur frequently in the translation of Chinese modality markers. Different from “de-obligational” shifts which all occur with corresponding forms of the Chinese modality markers in the TT, a considerable proportion of “strengthening” and “objectivization” shifts occur with

98

5 Modality Shifts in the Translation of Chinese Modality Markers

the omission of the Chinese modality markers in interpreting, which will be illustrated through the following cases. 要 The central modal 要 is the most frequently used modality marker (548 valid cases) in GPC texts, probably because of its polysemy. According to the Modern Chinese Dictionary (the 6th Edition, Institute of Linguistics, Chinese Academy of Social Sciences 2012: 1515), the Chinese auxiliary 要 can mean intend/want to (做某件事 的意志), should (应该), will/be going to (将要), etc. In the GPC corpus data, it is also found to mean need to (需要). Based on the corpus data, I consider the modal 要 as an obligation/inclination marker with a medium-high or medium value equivalent to need to, should, will, be going to or want to depending on the context. As revealed by Table 5.12, the central modal 要 contributes to 19.97% (121 cases) of “strengthening” shifts and 19.75% (80 cases) of “objectivization” shifts in the translation of Chinese modality markers. Out of the 121 cases of “strengthening” shifts involving the translation of the modal 要, 115 cases occur in the translation of premiers/foreign ministers’ speeches and only three cases occur in the translation of journalists’ speeches. Comparing these frequencies against the relative occurrence of Chinese modality markers with different speakers (F1 = 1909, F2 = 214), we can see that the probability of “strengthening” shifts in the translation of 要 is significantly affected by the identity of ST speaker (LL = 10.250; p < 0.01). The translation cases of 要 involving these two types of shifts can be categorized into three groups: translation cases witnessing the co-occurrence of “strengthening” and “objectivization” shifts; translation cases witnessing only the shift of “strengthening”; translation cases witnessing only the shift of “objectivization”. First, the majority of translation cases (F = 65) witness the co-occurrence of “strengthening” and “objectivization” shifts typically caused by the omission of 要 in translating speeches of Chinese officials. With these modality shifts, the speech function of “statement” remains unchanged, but the speaker sounds more assertive or direct in conveying the message. For example: (27) ST: 去年在经济下行压力加大的情况下, 我们果断决定啊全面审计政府性债务, 这本身表明要勇于面对。 [LT: Under the circumstance of the increasingly downward pressure on the Chinese economy last year, we decided to conduct a comprehensive audit of government debt, which shows that the Chinese government will face up to this challenge.] TT: In spite of this downward pressure on the Chinese economy, uh, last year, we con, conducted a comprehensive audit of government debt. That shows that the Chinese government has faced up to this challenge.

Example 27 involves the translation of a speech segment by Premier Li Keqiang in 2014 addressing China’s financial and debt risks. After acknowledging the international community’s worries about the downward pressure on China’s economic growth, Premier Li here states what the Chinese government did in the previous year (2013), leading to the conclusion that “债务风险是总体可控的”

5.2 Interpersonal Effects of the Major Types of Modality Shifts

99

(“the risks are on the whole under control”). He uses the modal 要 to express the government’s intention to face the challenge with courage, implying the government’s confidence in controlling the financial and debt risks. With the modal dropped and the present perfect tense used in the TT, the message about the government facing up to challenges is presented as an objective fact, so the premier (representing the Chinese government) sounds more confident and assertive in the TT. Second, there are 55 translation cases witnessing only the shift of “strengthening” without any shift in orientation, with 要 translated into higher-value modality markers such as must or have to. The speech function of “statement”, “offer” or “suggestion” remains unchanged, but the speaker sounds more assertive, sincere or forceful in the TT than in its literal translation. Examples 15 and 16 in Sect. 4.2.2 are two cases in point, so details will not be repeated here. Third, 15 cases witness only the shift of “objectivization” without any shift in modality value, with the obligation marker 要 typically rendered into a more objective marker such as it is important to in translating Chinese premiers/foreign ministers’ speeches. With these “objectivization” shifts, the speech function of “statement” or “suggestion” remains unchanged, but the subjectivity of the speaker as the source of the obligational pressure becomes less explicit. For example: (28) ST: 关于解决大学生就业和农民工就业的问题, 我们都已经制定了具体的政策, 要狠抓落实。 [LT: As for the solution to the unemployment issue for college graduates and migrant workers, we have already formulated very specific policies and we should work hard to ensure their implementation.] TT: We have already formulated very specific policies in addressing the unemployment issue for migrant rural workers and college graduates. What is now important is to ensure their implementation.

The ST of Example 28 is a speech segment of Premier Wen Jiabao in 2009 addressing the unemployment issue for migrant rural workers as well as university graduates. After admitting that unemployment is a very serious problem facing the Chinese government at the moment, Premier Wen here is reassuring the media that the government has formulated specific policies to ease this problem and emphasizes the government’s obligation to fully implement these policies with the modal 要 (closely equivalent to should). The modalized clause 要狠抓落实 can be considered either as an obligational statement of the government represented by the speaker, or as a personal suggestion to the government from the speaker. Either way, the speaker’s subjectivity as the source of the obligational pressure becomes more implicit with the translation of 要 into the objective marker what is important is to. 能/能够 As discussed in Sect. 3.1.2, the rare cases of 能/能够 to express “pure ability” have been discarded in my data. So the valid concordance lines (194) of 能/能 include its use as a low-value marker of “probability”, “inclination” or “obligation”, equivalent

100

5 Modality Shifts in the Translation of Chinese Modality Markers

to can or be able to in English. As Table 5.12 shows, the central modal 能/能够 contributes to 16.5% (100 cases) of “strengthening” shifts and 10.62% (43 cases) of “objectivization” shifts in the translation of Chinese modality markers. Out of the 100 cases of “strengthening” shifts in translating 能/能够, 86 occur in the translation of premiers/foreign ministers’ speeches, and 11 occur in the translation of journalists’ speeches. Comparing these frequencies against the relative frequency of Chinese modality markers used by different speakers (F1 = 1909, F2 = 214), we can see that the probability of “strengthening” shifts in the translation of 能/能够 is not affected by the identity of ST speaker (LL = −0.164; p = 0.685). In other words, the translation of 能/能够 contributes evenly to “strengthening” shifts in rendering speeches of Chinese officials and journalists. Because most of “objectivization” shifts with 能/能够 co-occur with “strengthening” shifts, the translation cases of 能/能够 involving these two types of shifts can be classified into two groups: translation cases witnessing only the shift of “strengthening” and translation cases witnessing the co-occurrence of “strengthening” and “objectivization” shifts. First, out of the 100 cases of “strengthening” shifts occurring in the translation of 能/能够, 58 cases occur alone (without any change in orientation), typically caused by the replacement of 能/能够 with the higher-value modal will. While translating Chinese premiers/foreign ministers’ speeches, “strengthening” shifts involving 能/ 能够 typically occur with the structure (我们)希望…能/能够… to convey a suggestion in the ST. As Example 10 in Sect. 4.2.2 shows, with the shift of “strengthening”, the basic speech function of “suggestion” in the relevant clause/sentence remains unchanged, but the speaker representing the Chinese government sounds more powerful. While translating journalists’ speeches, “strengthening” shifts involving 能/能够 typically occur when it is used in a question to request the addressee’s opinion, as illustrated by Example 13 in Sect. 4.2.2. With the shift of “strengthening”, the basic speech function of “question” remains unchanged, but a more confident and powerful position of the addressee (representing the Chinese government) is presupposed in the TT than in its literal translation. Second, there are 42 cases of “strengthening” shifts co-occurring with “objectivization”, typically caused by the dropping of 能/能够 in translation. The cases involving Chinese officials’ speeches typically witness 能/能够 used in stating the goal of a desired or intended action. With the modal dropped in the translation, the speech function of “statement” remains unchanged, but the premier/foreign minister’s statement sounds more assertive and concise, and the speaker’s subjectivity is less visible (see Example 29). The cases involving journalists’ speeches typically witness the use of 能/能够 as a politeness marker in raising a question. With the modal dropped in the translation, the speech function of “question” or “request” remains unchanged, but the journalist’s message sounds more direct and concise in the TT (Example 30). In both cases, the efficiency of communication between the speaker and audience is improved with the message conveyed more explicitly and concisely.

5.2 Interpersonal Effects of the Major Types of Modality Shifts

101

(29) ST: 但是我也相信, 不管怎么样, 历史是不能隐瞒的, 我们将继续调查这件事情, 使它能够水落石出。 [LT: But I also believe that history cannot be covered up no matter what. We will continue to investigate this matter so that the truth can come to the light.] TT: Having said that, I want to emphasize that I do believe no one is able to cover up historical truth, so investigation will continue till we really get the full picture. (30) ST: 这个, 这是澳门哪很多人都关心的一个热点问题, 想能不能请总理啊谈一点 看法和意见? [LT: This is a hot issue that many people in Macao are concerned about. So I wonder if I can ask Mr. Premier to share your views and suggestions.] TT: Because this is an issue of great interest to residents in Macao, therefore I raise this to seek your view.

Example 29 concerns the speech segment of Premier Zhu Rongji in 2001 in response to a foreign journalist’s question on the explosion incident in a rural primary school in Jiangxi Province. After acknowledging the skepticism in the media about the cause of the incident and apologizing for having not been able to avoid the incident, Premier Zhu here promises to continue the investigation so as to get the full picture. As a common hedging expression in Chinese, the modal 能够 is used in the ST in stating the desired goal of uncovering the real cause of the incident. With the modal dropped, the TT statement sounds more concise and assertive than the literal translation. With the modalized statement about the goal changed into an objective polar statement, the premier’s subjective assessment of the probability of realizing the goal is also less visible in the TT clause “till we really get the full picture”, although the speaker’s subjectivity in the whole sentence is conveyed with the use of other modality markers. Example 30 involves the speech segment of a journalist in 2001 during Premier Zhu Rongji’s press conference following two related questions, one on the possible changes of the economic role of Macao with the entry into WTO (World Trade Organization) of both the mainland and Taiwan, the other on the prospect of development in Macao. With an account of the Macao people’s interest as the justification, the journalist here is politely requesting the premier to share his view on the two questions with the use of the modal structure 能不能. Since this request is already indicated in the previous two questions, the literal translation would sound redundant. With the modal structure dropped in the translation, the TT is transformed into an explicit expression of the journalist’s request and justification, so the TT sounds more concise and direct, though less polite than its literal translation. 可以 With a total frequency of 197, 可以 is used in the GPC texts as a low-value modal marking obligation, inclination or probability, equivalent to can, may, or be able to in English. As revealed by Table 5.12, the central modal 可以 contributes to 15.18% (92 cases) of “strengthening” shifts and 16.05% (65 cases) of

102

5 Modality Shifts in the Translation of Chinese Modality Markers

“objectivization” shifts in the translation of Chinese modality markers. Among the 92 cases of “strengthening” shifts in translating 可以, 81 occur in the translation of premiers/foreign ministers’ speeches, and 10 occur in the translation of journalists’ speeches. The LL test shows that the probability of “strengthening” shifts in the translation of 可以 is not affected by the identity of ST speaker (LL = −0.081; p = 0.776), either. Like in the case of 能/能够, most of “objectivization” shifts in the translation of 可以 co-occur with “strengthening” shifts, so I will discuss the translation cases of 可以 involving these two types of shifts in two groups: translation cases witnessing the co-occurrence of “strengthening” and “objectivization” shifts and translation cases witnessing “strengthening” shifts alone. First, the majority of cases (60) witness the co-occurrence of “strengthening” and “objectivization” shifts, most of which are caused by the omission of 可以 in translation. In the translation cases of both Chinese officials’ and journalists’ speeches, 可以 (often in the phrase 可以说) is used typically to make a hedged “statement”. With 可以 dropped in the translation, the speech function of “statement” remains unchanged, but the speaker sounds more assertive and confident in the TT. For example: (31) ST: 中国在叙利亚问题上的立场可以概括为四点… [LT: The position of China on the Syrian issue can be summarized as four points…] TT: The position of the Chinese government on the Syrian issue is fourfold…

(32) ST: 最近两岸的问题可以说举世瞩目。 [LT: We can say the questions concerning the two sides of the Taiwan Straits have attracted worldwide attention recently.] TT: And recently the questions concerning the two parts of the Taiwan Straits have attracted worldwide attention.

The ST in Example 31 is a speech segment of Premier Wen Jiabao addressing the Syrian issue in 2012. In response to the journalist’s question “what exactly is China’s position on the issue of Syria”, Premier Wen uses the modal 可以 to make a hedged statement about the position of the Chinese government. Against the background of some foreign media’s criticism that China is biased for Syria because of its friendly relations with Arab countries, the premier actually means to make a clear statement of the Chinese government’s stance on the Syrian issue. So the modal is used as a cultural habit instead of showing uncertainty. By dropping the modal 可以, the interpreter renders the ST into a stronger and more objective statement, so that the premier sounds more assertive and explicit about the Chinese government’s stance towards Syria in the TT than in the literal translation. Example 32 presents a speech segment of a journalist in 2000 seeking Premier Zhu Rongji’s comment on the various interpretations in the media of the white paper on the One-China Principle and the Taiwan Question. Before raising the question, she makes the above statement to offer the background for the question.

5.2 Interpersonal Effects of the Major Types of Modality Shifts

103

The modal phrase 可以说 is used as a hedge to make her ST statement about the publicity of the Taiwan question sound tentative and polite. With the dropping of 可以说 in the translation, the TT is transformed into an objective fact, so the journalist sounds more assertive and direct in the TT than in the literal translation. Second, there are 31 translation cases witnessing “strengthening” shifts alone (no shift in orientation), with 可以 rendered into higher-value modality markers such as will, would like to and need to. In translating Chinese officials’ speeches, “strengthening” shifts involving 可以 typically occur with the use of 可以 to make a statement (see Example 11 in Sect. 4.2.2) or an offer (see Example 33) in the ST. With the shift of “strengthening”, the basic speech function of “statement” or “offer” hardly changes, but the speaker representing the Chinese government sounds more confident or sincere. While translating journalists’ speeches, “strengthening” shifts involving 可以 typically occur when it is used in a question to request the addressee’s opinion (see Example 34). With the shift of “strengthening”, the basic speech function of “question” remains unchanged, but a more confident and powerful position of the addressee (the premier or foreign minister) is presupposed in the TT than in its literal translation (similar to the case of 能/能够). (33) ST: 我可以告诉大家, 中央高度重视, 啊, 立即责成有关部门进行专门调查。目前 调查已经取得进展。 [LT: I can tell you that the central government takes this matter very seriously. We immediately asked the relevant authorities to investigate this incident, and so far progress has been made in the investigation.] TT: I would like to tell you that the central authorities have taken this matter very seriously. And after this incident took place, the central authorities immediately gave the instruction that the relevant authorities should start an investigation into this matter and so far progress has been made in the investigation. (34) ST: 还有就是说, 现在一些外资银行已经可以经营人民币的业务了, 那么目前您 认为大概最快什么时候可以全面开放人民币业务的经营? [LT: And now that some foreign banks are already allowed to conduct RMB business, how soon do you think China can fully open its RMB business to foreign banks?] TT: And my second question is some foreign banks are now already allowed to be engaged in the local currency, the RMB business, how soon will China open totally its RMB business to foreign banks?

Example 33 concerns the speech segment of Premier Wen Jiabao in 2012 addressing the “Wang Lijun incident”. After acknowledging the extensive public concern for this incident, Premier Wen here uses the modal 可以 to offer to inform the media of the attitude and actions of the central government concerning this incident. As a low-value inclination modal, 可以 is this context is closely equivalent to can in English. But the interpreter renders it into a higher-value modality marker “would like to”, to the effect that the premier sounds more earnest and ready in making the offer, and also sounds more confident about the following information he is revealing.

104

5 Modality Shifts in the Translation of Chinese Modality Markers

Example 34 concerns the speech segment of a journalist in 1999 asking Premier Zhu Rongji about the policy towards foreign banks conducting RMB business. The modal 可以 is used in the ST question to seek Premier Zhu’s opinion on the probability of fully opening the RMB business to foreign banks. With 可以 translated into a higher-value modal will, a more confident and powerful position of the premier towards the addressed issue is presupposed in the TT question. 我/我们想 As the most frequently used subjective modality marker in the corpus, 我/我们想 (followed by a clause) appears 158 times, including 157 cases of 我想 and 1 case of 我们想. Equivalent to I/we think, 我/我们想 is used typically in the GPC texts as medium-value subjective markers of probability. As revealed by Table 5.12, the structure 我/我们想 contributes to 16.01% (97 cases) of “strengthening” shifts and 19.01% (77 cases) of “objectivization” shifts in the translation of Chinese modality markers. Note that almost all of the translation cases of 我/我们想 involving “strengthening” shifts occur in translating Chinese officials’ speeches (95 out of 97), with only one case in translating a journalist’s speech. As the LL test reveals (LL = 13.659; p < 0.01), the probability of “strengthening” shifts in the translation of 我/我们想 is significantly affected by the identity of the ST speaker. Like in the case of 能/能够 and 可以, most of the “objectivization” shifts in the translation of 我/我们想 co-occur with “strengthening” shifts, so I will discuss the translation cases of 我/我们想 involving these two types of shifts in two groups: translation cases witnessing the co-occurrence of “strengthening” and “objectivization” shifts and translation cases witnessing the “strengthening” shifts alone. First, the majority of cases (F = 75) witness the co-occurrence of “strengthening” and “objectivization” shifts, most of which are caused by the omission of 我/ 我们想 (in fact, only 我想 is involved here) in translation. Here 我想 is typically used in the Chinese premiers/foreign ministers’ speeches to state a subjective opinion. With these modality shifts, the speech function of “statement” remains unchanged, but the Chinese officials’ subjectivity is less visible and the statement sounds more assertive in the TT. For example: (35) ST: 呃,我们啊, 国家的发展我想为我们同其他国家的, 这个, 呃,合作创造了更大 的空间。 [LT: Well, the development of China, I think, has created broader space for our cooperation with other countries.] TT: The development of China has offered broad space for China’s cooperation with other countries.

Example 35 concerns the speech segment of Foreign Minister Yang Jiechi in 2011 on the principles guiding China’s relations with other countries. After emphasizing China’s adherence to the Five Principles of Peaceful Coexistence, Mr. Yang here is talking about the contribution of China’s development to other countries (especially African countries) to illustrate the “mutually beneficial cooperation”. 我想 is used as a hedging device in the ST to make a humble

5.2 Interpersonal Effects of the Major Types of Modality Shifts

105

statement of China’s contribution. With the omission of this subjective modality marker in the translation, the TT statement sounds more objective and assertive and the foreign minister sounds more confident to the English-speaking audience. Second, another considerable number of cases (F = 21) witness the shift of “strengthening” alone, with 我想 typically rendered into the higher-value modality marker I believe. Similar to the first group, 我想 is also used in the Chinese premiers/foreign ministers’ speeches to state a subjective opinion. With the “strengthening” shift, the speech function of “statement” remains unchanged, but the Chinese official sounds more confident with his statement in the TT. A case in point is Example 17 presented in Sect. 4.2.2.

5.2.2

De-obligation

应该 With a total frequency of 144, 应该 is used in GPC texts as a medium-value modal to express mainly obligation and occasionally probability, and it is closely equivalent to should in English. As Table 5.12 shows, 10 translation cases of 应该 undergo “de-obligation”, accounting for 33.33% of all “de-obligation” shifts in the translation of Chinese modality markers. These shifts mostly involve the translation of the obligation modal 应该 into inclination markers such as will, be going to and want to. First, the “de-obligation” cases involving Chinese premiers/foreign ministers as ST speakers typically witness 应该 rendered into will or want to, through which the “statement” about the Chinese government’s obligation is changed into an “offer” (see Example 19 in Sect. 4.2.3) or the speaker’s “offer” becomes more explicit (see Example 36), so that the TT sounds more responsive, sincere and appealing to the English-speaking audience. (36) ST: (发生这样一个事情, 特别是发生在江泽民主席对于这一种爆炸事件多次的 批示的情况底下, 国务院没有尽到它的责任。 ) 我感到心情沉重, 应该进行检讨。 [LT: I feel sad with a heavy heart, and I should apologize and criticize my own work.] TT: (I feel very sad because this incident occurred despite that President Jiang made lots of important instructions on efforts to prevent those explosions and accidents. So to put it differently, the State Council has not performed its mission properly.) I feel very sad and I carry a very heavy heart. I want to apologize and review and reflect my…reflect on my own work.

The ST of Example 36 is a speech segment of Premier Rongji in response to a foreign journalist’s question on the explosion incident in a rural primary school in Jiangxi Province. Before reporting the results of official investigations, Premier Zhu here expresses his deep sorrow over this event and uses the modal 应该 to state his obligation to apologize and reflect on his own work, which also functions as an

106

5 Modality Shifts in the Translation of Chinese Modality Markers

“offer”, because the statement 应该进行检讨 itself performs the act of apologizing. With the obligation modal 应该 replaced by the inclination marker want to, this offer of making an apology sounds more direct and sincere in the TT. Second, the “de-obligation” cases in the interpretation of journalists’ questions typically witness 应该 rendered into be going to, with the Chinese government or the State Council as the modal subject. The speech function of “question” remains unchanged but the focus of the question is shifted from the obligation of the Chinese government to the probable future action, which more closely reflects the questioner’s intention and the audience’s interest, and also makes the journalist sound more polite in the TT. Example 25 and Example 26 in Sect. 4.2.3 are two cases in point. 要 As shown in Table 5.12, 5 translation cases of 要 undergoes “de-obligation”, accounting for 16.67% of all “de-obligation” shifts in the translation of Chinese modality markers. These shifts involve the translation of the obligation modal 要 (closely equivalent to should or need to) into the inclination/probability marker will or be ready to, typically with Chinese premiers/foreign ministers as ST speakers. With these “de-obligational” shifts, either the speech function of “statement” remains unchanged, or the “statement” of the government’s obligation is transformed into an “offer”, and the TTs sound more comprehensible and appealing to the English-speaking audience. (37) ST: 要让政府有更多的精力来完善和创新宏观调控, 尤其是加强事中事后的监 管。 [LT: We should ask the government to exert more energy to improve and reform macro-control, especially to tighten its oversight over delegated matters in process and afterwards.] TT: The government, in this way, will have extra energy to focus its attention on pursuing a creative and better macro-control. And also in particular, the oversight over delegated matters will be tightened when these matters are being handled or after they have been handled. (38) ST: 我们要更加善于应对挑战, 抓住机遇。我们要化挑战为机遇, 变机遇为现 实。 [LT: We need to be better at meeting challenges and seizing opportunities. We need to turn challenges into opportunities, and opportunities into realities.] TT: We are ready to meet challenges, seize opportunities and translate challenges into opportunities, and opportunities into a reality.

The ST in Example 37 is a speech segment of Premier Li Keqiang in 2014 on “简政放权” (“streamlining administration and delegating power”). Following the clarification that “我们讲的是放管结合” (“What we want is both the power delegation and tightened oversight on part of the government.”), Premier Li here uses the obligation modal 要 to state “our” obligation to allow the central government to

5.2 Interpersonal Effects of the Major Types of Modality Shifts

107

focus more on improving macro-control and oversight instead of reviewing everything that can be done by lower-level governments. It is easy for the Chinese native-speakers to understand that the intention of the speaker here is not really to stress anyone’s obligation to do anything, but to explain the reason and the essence of “streamlining administration and delegating power”. However, the literal translation would have been confusing to the English-speaking audience as the carrier of obligation is not clear. With the obligation modal 要 rendered into the probability modal will and the modal subject clarified, the TT states the government’s motivation for delegating power more clearly. Example 38 concerns Foreign Minister Yang Jiechi’s speech segment in 2008 addressing the opportunities and challenges for China’s diplomacy in the next five years. After admitting that “the international situation will become more complex in the next 5 years”, Mr. Yang here uses the modal 要 to state the government’s internal obligation to be better at handling challenges and seizing opportunities. With the obligation modal 要 replaced by the inclination marker be ready to, the statement becomes an offer with more performative force, so the TT sounds more relevant and appealing to the English-speaking audience than the literal translation. 可以 Similar to the modal 要, 5 translation cases of 可以 undergoes “de-obligation”, accounting for 16.67% of all “de-obligation” shifts in the translation of Chinese modality markers. These shifts typically involve the replacement of the obligation modal 可以 (equivalent to can/may) with the probability marker will or the inclination marker would like to in translating Chinese officials’ speeches. With these “de-obligational” shifts, the speech function of “statement” or “suggestion” remains unchanged, but the speaker or the Chinese government sounded more proactive in the TTs. (39) ST: (北京市已经有了这样的打算, 将亚运会的这些设施将来将有效的利用起 来。) 譬如说有些运动员的住宅可以作为公寓给缺房子的人住… [LT: For example, some of the residential places for athletes can be used as the apartment buildings to those people in need of a house.] TT: (The municipal government of Beijing has this intention, that is, after the Asian Game the facilities of this Asian Game will be made full use of.) Some of the residence places for athletes and sportsmen this time will be given… made …will be used as the apartment buildings to those people in need of house. (40) ST: (我们国务委员兼北京市市长陈希同同志已经在去年人民, 人大常委会的会 议上就六四事件做了非常详尽的报告。) 如果诸位有兴趣的话还可以去翻一翻。 [LT: If any of you are interested, you can go ahead to read it.] TT: (I think about the June 4th Event, Comrade Chen Xitong who’s the State counselor and Mayor of Beijing gave a very detailed report and account about the June 4th Event in his report released last year at a standing committee meeting of the National People’s Congress.) So if anyone is interested in this matter, I would like to refer you to that report.

108

5 Modality Shifts in the Translation of Chinese Modality Markers

Example 39 concerns Premier Li Peng’s speech segment in 1989 addressing the cost and benefit of hosting the Asian Games in 1992. In response to some people’s opinion that this hosting of the Asian Games would be an unnecessary waste of money for China, Premier Li here is talking about the Beijing municipal government’s plan to make good use of the facilities of this Asian Games afterwards. With the low-value obligation modal 可以 (can/may), he states the municipal government’s possible permission for the needy people to use the residential places for athletes during the Games as an example of not wasting the facilities. The interpreter opts to render 可以 into the probability modal will, so the focus of the TT becomes a strong prediction about the future use of these facilities instead of the possible permission. With this shift, the Chinese government sounds more proactive and the Beijing municipal government’s “intention” sounds more convincing to the English-speaking audience. Example 40 concerns Premier Li Peng’s speech segment in 1990 in response to a foreign journalist’s question about some details of the June 4th Event. Here Premier Li uses the obligation modal 可以 (closely equivalent to can) to realize the indirect act of “suggestion” for the journalist and any other people interested to read the report of Comrade Chen Xitong (Mayor of Beijing) on this event released last year, which covers all the details. As the original question from the journalist is very provocative and face-threatening (“So Mr. Li Peng, do you or do you not accept responsibility for the troops firing at the people? Was there any order for the troops to fire at anyone that night and who gave it?”), Premier Li chooses not to give a direct answer here, but gives the important reference where the journalist could find all relevant information. With the obligational modal 可以 rendered into the inclination marker would like to and the modal subject shifted from the implicit you to the explicit I, the premier sounds more proactive and responsible in giving the suggestion. 需要 With a total occurrence of 37, the semi-modal 需要 (followed by a verb) is used in the GPC texts as an obligation marker equivalent to need to in English. As Table 5.12 reveals, 7 translation cases of 需要 undergoes “de-obligation”, accounting for 23.33% of all “de-obligation” shifts in the translation of Chinese modality markers. These shifts mostly involve the translation of the obligation modal 需要 into inclination markers such as will and would (including the phrase “would like to”) in Chinese premiers/foreign ministers’ speeches. With these “O-I” shifts, either the “statement” of the government’s obligation is turned into an “offer”, or the “offer” is realized in a more explicit way to the effect that the Chinese government sounds more proactive and the TT sounds more appealing to the English-speaking audience. Example 20 and Example 23 discussed in Sect. 4.2.3 are two cases in point, so illustrations will not be repeated here.

5.3 Summary

5.3

109

Summary

In this chapter, I have discussed the general patterns of modality shifts in the translation of high-frequency Chinese modality markers in Chinese-English GPC interpreting and also explored the interpersonal effects of the major types of modality shifts through case studies of the major contributors. Analysis of the semantic shifts of modality through the translation of Chinese modality markers reveals that modality shifts occur less frequently than zero shifts in Chinese-English GPC interpreting, but the degree of modality shifts is still fairly high (39.8%), most of which occur with corresponding forms in the TT. The distribution of modality shifts in the three dimensions follows the pattern that shifts of modality value are the most common, shifts of orientation fairly common and shifts of modality type the least common. Within the three dimensions of semantic shifts, “de-obligation”, “objectivization” and “strengthening” shifts are predominant respectively. Comparative analysis of modality shifts in translation cases with different ST speakers shows that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the translation of Chinese modality markers in GPC interpreting but does affect the degree of the “strengthening” shift, which features more prominently in translating speeches of Chinese officials than journalists (mainly in the translation of 要 and 我/我们想). Bringing the findings from Chaps. 4, 5 together, we can see different patterns as well as shared features in terms of modality shifts from the two directions. Shared features include: (1) The general degree of modality shifts is similar in two directions, both significantly lower than that of zero-shifts, but the degree of modality shifts (around 40% in both directions) is still high considering the availability of equivalent Chinese/English modality markers. (2) The distribution of modality shifts in the three dimensions follows the same pattern that shifts of modality value are the most common, shifts of orientation fairly common and shifts of modality type the least common. (3) “De-obligation” is predominant in terms of shifts in modality type in both directions. (4) The identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the reconstruction of modality but does affect the degree of the “strengthening” shift, which features more prominently in translating speeches of Chinese officials than journalists (mainly reflected in the use of must and I/we believe, and the translation of 要 and 我/我们想). Different features include: (1) The contribution of the omission of Chinese modality markers to modality shifts in the translation of Chinese modality markers is not so prominent as the role of the addition of English modality markers in causing modality shifts in the use of English modality markers, and the relative frequency of omitting Chinese modality markers is also significantly lower than adding English modality markers against the total occurrence of Chinese/ English modality markers. In other words, the GPC interpreters prefer the “addition” of English modality markers to the “omission” of Chinese modality markers in terms of macro-shifts. (2) Different from the predominance of “subjectivization” and “weakening” shifts in the use of English modality markers, the translation of

110

5 Modality Shifts in the Translation of Chinese Modality Markers

Chinese modality markers reveal the predominance of “objectivization” and “strengthening” shifts. As discussed in Sect. 5.1.1, a close look at the data from the two directions (see Tables 4.7, 4.9, 5.7, 5.9) leads to two findings: first, the tendencies of “objectivization” and “subjectivization” offset each other to a high extent, but “subjectivization” (mainly caused by the addition of English modality markers) takes the upper hand in the reconstruction of modality; second, the tendencies of “weakening” and “strengthening offset each other to a large extent, but the shift of “weakening” (also mainly caused by the addition of English modality markers in interpreting) outnumbers “strengthening” in the overall reconstruction of modality. As discussed in Sect. 5.2, the interpersonal effects of the major types of modality shifts in the translation of Chinese modality markers are revealed through the case studies of their major contributors: (1) With the shifts of “strengthening” and “objectivization” (most of “objectivization” shifts co-occur with “strengthening” shifts) mostly occurring in the translation of 要, 能/能够, 可以, 我/我们想, either the Chinese premiers/foreign ministers are presented as more confident and authoritative, or the journalists sound more direct and concise. Like in the cases of must and I believe discussed in Sect. 4.2.2, the effect of the ST speaker’s identity on the degree of “strengthening” shifts is mainly reflected in the translation of 要 and 我/我们想, probably because these two modality markers are typically used to express the opinion or attitude of the speaker rather than that of the addressee, and the interpreter is not as motivated to strengthen journalists’ opinions/attitudes as to strengthen Chinese officials’. (2) With the “de-obligation” shifts mainly occurring in the translation of 应该, 要, 可以 and 需要, either the Chinese officials sound more sincere, proactive and responsive, or the journalists sound more polite and respectful. Similar to what’s revealed by the use of English modality markers, despite the frequent shifts in the semantics of modality in the translation of Chinese modality markers, the basic interpersonal speech function (statement/question/offer/command) rarely changes. In a word, the interpreters frequently allow modality shifts instead of always obeying the rule of “equivalence” in their reconstruction of modality in Chinese-English GPC interpreting mainly for achieving a better communicative effect. So far, the corpus-based data analysis and case studies in Chaps. 4, 5 have revealed the interpreter’s active role as a mediator between the ST speaker and the TT audience, and the interpersonal effects of the major types of modality shifts have also pointed to the importance of sociocultural factors in interpreting choices, but the systematic relationship between the reconstruction of modality and the sociocultural context of GPC interpreting is yet to be discussed. To achieve a better understanding of the sociocultural factors shaping the interpreting choices in the reconstruction of modality and a deeper understanding of GPC interpreting as a socially situated practice, I will discuss the relationship between the interpreting choices concerning modality and the sociocultural context in the next chapter, with the focus on the institutional norms and the typological features of Chinese-English GPC interpreting.

References

111

References Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese-English): face management and information management. Ph.D. thesis, Heriot-Watt University. Monacelli, Claudia. 2005. Surviving the role: a corpus-based study of self-regulation in simultaneous interpreting as perceived through participation framework and interactional politeness. Ph.D. thesis, Heriot-Watt University. Wang, Binhua王斌华. 2009. 口译规范描写及其应用——一项基于中国总理“两会”记者招待 会交传语料的研究 [Description of norms in interpreting and its application—a study based on the corpus of consecutive interpreting in Chinese premier press conferences]. Ph.D. thesis, Guangdong University of Foreign Studies. Wang, Binhua. 2012. A descriptive study of norms in interpreting: based on the Chinese-English consecutive interpreting corpus of chinese premier press conferences. Meta 57 (1): 198–212.

Chapter 6

The Reconstruction of Modality in Relation to Its Sociocultural Context

As mentioned in Sect. 3.1.3, due to the shared temporal and spatial setting of the ST and TT and the co-presence of primary participants in the GPC interpreting events, I will follow interpreting scholars such as Alexieva (1997), Diriker (2004) and Pöchhacker (2009) and discuss the relevant sociocultural parameters of GPC interpreting as the integrated context of this interpreter-mediated communicative activity rather than discuss the context of the ST and TT separately, which is more applicable to studies of written translation. I will approach the sociocultural context of Chinese-English GPC interpreting from two perspectives: the meta-discursive representation of diplomatic interpreting which indicates the institutional norms of GPC interpreting, and the multi-parameter typological analysis of GPC interpreting. The meta-discourse will be juxtaposed with the regular patterns in reconstructing modality to explore the governing power of institutional norms, and then the typological features will be analyzed as contextual clues to explain the interpreting choices concerning modality.

6.1

Modality Reconstruction in GPC Interpreting in Relation to Its Meta-Discourse

As Diriker (2004, p. 25) rightly points out, the meta-discourse of interpreting is not “just a neutral description of intrinsic professional features”, but “presents a selected and hierarchized set of ‘norms’”, which embody “the values and ideas” shared by a community. Although translational/interpreting norms extracted from these “extratextual” sources may contradict with the actual translational/interpreting behavior, they nevertheless offer “a possible key to the analysis of actual behavior” (Toury 1995, pp. 65–66). Since the GPC interpreting in this study is a typical type of diplomatic interpreting performed by the professional interpreters from the Department of Translation and Interpreting affiliated to the Ministry of Foreign © Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_6

113

114

6 The Reconstruction of Modality in Relation …

Affairs, I will analyze the meta-discourse on diplomatic interpreting from the Ministry of Foreign Affairs, which will reveal the institutional norms of GPC interpreting. Due to this institutional background, the most relevant meta-discourse of GPC interpreting would be written normative documents such as “codes of practice” or “official guidelines” for in-house interpreters at the Department of Translation and Interpreting, Ministry of Foreign Affairs. Unfortunately, these documents are not available to the public for confidential reasons. Thus I will turn to the academic publications by former leaders of the Department of Translation and Interpreting, Ministry of Foreign Affairs and interviews with them to extract the institutional norms of GPC interpreting, which will then be compared against the interpreting choices in reconstructing modality discussed in Chaps. 4 and 5.

6.1.1

The Meta-Discourse on Diplomatic Interpreting in China

6.1.1.1

Published Discourse in the Academia from Institutional Insiders

Although the official normative statements on how to interpret from the Ministry of Foreign Affairs are not accessible to the outsiders, the past 25 years does witness the publication of articles in the academia by former directors or deputy-directors of the Department of Translation and Interpreting, Ministry of Foreign Affairs (Wang 1991, 2002; Xu 2000; Ren 2004; Guo 2004; Zhang 2004; Shi 2007, 2009) on diplomatic translation. Due to these authors’ positions as institutional insiders and leaders, their critical, evaluative, or introspective comments in these articles serve as an important source to extract the institutional norms that are supposed to govern GPC interpreting. Note that interpreting norms in this study refer to the “initial norms” instead of “operational norms” according to Toury’s (1995) classification of translation norms. Although he (ibid., 56) used the two terms “adequacy” and “acceptability” to refer to the initial choice of the translator between adherence to norms of the source language/culture and adherence to norms of the target language/culture, the term “adequacy” is rather confusing because there is a wide-spread use of “adequacy” to mean roughly what Toury means by “acceptability” among other scholars (Chesterman 1996, p. 200). Thus I choose to use “fidelity” instead of “adequacy” to refer to the ST-oriented norm/principle and use “acceptability” to refer to the TT-oriented norm/principle in the context of this study. Note that the two principles are most of the time compatible in reality, as being faithful/accurate or acceptable is a matter of degree, rather than a matter of “yes or no”. The ideal diplomatic interpreting performance in China aims to realize both “fidelity” and “acceptability”, as reflected in Premier Zhou Enlai’s requirement “complete and accurate, smooth and intelligible” (in Shi 2009, p. 10, my translation). Nevertheless, when the two norms are considered as two poles with opposing

6.1 Modality Reconstruction in GPC Interpreting in Relation …

115

orientations, the former leaders of the Department of Translation and Interpreting, Ministry of Foreign Affairs seem to agree to the overriding position of “fidelity” in diplomatic interpreting: Chinese-English [diplomatic] translation is required to achieve “faithfulness, expressiveness and elegance”, but among the three “faithfulness” is the most important. There is no point of talking about “expressiveness and elegance” without achieving “faithfulness” first. (Wang 1991, p. 7, my translation) Our government leaders usually voice their attitudes to the foreign audience with deliberation over the wording; therefore whether the interpreters can render the leaders’ speeches accurately makes a great difference in diplomatic affairs. (Xu 2000, p. 35, my translation) [Diplomatic] interpreting only serves as a bridge for communication. Therefore, the interpreter cannot change or omit words in the source text as s/he likes. S/he cannot abridge the source speech because s/he considers the speaker too wordy. (ibid., 38, my translation) Diplomatic interpreting is politically very sensitive, so accuracy is taken as a top criterion in the training. (Ren 2004, p. 62, my translation) All translation practitioners hope that their products can meet the criteria of “faithfulness, expressiveness and elegance”, and diplomatic translation is no exception. “Faithfulness”, referring to the accurate understanding and re-expressing of the source text, is the most crucial among the three. (Zhang 2004, p. 55, my translation) Diplomatic interpreting is highly sensitive politically, so we must interpret with precision and avoid deviations from the source speech.(Shi 2007, p. 57, my translation)

All the above comments highlight the importance of “faithfulness, completeness and accuracy” in diplomatic interpreting, which points to the primary norm of “fidelity”. Since diplomatic interpreting (especially Chinese-English interpreting) in China has the primary function of publicizing the Chinese government’s policies and political stances to the outside world, this emphasis on “faithful translation” can be regarded as a means to support the “proactive public diplomacy campaign” (Sun 2013, p. 59) of China to “present an accurate picture of itself to the world” (Zhao 2007a). It is noteworthy that these institutional insiders do not identify “faithfulness, completeness and accuracy” with “word-for-word translation”. Instead, they encourage the diplomatic translators/interpreters to be faithful to the intended meaning and spirit of the source text: In order to accurately convey the meaning of the source text, the translator/interpreter sometimes needs to add or omit certain words in translation. (Wang 1991, p. 9, my translation) A noticeable problem is that some [diplomatic] translators/interpreters are used to “playing safe” at work in order to achieve “accuracy” or to avoid making mistakes to be exact. That is, they prefer to pursue the formal equivalence between the target text and the source text, resort to literal translation most of the time instead of making efforts to thoroughly understand the meaning of the source text and pursuing more accurate re-expression. (Zhang 2004, p.55, my translation)

With this understanding of “faithfulness and accuracy”, Ms. Shi Yanhua, the former Director of the Department of Translation and Interpreting, Ministry of

116

6 The Reconstruction of Modality in Relation …

Foreign Affairs, points out the growing attention to “acceptability” in diplomatic translation/interpreting: Recent years witness an orientation that diplomatic interpreters are striving for. That is, taking into consideration the specific settings and audience of the speeches delivered by state and ministerial leaders, we try to render them with languages readily understandable by foreign audience. In other words, we are trying to make the target speech more smooth and intelligible with faithfulness to the original text ensured. (Shi 2009, p. 12, my translation)

To sum up, the published articles by the former directors and deputy-directors of the Department of Translation and Interpreting, Ministry of Foreign Affairs point to the primary norm of “fidelity” (i.e., “faithfulness, completeness and accuracy”) and the secondary norm of “acceptability” (i.e., “fluency, smoothness and intelligibility”) and also reveal the ideal of realizing both “fidelity” and “acceptability” in diplomatic interpreting. According to these institutional insiders, the top principle of “faithfulness, completeness and accuracy” in diplomatic interpreting does not always lead to literal translation. Instead, deviation from formal equivalence is sometimes necessary to ensure the accurate conveyance of the source message.

6.1.1.2

A Personal Interview with an Institutional Insider

Besides collecting the academic publications from the institutional insiders, I also had the honor of interviewing Mr. Chen Mingming1 (Chen 2015), the former Director of the Department of Translation and Interpreting, Ministry of Foreign Affairs over the phone. As an experienced diplomatic translator/interpreter working for the Ministry of Foreign Affairs for decades, Mr. Chen’s remarks will undoubtedly shed light on the institutional norms of GPC interpreting. In fact, my interview covers both questions on the norms/principles of GPC interpreting, and questions related to the typological features of GPC interpreting to be discussed in Sect. 6.2. Here I will only present his remarks on the norms/principles of GPC interpreting. Talking about the standard in GPC interpreting, like all the other institutional insiders mentioned above, Mr. Chen clearly regards “faithfulness, completeness and accuracy” as the overriding principle and places “acceptability” in the second place: The most important principle is to be faithful to the source speech and to render the messages in the source speech completely and accurately. Having achieved this, we should also try our best to present the target speech fluently and make it acceptable and comprehensible to the target audience. For example, when there are unintended repetitions in the source speech, the repeated parts can be dropped in the target speech to make it more presentable. If the literal translation of the source speech leads to logical ambiguity or misunderstanding, we should make adjustments in the interpreting to convey the intended message clearly to the audience. However, generally speaking, the room for target-audience-oriented adjustments is rather limited in GPC interpreting. (Chen 2015, my translation)

1

The interview was conducted on April 27, 2015 over the phone in Chinese. For a full script of the interview, see Appendix D.

6.1 Modality Reconstruction in GPC Interpreting in Relation …

117

Through personal communication, I could tell that there is a balance in Mr. Chen’s mind between “fidelity” and “acceptability”, but the balance is clearly tilted towards the former. On the one hand, Mr. Chen agrees with Zhang (2004) and Shi (2009) that being faithful does not mean rigid word-for-word translation and diplomatic interpreters should strive to make their target speech comprehensible to the target audience. On the other hand, he has strong reservations about the interpreter’s flexibility in adding, omitting or adjusting the strength of certain words in the source text such as modality markers in order to promote friendly communication between the speaker and the audience. For example, he agrees to the view that “being faithful to the source text means adopting literal translation as much as possible on condition that the target text is comprehensible” (Chen 2015, my translation). When talking specifically about the translation of modality markers, he holds that “GPC interpreters are encouraged to render the Chinese modality markers into their equivalent English modal expressions as often as possible” (ibid., my translation). When asked whether interpreters are encouraged to adjust the modal strength in the source text or add modality markers in the target text to reduce the power distance between the speaker and the audience and facilitates their communication, he replies without hesitation, “No, it’s not necessary to do so, because the adjustment of modal strength or the addition of modality markers will result in the target message deviating from the source message” (ibid., my translation). To sum up, my personal interview with Mr. Chen Mingming reveals a similar institutional norm to that indicated from published discourse by institutional insiders: “fidelity” first and “acceptability” second. Besides, Mr. Chen explicitly tells the condition under which interpreters should deviate from literal translation: “only when the literal translation leads to logical ambiguity, misunderstanding or unacceptable English” (ibid., my translation). More specifically, talking about translation concerning modality, he does not think interpreters should be encouraged to change the modal strength (i.e., modality value) or add/omit modality markers in the translation even if these changes could promote a more friendly relationship between the speaker and audience (ibid., my translation).

6.1.2

Juxtaposing the Reconstruction of Modality in GPC Interpreting with the Meta-Discourse

Juxtaposing the regular patterns in the reconstruction of modality in ChineseEnglish GPC interpreting discussed in Chaps. 4 and 5 with the meta-discursive representation of diplomatic interpreting in the Chinese setting discussed in Sect. 6.1.1, we can see a relationship of “partial conformity and partial contradiction”. On the one hand, in a loose sense, the overriding norm of “fidelity” (i.e., “faithfulness, completeness and accuracy”) is reflected in the preference for “zero semantic shifts” to “semantic shifts” and the preference for “additions” to “omissions” of

118

6 The Reconstruction of Modality in Relation …

modality markers in the reconstruction of modality in Chinese-English GPC interpreting. As the corpus-based data analysis (see Tables 4.2 and 5.2) reveals, the frequency of “zero semantic shifts” is significantly higher than “semantic shifts” in the translation of Chinese modality markers as well as the use of English modality markers. That is to say, using an equivalent expression in modality type, modality value and orientation (usually “stock equivalents” noted in Chinese-English dictionaries) to translate the Chinese modality markers into English is the preferred choice in the reconstruction of modality in GPC interpreting, which conforms to the primary principle of “faithfulness, completeness and accuracy” revealed by the meta-discourse. For example, the English modal should is mostly used in the interpreted texts to render its Chinese equivalent modal 应该 or 要 in the source texts, while the Chinese semi-modal 必须 occurring in the source texts is mostly translated into its English equivalent must in the target texts. When semantic modality shifts are caused by the addition or omission of modality markers, “addition” (650 out of 3013) is clearly more common than “omission” (318 out of 2144). This is understandable because the addition of a modality marker in the TT can be considered as a means to faithfully render the intended meaning of the ST speaker, but the omission of an existing modality marker in the ST is more likely to be considered as incomplete rendition. On the other hand, in a strict sense, the institutional norm “fidelity first and acceptability second” is contradicted by the high frequency of modality shifts, which occur as optional choices of the interpreter instead of obligatory shifts demanded by linguistic differences. As mentioned in Chaps. 4 and 5, although not as frequent as “zero shifts”, the frequency of “modality shifts” is still fairly high in Chinese-English GPC interpreting. To be exact, 41% (1236 out of 3013) of all cases of high-frequency English modality markers and 39.8% (853 out of 2144) of all cases of high-frequency Chinese modality markers involve modality shifts (optional). Although the institutional insiders agree that being faithful and accurate does not mean rigid word-for-word translation, my personal interview with Mr. Chen Mingming shows that modality markers as important lexical realizations of interpersonal functions are not among the words that interpreters are encouraged to translate with high flexibility. The former leaders of the Department of Translation and Interpreting, Ministry of Foreign Affairs seem to agree that interpreters are encouraged to make flexible adjustments in diplomatic interpreting if the literal translation of the source speech is not comprehensible to English-speaking audience or cannot convey the source message accurately to the target audience. However, most of the modality shifts in our corpus data occur when the alternative of literal translation could also have communicated the message accurately and intelligibly to the target audience. For instance, Premier Li Keqiang’s polar statement “需要持久的世界和平环境” in Example 3 (see Sect. 4.2.1) could have been rendered more “faithfully” into “that requires an international environment of lasting peace” without the addition of would. Premier Zhu Rongji’s polar statement “而且现在还言之过早” in Example 9 could have been translated more “faithfully” into “it is still too early now to discuss it” without the addition of I think. Premier Zhu Rongji’s speech segment “应该进行检讨” in Example 36 (see Sect. 5.2.2)

6.1 Modality Reconstruction in GPC Interpreting in Relation …

119

could have been rendered more literally into “I should apologize and criticize my own work” instead of the actual “de-obligational” version “I want to apologize and review and reflect my…reflect on my own work”. These examples illustrate the overall tendency towards “weakening”, “subjectivization” and “de-obligation” shifts in the reconstruction of modality in Chinese-English GPC interpreting, which make the speakers sound more friendly, responsive and sincere in the TT. Given the previous suggestions that English political speeches have a more personal touch (Wang 2008; Zhang 2013; Pan 2014) and they use the lower-value modals more frequently and the obligation modals less frequently than Chinese political speeches (Wang et al. 2011; Li and Hu 2013; Zhang 2013; Pan 2014), these tendencies of “weakening”, “subjectivization” and “de-obligation” can be regarded as interpreters’ efforts to reduce the style gap between Chinese and English GPC speakers and to facilitate the acceptance of the Chinese ST message for the English-speaking TT audience. Here the balance between “fidelity” and “acceptability” is slightly tilted towards the latter, as the TT deviates from strict-sense faithfulness to the ST for better acceptability to the TT audience. Besides the general preference for “weakening”, “subjectivization” and “de-obligation” shifts in the reconstruction of modality (“weakening” and “subjectivization” are mainly caused by the addition of English modality markers in translation), the shifts of “strengthening” and “objectivization” are found to be dominant in the translation of existing Chinese modality markers, and they make the premier/foreign minister sound more assertive, powerful, determined or confident in the TT on politically crucial issues. For instance, the modal 要 in Foreign Minister Li Zhaoxing’s speech segment “日方要以严肃认真的态度处理好历史问题” (see Example 16, Sect. 4.2.2) is rendered into the higher-value must to exert high pressure on the Japanese government regarding their improper behavior of “paying homage to Class-A war criminals”. The modal 可以 is dropped in the interpretation of Premier Wen Jiabao’s speech segment “中国在叙利亚问题上的立场可以概括 为四点” (see Example 31, Sect. 5.2.1) to strengthen and objectify the statement of the Chinese government’s stance on the Syrian issue. The subjective marker 我想 in Foreign Minister Yang Jiechi’s speech segment “我们啊, 国家的发展我想为我们 同其他国家的, 这个, 呃, 合作创造了更大的空间” (Example 35, Sect. 5.2.1) is dropped in the interpretation so that the contribution of China’s development to other countries is objectified and the speaker sounds more confident in the statement. These “strengthening” and “objectivization” shifts also conflict with the strict-sense principle of “fidelity”, but they do not contribute to better “acceptability” either. Instead, they signal the interpreters’ institutional alignment with the Chinese officials as the Chinese government’s stance on crucial issues is strengthened or is explicitated. To sum up, the institutional norm of “fidelity first and acceptability second” indicated in the meta-discourse is generally reflected in the reconstruction of modality in Chinese-English GPC interpreting, but the interpreters in actual GPC settings seem to apply the norm more flexibly in the reconstruction of modality than what the former institutional leaders suggest in the meta-discourse. This confirms previous findings that “the interpreters working for these press conferences do not

120

6 The Reconstruction of Modality in Relation …

always follow the principle of ‘faithfulness, accuracy, completeness’” (Liu 2010, p. 193), and the GPC interpreters are “rather proactively engaged in the construction of interactional meaning” (Sun 2013, p. 118) as institutional allies with the Chinese government. Since the interpreting of journalists’ questions are not mentioned in the meta-discourse, our discussion of GPC interpreting in this section only covers the interpreting of Chinese government officials’ speeches. More comprehensive explanations for all the regular patterns in the reconstruction of modality including the different degree of “strengthening” shifts with different ST speakers will be discussed in the next section from the perspective of typological analysis.

6.2

Modality Reconstruction in GPC Interpreting in Relation to its Typological Features

As Sun (2013, p. 5) points out, GPC interpreting is “a distinct type of interpreting compared to the more commonly known ‘public service/community/dialogue/ liaison interpreting’ and ‘conference interpreting’”. Generally speaking, it is more formal than dialogue interpreting in community settings and more interactive than the strict-sense conference interpreting. As a distinct type of interpreting with Chinese characteristics, its typological features offer important contextual clues for us to understand the interpreters’ choices in modality reconstruction discussed in Chaps. 4 and 5. Inspired by Alexieva (1997) and Pöchhacker (2009), I will present a multi-parameter typological description of GPC interpreting in China according to the following inventory: (1) participants and goals of communication; (2) temporal and spatial setting; (3) topics and text type; (4) working mode and directionality. At the same time, I will discuss how these typological parameters relate to the regular patterns of modality reconstruction in GPC interpreting.

6.2.1

Participants and Goals of Communication

Different from Alexieva (1997, p. 222) who considers the interpreter as the secondary participant in an interpreting event, I regard the interpreters as primary participants along with the ST speakers and the TT listeners, because direct communication between the ST speakers and the TT listeners is not possible without their mediation. In the context of the Chinese-English GPC interpreting, the ST speakers are primarily the Chinese government officials (premiers or foreign ministers) and occasionally the journalists raising questions in Chinese (either from China or abroad), while the TT listeners are primarily the foreign journalists present at the GPCs (when the Chinese ST is uttered by a journalist, the addressee of the ST is the Chinese official, but the English TT listeners are still the foreign journalists). The interpreters are “civil servants” (Chen 2015) working for the Department of

6.2 Modality Reconstruction in GPC Interpreting in Relation …

121

Translation and Interpreting affiliated to the Ministry of Foreign Affairs, and they have been through intense competition while being recruited and then received intensive in-house training for translation and interpreting skills, so they “represent the top-level of the translating/interpreting procession in China” (Ren 2004, p. 61, my translation). The secondary participants who are present at the GPCs but not directly involved in the speaker-listener communication include the initiator/ organizer of the GPC interpreting events which is the National People’s Congress (Chen 2015), and the Chairperson/moderator (typically a senior official from the standing committee of the National People’s Congress) who hosts the press conference by giving opening and concluding remarks and decides who among the journalists gets the chance to ask questions. Another important secondary participant, the quality assessor of the interpreting performance, is a team of experienced translation/interpreting experts at the Department of Translation and Interpreting, who “analyze the interpreting behavior through on-site observation or viewing the video records and then give their evaluation and feedback to the interpreters” (ibid., my translation). In the communicative event of GPC interpreting, different but non-conflicting goals of communication apply to the primary participants. The Chinese government official as the primary speaker aims to convey state policies and express government stances and attitudes on international as well as domestic issues to the international media through the form of press conference, and the invited journalists as the primary listeners aims to get the firsthand message about the Chinese government’s policies and stances towards important national and diplomatic issues. Sometimes journalists may ask challenging questions concerning the relation between the Chinese Mainland and Taiwan, corruption among Chinese officials, etc., but generally speaking, the two sides are cooperative, with one side seeking information and the other side giving information. As the mediator, the interpreter’s goal is to render the government official and journalists’ utterances effectively and appropriately so as to facilitate the communication between them. As institutional insiders and aligners, the interpreter also has the goal of helping the Chinese government publicize China’s policies to the world. The power relations between these participants and their goals of communication have important implications for the reconstruction of modality in GPC interpreting: First, different from literary translation where the act of translation is typically initiated in the target culture (Toury 1995, p. 147), the Chinese-English GPC interpreting events are initiated and organized by the National People’s Congress (the highest organ of state power in China), an institution in the source culture. To be exact, the initiator, the interpreter, the quality assessor and the primary ST speaker of the GPC interpreting events are all institutional aligners of the Chinese government, so it is natural that the interpreting behavior is governed more by the ST-oriented norm (fidelity to the ST speaker) than the TT-oriented norm (acceptability to the TT listener), which explains why “zero modality shifts” is predominant and why “addition” is preferred to “omission” in the reconstruction of modality in Chinese-English GPC interpreting.

122

6 The Reconstruction of Modality in Relation …

Second, as civil servants working for the Ministry of Foreign Affairs, the interpreters would consider themselves institutional aligners with the Chinese official who answers questions but not of the journalists who raise questions. Moreover, the system that the Chairperson who is a government representative decides who among the journalists present at the GPC has a chance to raise questions gives more power to the primary speaker than the journalists. Therefore the degree of “strengthening” shifts with the effect of making the speaker sound more confident, assertive, straightforward or sincere is significantly higher in the translation of Chinese officials’ speeches than that of journalists’ speeches. Third, in the Chinese context of diplomatic interpreting, the interpreters not only function as professional interpreters, but also play the role of “diplomatic workers” and “government representatives” (Guo 2004, p. 14). Thus the in-house interpreters at the GPCs have the dual role of translating the Chinese premiers or foreign ministers’ speeches into English and effectively publicizing China’s policies to the world. It is probably the awareness of this dual role that motivates the interpreters to frequently allow modality shifts in terms of modality value, orientation and modality type instead of always using literal translation for the reconstruction of modality. In particular, the overall preference for “weakening”, “subjectivization” and “de-obligation” shifts reflect the interpreters’ efforts to convey the Chinese government officials’ messages in a more audience-friendly style so as to achieve effective communication between the Chinese government and the foreign media. Fourth, as the primary speakers’ goal of communication at the Chinese-English GPC interpreting events is to accurately convey the Chinese government’s policies and stances on political and diplomatic issues to the foreign media, it is crucial for the interpreters to accurately and effectively transfer the interpersonal meanings of modality from the ST to the TT. Since shifts of modality type are far more likely to cause differences in the basic interpersonal speech function of the corresponding ST and TT clauses than shifts of modality value and orientation, they occur with a much lower frequency than the other two types of shifts in the translation of Chinese modality markers as well as the use of English modality markers. As a result, the basic interpersonal speech function of the relevant clause (statement/ question/offer/command) rarely changes despite the frequent shifts in the semantics of modality in GPC interpreting. Last, despite the limited utterances from the journalists, the interpreters also have an overall tendency towards “weakening”, “subjectivization” and “de-obligation” shifts in reconstructing modality of their speeches (see Tables 4.11 and 5.11), to the effect that the journalists sound more polite and respectful (see Example 2, 6 and 25) when raising questions to Chinese premiers and foreign ministers. Although the primary listeners of the TT here are not the Chinese officials themselves but the foreign journalists present at the Press Conference, the interpreters have reconstructed a proper power relationship between the question-asking journalists and their addressees in the TT through these “weakening”, “subjectivization” and “de-obligation” shifts, so that the “face” of the Chinese government has been protected in front of an international audience. This again reflects the interpreter’s institutional and emotional alignment with the Chinese government.

6.2 Modality Reconstruction in GPC Interpreting in Relation …

6.2.2

123

Topics and Speech Type

Generally speaking, speeches made at GPCs belong to a specialized political-diplomatic genre mostly in the form of question and answers, so topics discussed depend on the questions raised by the journalists. But the Chairperson plays an important role in confining the questions to prepared topics, mainly on political, diplomatic, economic issues and military affairs concerning China. The number of journalists at one GPC could be as large as 800 (Zhao 2007b, p. 36), which creates a formal atmosphere for the interpreting event. In addition, as the GPCs during the NPC and CPPCC sessions are accessible to the public through live TV broadcasts, the numerous off-site TV viewers should be counted as “ratified” but “unaddressed” listeners (see Goffman 1981, pp. 131–132). Due to the high level of publicity and the political sensitivity of its topics, a fairly high degree of planning is involved while the government official who is the primary speaker utters the ST. The organizers of the GPC will predict and prepare for possible questions through careful media analysis although they will not know the exact questions to be asked.2 In other words, the STs of Chinese-English GPC interpreting are mostly half-prepared speeches instead of fully improvised speeches, but the degree of spontaneity is higher than the STs of the narrow-sense conference interpreting in technical settings. The politically sensitive topics and the formal, half-prepared speech style of the STs at Chinese GPCs serve as important motivators for the interpreters to prefer “zero modality shifts” to “modality shifts” in the reconstruction of modality in GPC interpreting, and prefer the “addition” of English modality markers to the “omission” of Chinese modality markers in terms of macro-shifts. First, because communication at GPCs revolves around politically sensitive topics and a large audience is involved, Chinese premier or foreign minister as the primary speaker usually state the policies and opinions in a formal style “with deliberation over the wording” (Xu 2000, p. 35). Presumably modality markers as important realizations of interpersonal meaning are also carefully chosen in their speeches, so interpreters generally prefer to use literal translation in the reconstruction of modality to ensure a faithful and accurate rendition of the ST. Second, due to such a high degree of planning in the production of the source speech, interpreters prefer “addition” to “omission” in terms of macro-shifts, as the omission of the existing Chinese modality markers (apart from obligatory shifts out of linguistic differences) in the ST is more likely to be criticized as incomplete and unfaithful translation than the addition of English modality markers in the TT, which can be considered as a means to faithfully render the intended meaning of the ST speaker. For instance, the dropping of the subjective marker 我想 in the interpretation of Foreign Minister Yang Jiechi’s speech segment “呃, 我们啊, 国家的发展 我想为我们同其他国家的, 这个, 呃, 合作创造了更大的空间” (see Example 35, 2

China News Week, 16 March, 2013, http://news.china.com.cn/2013lianghui/2013-03/16/content_ 28261953.htm, last accessed on 24 July 2014.

124

6 The Reconstruction of Modality in Relation …

Sect. 5.2.1) is likely to be regarded as incomplete translation which makes the speaker sound unnecessarily assertive. By contrast, the addition of the subjective marker I think in the interpretation of Premier Zhu Rongji’s speech segment “这个问 题不应该由我来回答, 而且现在还言之过早” (see Example 9, Sect. 4.2.1) is more likely to be regarded as the faithful rendition or proper explicitation of Mr. Zhu’s intended message.

6.2.3

Temporal and Spatial Setting

As the Chinese GPCs in this study are confined to the press conferences held during the “Two Sessions”, and the corpus data further restricts the time to the period between 1989 and 2014, we can describe the temporal and spatial setting as contemporary China. The 25 years witness not only the development of China’s public diplomacy and the GPC system but also the revitalization of diplomatic translation/ interpreting in China. According to Ms. Shi Yanhua (Shi 2009, pp. 9–12), the 60-year history of diplomatic translation/interpreting can be divided into four periods: (1) the establishment of the translation/interpreting team of the Ministry of Foreign Affairs at the founding of the People’s Republic of China (1949); (2) the blossoming of the translation/interpreting team of the Ministry of Foreign Affairs with the care and trust from Premier Zhou Enlai (1950–1965); (3) the stagnation of diplomatic translation/interpreting during the Cultural Revolution (1966–1976); (4) the revitalization of diplomatic translation/interpreting since the implementation of the Reform and Opening-up Policy (1978–2009). Right after the founding of the People’s Republic of China in 1949, a team of translators/interpreters from the Ministry of Foreign Affairs was established to help develop diplomatic relations with friendly countries and combat anti-China activities from some western countries so as to win international recognition of the “New China”. This translation/interpreting team soon developed into a specialized department for diplomatic translation and interpreting (now known as the Department of Translation and Interpreting) affiliated to the Ministry of Foreign Affairs to perform numerous interpreting tasks for government leaders as well as translation tasks for important political documents. Between 1950 and 1965, this specialized translation/interpreting team grew quickly (in both size and quality) with the care and trust from Premier Zhou Enlai. Besides the high expectation for diplomatic translators/interpreters to be knowledgeable and conscientious, Premier Zhou also attached great importance to their role in diplomatic publicity: “Translators/interpreters are not just sounding machines. Instead, you are human beings, Party (Chinese Communist Party) members, and revolutionary cadres. You should have political consciousness and subjective agency, so that you can play a greater role. Of course you cannot make irresponsible remarks. Your first task is to translate/interpret, and your second task is to publicize China’s stances and principles to the outside world.” (ibid., p. 10, my translation) This healthy development of diplomatic translation/interpreting was halted in the next 10 years (1966–1976)

6.2 Modality Reconstruction in GPC Interpreting in Relation …

125

as the whole country was thrown into turmoil by the Cultural Revolution. Due to the chaotic political environment, the translation/interpreting team within the Ministry of Foreign Affairs was almost dismissed and the quality of translation/interpreting went downhill. For example, “最最最强烈的抗议” at that time was translated into “strongest and strongest protest” regardless of the English language norms. (ibid., p. 11) Fortunately, this chaotic period was soon ended and diplomatic translation/interpreting in China has been revitalized since the implementation of the Reform and Opening-up Policy in 1978. With the increase of diplomatic activities and international communication, the workload of the Department of Translation and Interpreting increased and a rigorous selection and training system was established within the Ministry of Foreign Affairs to ensure the quality of the translators/interpreters on the team. As the theme of the political environment changed from “revolution and fighting” into “peace and development”, the diplomatic translators/interpreters’ role as mediators or facilitators for international communication was foregrounded. Note that the comment of Ms. Shi (ibid., p. 12; see Sect. 6.1.1.1) on the recent attempts of diplomatic interpreters to “make the target speech more smooth and intelligible with faithfulness to the original text ensured” was made in her discussion of the fourth period (1978–2009), which to a large extent overlaps with the span of the corpus data in the present study (1989–2014). As for the specifics of the spatial setting, the GPC interpreting events take place in the Great Hall of the People, Beijing, which is the venue for legislative (including the NPC and CPPCC sessions) and ceremonial activities organized by the central government of China. So the location is home to the primary speaker, the interpreter, the Chairperson, and the Chinese journalists, but foreign to journalists from other countries. In terms of the inter-personal distance, the interpreter usually sits by the side of the government official at the rostrum and the journalists sit at a longer distance from the official in the audience seats facing the rostrum. In addition, the “space” in which the interpreting event takes place is not exclusively reserved for the primary and secondary participants, but shared by numerous off-site TV viewers, thus there is relatively “less privacy” and “more self-monitoring” (Alexieva 1997, p. 229) on the part of speakers and interpreters in the GPC interpreting than in community interpreting. That is to say, although the primary speakers and the interpreters are pretty familiar with the venue, a fairly high degree of planning is involved in their speeches because of the formal setting and the publicity of the event. The temporal and spatial setting also sheds light on our understanding of the regular patterns of modality reconstruction in GPC interpreting: First, the GPC interpreting investigated in the present study is situated in the post-Reform-and-Opening-up “revitalization” period of diplomatic interpreting which witnesses the efforts of interpreters to attend to not only the ST-oriented principle of “fidelity” but also the TT-oriented principle of “acceptability”. Therefore, along with the predominant choice of “literal translation” for modality reconstruction, interpreters also frequently allow modality shifts especially shifts of “weakening” and “subjectivization” (often through the addition of modality

126

6 The Reconstruction of Modality in Relation …

markers, see Example 1, 3 and 9 in Sect. 4.2.1) to make the Chinese premiers or foreign ministers sound more audience-friendly in communicating their policies and stances. In fact, as the GPC interpreting is a cognitively demanding task, using “stock equivalents” for the translation of modality would be a more economical choice in line with Lévy’s “minimax strategy” (1967, p. 1179), yet interpreters frequently choose to reconstruct modality in a flexible way with the overall preference for “weakening”, “subjectivization” and “de-obligation” shifts. These interpreting choices are motivated by the friendly and open political environment in the post-Reform-and-Opening-up period, which witnesses interpreters enjoying more subjective agency than the previous period of the Cultural Revolution. Second, the Chinese officials and the interpreters’ familiarity with the venue of the press conferences and the physical proximity between them reinforces the interpreters’ emotional alignment with the Chinese officials and their emotional distance from the journalists, which explains why interpreters have a stronger tendency to strengthen the modality in Chinese premiers or foreign ministers’ speeches than in journalists’ speeches. As Example 14, 16 and 17 in Sect. 4.2.2 and Example 27 and 35 in Sect. 5.2.1 show, interpreters often replace the Chinese modality markers with higher-value English modality markers or dropping the Chinese modality markers (应该 ! must; 要 ! must; 我想 ! I believe; 要 ! 0; 我想 ! 0) in translating Chinese officials’ speeches, so that the Chinese officials sound more assertive, powerful and sincere in their statements, suggestions and offers. Third, the “open space” for the general public to get involved as “ratified” listeners of the press conferences foregrounds the high-level publicity and formality of the interpreting events, which motivates both the Chinese officials and interpreters to be very cautious over their wording. As discussed in Sect. 6.2.2, this formal setting and high degree of “self-monitoring” on the part of the primary speakers explains why interpreters prefer “zero modality shifts” to “modality shifts” in the reconstruction of modality and prefer the “addition” of English modality markers to the “omission” of Chinese modality markers in terms of macro-shifts.

6.2.4

Working Mode and Directionality

In terms of the working mode, two types of interpreting are usually distinguished: simultaneous interpreting (the speaker delivers the ST non-stop and the interpreter renders it almost simultaneously) and consecutive interpreting (the speaker delivers a segment of the ST, and then pauses for the interpreter to render it). Different from the European tradition of employing the simultaneous mode for conference interpreting, GPC interpreting investigated in this study follows the consecutive mode, the “‘classic’ consecutive” (Pöchhacker 2009, p. 19) interpreting mode with note-taking, to be exact. According to the International Association of Conference Interpreters, directionality of interpreting is distinguished with reference to the interpreter’s working

6.2 Modality Reconstruction in GPC Interpreting in Relation …

127

languages. The interpreter’s native or best “active” language is classified as A language, his/her “active” language commanded with near-native proficiency is classified as B language, and his/her “passive” language allowing “complete understanding” is called C language. (Pöchhacker 2009, p. 21) For Chinese in-house interpreters, Chinese is their A language and English their B language. Different from the western tradition of interpreting into the interpreter’s A language for important conferences (ibid., 21), the Chinese GPCs involves bi-directional interpreting performed by the same interpreters, i.e., both B-to-A (English to Chinese) and A-to-B (Chinese to English) interpreting, but A-to-B interpreting is only needed when journalists ask questions in English. Most of the time during the Chinese GPC event, the ST is in Chinese, so A-to-B interpreting is predominant. In fact, for the convenience of text alignment, my corpus only includes the A-to-B (Chinese to English) interpreting data at the GPCs, which is the subject of the present study. As for why the Chinese government asks the Chinese interpreters to perform the so-called “inverse translation” (Wang 2011, p. 902) at GPCs, the reason is probably both practical (it is not easy to find English native-speakers with the ability to render the Chinese GPCs) and ideological [the Chinese government needs to ensure the interpreter’s “political allegiance” (Beeby 2009, p. 86)]. Besides this perspective of A-to-B interpreting, another important aspect of the directionality of GPC interpreting concerns the two distinctive languages: Chinese and English. Since modality shifts in the present study exclude obligatory shifts caused solely by the difference in the two linguistic systems, grammatical differences between Chinese and English will not be a major factor to be considered here. However, it is important to discuss the different communicative conventions between Chinese and English. As suggested by previous studies (Hu and Cao 2011; Zhao and Liang 2013; Li and Hu 2013; Zhang 2013; Pan 2014), remarkable differences in communicative conventions exist between the two languages and particularly between Chinese and English political speeches: the Chinese communicative convention is content-oriented and authority-oriented and thus is characterized by the frequent use of strong modals, while the English communicative convention is addressee-oriented and argumentation-oriented and thus is characterized by the frequent use of weak modals; the English political speeches are more personal and consultative in style than the Chinese ones; obligation modals are used more frequently in Chinese political speeches than English ones. Also of relevance here is the difference between what the anthropologist Hall (1976) describes as “low-context” and “high-context” cultures. In order to understand the basic differences in communication style between communities, Hall (ibid.) categorized cultures on a scale from high to low context. In a high-context culture (such as China), people typically communicate in the way that “most of the information is already in the person, while very little is in the coded, explicit, or transmitted part of the message” (ibid., 91, original emphasis). In a low-context culture (such as the US and other English-speaking countries), the style of communication is “just the opposite; i.e., the mass of information is vested in the explicit code” (ibid., 91). Thus, comparatively speaking, communication in the Chinese-speaking culture is more dependent on the non-verbal context than in the English-speaking culture.

128

6 The Reconstruction of Modality in Relation …

These differences coupled with the interpreting direction from Chinese to English help account for the patterns of modality shifts in GPC interpreting. The working mode and directionality of GPC interpreting have important implications for the interpreters’ choices in modality reconstruction: First, at GPCs investigated in this study, i.e., press conferences of Chinese premiers or foreign ministers during the annual “Two Sessions” of the NPC and CPPCC, interpreters are asked to work in the “classic consecutive” mode instead of the more time-saving simultaneous mode to “ensure the accuracy of interpreting” and to “give the government leaders more time to prepare for their answers” (Chen 2015). The choice of the working mode shows the high expectation of “accuracy” in GPC interpreting and allows the primary speakers to weigh their words (including modality markers) before uttering them, so it is natural that interpreters prefer “zero modality shifts” to “modality shifts”, and prefer “addition” to “omission” in the reconstruction of modality. Second, as the Chinese premiers or foreign ministers often utter a long segment of speech before pausing for rendition, the interpreters have to take notes carefully while listening to the speech segments. However, in classic consecutive interpreting, “[n]otes do not reproduce the speech; they are only written indications to help the interpreter remember it” (Gile 1995, p. 181). In other words, interpreters have to take notes in a highly selective way to ensure sufficient “processing capacity” for “listening and analysis” (ibid., 179–181). Thus the GPC interpreters probably did not choose to take down the exact modality markers as the key content words, numbers and proper nouns were more challenging to their memory capacity. Among the 1236 cases (41%) of modality shifts in the use of English modality markers and 853 cases (39.8%) of modality shifts in the translation of Chinese modality markers, some may have been caused by their failure to remember the exact modality marker used in the source speech. Third, the exclusive employment of Chinese interpreters to do the “A-to-B” (i.e., Chinese to English) interpreting primarily for the Chinese premiers or foreign ministers at the GPCs shows that an accurate understanding and faithful reproduction of the ST is more important than a smooth production of the TT in this Chinese GPC interpreting context, which explains why literal translation is preferred in the reconstruction of modality. Moreover, the direction of “A-to-B” translation ensures the interpreter’s political alignment with the primary speaker, which explains why interpreters prefer the shifts of “strengthening” and “objectivization” in rendering existing Chinese modality markers to make the Chinese premier or foreign minister sound more confident, sincere, powerful or authoritative on important political issues. Last but not least, due to the different communicative conventions between the two languages/cultures and particularly between Chinese and English political speeches discussed above, the direction of interpreting from Chinese to English prompts the interpreters to prefer “weakening”, “subjectivization” and “de-obligation” shifts among all options of modality shifts so as to accommodate the communicative convention of the English language and English political speeches as being more consultative, personal and less obligational. This effect of

6.2 Modality Reconstruction in GPC Interpreting in Relation …

129

directionality is clearly revealed through a comparison of the overall tendency towards “weakening” in Chinese-English interpreting found in the present study with the “strengthening” tendency found in English-Chinese translation by Zhao and Liang (2013). Also the direction of translation from a high-context culture to a low-context culture naturally encourages the interpreters to frequently add English modality markers rather than omit Chinese modality markers in the reconstruction of modality. As the source culture (Chinese-speaking culture) is a high-context one, the modal meanings are likely to be hidden in the non-verbal context in the Chinese ST, the interpreters who are Chinese native-speakers and colleagues of the primary ST speakers can well detect the implicit messages and then render them more explicitly into the English TT, so that the intended messages of the Chinese officials can be communicated successfully to the English-speaking audience, who belong to a relatively low-context culture.

6.3

Summary

In this chapter, I have discussed the relationship between the regular patterns in the reconstruction of modality in Chinese-English GPC interpreting and its sociocultural context, which includes the meta-discourse on diplomatic interpreting in China and the typological features of Chinese GPC interpreting. The meta-discourse on Chinese diplomatic interpreting, i.e., the published discourse from institutional insiders of the Department of Translation and Interpreting, Ministry of Foreign Affairs where the GPC interpreters work, and my personal interview with the former Director Mr. Chen Mingming reveals the institutional norm of “fidelity first and acceptability second”, which is reflected in the general preference for “zero shifts” to “modality shifts” in terms of semantic shifts and the preference for “addition” to “omission” in terms of macro-shifts, but is contradicted by the high frequency of optional “modality shifts” in the reconstruction of modality. Thus I conclude that the interpreters’ linguistic choices in the Chinese GPC settings are constrained by the institutional norm of “fidelity first and acceptability second”, but they apply the norm to modality-related translation more flexibly than what their former leaders suggest in the meta-discourse. The discussion of the relationship between the typological features of GPC interpreting with the interpreting choices concerning modality further reveals the nature of GPC interpreting as a socially situated communicative act mediated by the interpreters. Its typological features in terms of the participants, goals of communication, temporal and spatial setting, topics, text type, working mode, and directionality serve as important contextual clues for us to understand the regular patterns in modality reconstruction discussed in Chaps. 4 and 5. The institutional alignment between the initiator, the interpreter, the quality assessor and the primary ST speaker of the GPC interpreting events, the politically sensitive topics, the “open space” for numerous TV viewers to get involved as the off-site audience, the formal speech style of the Chinese officials’ speeches, the “classic consecutive” working

130

6 The Reconstruction of Modality in Relation …

mode and the direction of “A-to-B” interpreting all serve as motivators for interpreters to prefer “zero modality shifts” to “modality shifts” and prefer “addition” to “omission” in the reconstruction of modality. The unsymmetrical power relations between the Chinese premiers/foreign ministers, the interpreters and the question-asking journalists, their different degree of familiarity with the GPC venue and the different physical distance between them explain why the degree of “strengthening” shifts is significantly higher in the interpretation of Chinese officials’ speeches than journalists’ speeches. The Chinese government officials and the interpreters’ goals of communication explain why shifts of modality type occur with a much lower frequency than shifts of modality value and orientation. The emotional alignment between the interpreters and the Chinese officials also prompts the interpreters to prefer “weakening”, “subjectivization” and “de-obligation” shifts in translating journalists’ speeches to make them sound more polite and respectful. The dual goal of the GPC interpreters to translate the Chinese government officials’ speeches into English and to effectively publicize China’s policies to the world, the temporal setting (the Post-Reform-and-Opening-up “revitalization” period of diplomatic interpreting) of the interpreting data in this study, and the direction of “Chinese-to-English” interpreting coupled with the different communicative conventions between the two languages/cultures explain why interpreters frequently allow modality shifts especially the semantic shifts of “weakening”, “subjectivization” and “de-obligation” and the macro-shifts of “addition” for a better communicative effect. In addition, the “classic consecutive” working mode implies the possibility of unintentional modality shifts, which contributes to the high frequency of modality shifts, and the “A-to-B” direction of GPC interpreting also explains why interpreters prefer the shifts of “strengthening” and “objectivization” in rendering existing Chinese modality markers to make the Chinese official sound more confident, sincere, powerful or authoritative on important political issues. To sum up, both the meta-discursive analysis and the multi-parameter typological analysis contribute to our understanding of the reconstruction of modality in Chinese-English GPC interpreting and our understanding of the nature of interpreting as a socially situated communicative act. The institutional norm of “fidelity first and acceptability second” for Chinese diplomatic interpreting revealed by the meta-discourse proves to have strong but not absolute governing power for GPC interpreters, and the actual interpreting choices in the reconstruction of modality are more comprehensively explained by the distinct typological features of GPC interpreting.

References Alexieva, Bistra. 1997. A typology of interpreter-mediated events. The Translator 3 (2): 153–174. Beeby, Allison. 2009. Directionality. In Routledge encyclopedia of translation studies, ed. Mona Baker, and Gabriela Saldanha, 84–88. London/New York: Routledge. Chen, Mingming 陈明明. 2015. Telephone interview. 16:00–16:30, 27 April.

References

131

Chesterman, Andrew. 1996. Review of descriptive translation studies and beyond, by Gideon Toury. Target 8 (1): 197–201. Diriker, Ebru. 2004. De-/re-contextualizing conference interpreting. Amsterdam/Philadelphia: John Benjamins. Gile, Daniel. 1995. Basic concepts and models for interpreter and translator training. Philadelphia: John Benjamins. Goffman, Erving. 1981. Forms of talk. Oxford: Basil Blackwell. Guo, Jiading 过家鼎. 2004. 岁月不曾流逝的记忆——外交翻译生涯三十余年点滴谈 [Reflections on my experience of diplomatic translation for over 30 years]. 对外大传播 [International Communications] 9: 12–14. Hall, Edward T. 1976. Beyond culture. New York: Doubleday. Hu, Guangwei, and Feng Cao. 2011. Hedging and boosting in abstracts of applied linguistics articles: A comparative study of English- and Chinese-medium journals. Journal of Pragmatics 43 (11): 2795–2809. Lévy, Jirí. 1967. Translation as a decision process. In To Honor Roman Jakobson, 1171–1182. The Hague: Mouton. Li, Xin 李鑫, and Kaibao Hu 胡开宝. 2013. 基于语料库的记者招待会汉英口译中情态动词的 应用研究 [A Corpus-based Study of Modal Verbs in Chinese-English Government Press Conference Interpretation]. 外语电化教学 [Computer-Assisted Foreign Language Education] 74 (3): 26–32. Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese-English): Face management and information management. Ph.D. Thesis, Heriot-Watt University. Pan, Feng 潘峰. 2014. 基于语料库的汉英会议口译中模糊限制语的应用研究 [A corpus-based study of the application of hedges in Chinese-English Conference Interpreting]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] (4): 24–29. Pöchhacker, Franz. 2009. Introducing interpreting studies. Shanghai: Shanghai Foreign Language Education Press. Ren, Xiaoping 任小萍. 2004. 外交部高级翻译培训 [The training of senior translators/interpreters at the Ministry of Foreign Affairs]. 中国翻译 [Chinese Translators Journal] (1): 61–62. Shi, Yanhua 施燕华. 2007. 怎样做好外交口译工作 [Tips on diplomatic interpreting]. 中国翻译 [Chinese Translators Journal] (3): 57–60. Shi, Yanhua 施燕华. 2009. 外交翻译60年 [Sixty years of diplomatic interpreting]. 中国翻译 [Chinese Translators Journal] (5): 9–12. Sun, Tingting. 2013. Interpreting China: Interpreters’ mediation of Government Press Conferences in China. Beijing: Foreign Language Teaching and Research Press. Toury, Gideon. 1995. Descriptive translation studies and beyond. Amsterdam and Philadelphia: John Benjamins Publishing Company. Wang, Nongsheng 王弄笙. 1991. 外事汉英翻译中的几点体会 [Reflections on Chinese-English diplomatic translation/interpretation]. 中国翻译 [Chinese Translators Journal] (3): 6–12. Wang, Nongsheng 王弄笙. 2002. 近年来汉英翻译中出现的一些新问题 [Some new problems in Chinese-English translation]. 中国翻译 [Chinese Translators Journal] (1): 23–28. Wang, Wangni 王望妮. 2008. 中美政府首脑记者招待会语言对比研究 [Evaluating the way government practice public communication: A comparative approach]. Ph.D. Thesis, Shanghai International Studies University. Wang, Baorong. 2011. Translation practices and the issue of directionality in China. Meta 56 (4). Wang, Hesi 王和私, Pi’an Yin 尹丕安, and Furong Wang 王芙蓉. 2011. 中英文政治演说的情 态对比研究 [A comparative study on modality in Chinese and English political speeches]. 西安外国语大学学报 [Journal of Xi’an International Studies University] (2): 38–42. Xu, Yanan 徐亚男. 2000. 外交翻译的特点以及对外交翻译的要求 [Features of diplomatic translation/interpretation and the required qualities for diplomatic translators/interpreters]. 中国 翻译 [Chinese Translators Journal] (3): 35–38.

132

6 The Reconstruction of Modality in Relation …

Zhang, Yuanyuan 张援远. 2004. 谈谈领导人言论英译的几个问题 [Several issues concerning the English translation of Chinese leaders’ speeches]. 中国翻译 [Chinese Translators Journal] (1): 55. Zhang, Guanfang 张冠芳. 2013. 系统功能语言学视角下的英汉政治演讲的情态对比研究 [A contrastive analysis on modality in English and Chinese political speeches from the perspective of Systemic Functional Linguistics]. M.A. Thesis, Ningbo University. Zhao, Hongyan 赵鸿燕. 2007a. 政府记者招待会: 历史, 功能与问答策略 [Government Press Conferences: History, functions and Q & A strageties]. Beijing: Communication University of China Press. Zhao, Qizheng. 2007b. Better public diplomacy to present a truer picture of China. People’s Daily Online, 30 March. http://en.people.cn/200703/30/eng20070330_362496.html. Accessed 14 Sept 2015. Zhao, Qiurong 赵秋荣, and Maocheng Liang 梁茂成. 2013. 认识型情态动词 may 和 might 汉译 强度变化研究 [The modal strength of epistemic verbs may and might in English-Chinese Translation]. 山东外语教学 [Shandong Foreign Languages Teaching Journal] 108 (6): 96–99.

Chapter 7

Conclusion

Based on an interpreting corpus of 21 parallel texts (293,284 tokens), the present study has attempted to investigate the reconstruction of modality in ChineseEnglish GPC interpreting and discuss the regular patterns of modality shifts in relation to the sociocultural context. The broad framework of DTS and the modality theory from SFG have been adopted to analyze and describe the regular patterns of modality shifts at the semantic level, and the meta-discourse and typological features of GPC interpreting have been analyzed as the sociocultural context, which proves to be important background for understanding the interpreters’ choices in reconstructing modality in GPC interpreting. As a conclusion to the study, this chapter aims to summarize the major findings of the present study, discuss its implications and limitations and offer suggestions for future research.

7.1

Major Findings

The major findings of the present study can be summarized in relation to the research questions raised in Chap. 1, which are reproduced here for reference: 1. How are modal meanings reconstructed in Chinese-English GPC interpreting? 1a. What are the regular patterns in the reconstruction of modality in Chinese-English GPC interpreting? 1b. Does the identity of ST speaker affect the degree or major types of modality shifts in Chinese-English GPC interpreting? 1c. What are the interpersonal effects of the major types of modality shifts? 2. What is the relationship between the interpreters’ linguistic choices of reconstructing modality and the sociocultural context of Chinese-English GPC interpreting? © Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2_7

133

134

7 Conclusion

2a. How do the regular patterns of modality shifts relate to the institutional norms of Chinese-English GPC interpreting indicated in the metadiscourse? 2b. How does the regular patterns of modality shifts relate to the typological features of Chinese-English GPC interpreting? First, to answer Research Question 1, I have attempted to analyze semantic shifts of modality in the corpus from two directions: the use of English modality markers in the TT and the translation of Chinese modality markers in the ST, which reveal both shared patterns and different features. Taken together, they depict the overall patterns of modality shifts in the reconstruction of modality in Chinese-English GPC interpreting. Corpus-based analyses from both directions reveal the following shared patterns: (1) The general degree of modality shifts is significantly lower than that of zero-shifts, indicating the overall preference for “literal translation” or “equivalent translation” of modality, but the degree of modality shifts (around 40% of the modality markers involve modality shifts) is still fairly high considering the availability of equivalent Chinese/English modality markers. (2) The distribution of modality shifts in the three dimensions follows the pattern that shifts of modality value are the most common, shifts of orientation fairly common and shifts of modality type the least common. (3) Despite the low frequency compared with shifts in other dimensions, the interpreters have a clear tendency towards “de-obligation” in terms of shifts of modality type. Analyses from the two directions also reveal the following different features: (1) The contribution of macro-shifts (i.e., omissions or additions) to modality shifts in the translation of Chinese modality markers is not so prominent as in the use of Chinese modality markers, and the frequency of adding English modality markers is also significantly higher than that of omitting Chinese modality markers against the total occurrence of Chinese/English modality markers. It is safe to conclude that the GPC interpreters prefer the “addition” of Chinese modality markers to the “omission” of Chinese modality markers in terms of macro-shifts. (2) Different from the predominance of “subjectivization” and “weakening” shifts in the use of English modality markers, the translation of Chinese modality markers reveal the predominance of “objectivization” and “strengthening” shifts. Viewed together (see Tables 4.7, 4.9, 5.7 and 5.9), data from the two directions reveal the interpreters’ overall preference for “weakening”,1 “subjectivization” and “de-obligation” in the reconstruction of modality with “weakening” and “subjectivization” mainly caused by the addition of English modality markers, but there is a clear tendency towards “strengthening” and “objectivization” in rendering existing Chinese modality markers in the ST (Research question 1a). Analyses from both directions reveal that the identity of ST speaker doesn’t affect the general degree or predominant types of modality shifts in the reconstruction of modality but does affect the degree of the “strengthening” shift, which This overall tendency towards “weakening” supports the hypothesis discussed in Sect. 2.2.1.

1

7.1 Major Findings

135

occurs with a higher probability in the translation of Chinese government officials’ speeches than journalists’ speeches (mainly reflected in the use of must and I/we believe, and the translation of 要 and 我/我们想) (Research Question 1b). I have conducted case studies of the major contributors of the major types of modality shifts in both the use of English modality markers and the translation of Chinese modality markers, which can be brought together to cover the interpersonal effects of “de-obligation”, “weakening”, “subjectivization”, “strengthening” and “objectivization” in the reconstruction of modality: (1) The “de-obligation” shifts mainly occurring in the use of will, would and be going to and the translation of 应 该, 要, 可以 and 需要 either render the Chinese officials’ speeches more sincere and responsive to the audience, or make the journalists’ questions more respectful to the speaker and appealing to the audience. (2) The shifts of “subjectivization” and “weakening” mostly occurring with the addition of will, would, can and I think foreground the Chinese government’s subjectivity and make the Chinese officials or journalists sound more friendly and polite. (3) The “strengthening” shifts alone mainly occurring in the use of will, must and I/we believe and in the translation of 要, 能/能够, 可以 and 我/我们想 help present the Chinese government as more confident, sincere and powerful on important political issues, and the shifts of “objectivization” that mostly co-occur with “strengthening” shifts in the translation of 要, 能/能够, 可以 and 我/我们想 either make the Chinese government speakers sound more confident and authoritative or make the journalists sound more straightforward and concise. These case studies also reveal that the basic interpersonal speech function—statement/question/offer/command—rarely changes despite the frequent occurrence of modality shifts in the reconstruction of modality. Thus I conclude that the interpreters frequently allow modality shifts instead of always using “stock equivalents” in their reconstruction of modality in Chinese-English GPC interpreting mainly for achieving a better communicative effect (Research Question 1c). Second, I have addressed Research Question 2 by mapping the regular patterns of modality shifts in Chinese-English GPC interpreting onto the sociocultural context of GPC interpreting, which includes the meta-discourse on diplomatic interpreting and the typological features of GPC interpreting in China. The meta-discourse on Chinese diplomatic interpreting (the published discourse from institutional insiders and the personal interview with Mr. Chen Mingming, the former Director of the Department of Translation and Interpreting, Ministry of Foreign Affairs) reveals the institutional norm of “fidelity first and acceptability second”, which is in a loose sense reflected in the preference for “zero semantic shifts” to “semantic shifts” and the preference for “additions” to “omissions” of modality markers, but is in a strict sense contradicted by the high frequency of optional “modality shifts” in Chinese-English GPC interpreting. Thus I conclude that the GPC interpreters’ linguistic choices concerning modality are constrained by the institutional norm of “fidelity first and acceptability second”, but they seem to apply the norm more flexibly than what their former leaders suggest in the meta-discourse. This confirms Diriker’s (2004, p. 139) finding that the norms of interpreting revealed by the meta-discourse serve as “performance instructions”

136

7 Conclusion

whose actual turnout is actively “negotiated by the interpreters” (original emphasis) (Research Question 2a). The typological features of GPC interpreting including its participants and their goals of communication, its temporal and spatial setting, topics, text type, working mode, and directionality all prove to be important motivating factors accounting for the interpreters’ regular choices concerning modality. For example, the institutional alignment between the initiator, the interpreter, the quality assessor and the primary ST speaker of the GPC interpreting events, the politically sensitive topics, the “open space” for the general public to get involved as off-site audience of the events, the formal speech style of the Chinese officials’ speeches, the “classic consecutive” working mode and the direction of “A-to-B” interpreting all serve as motivators for interpreters to prefer “zero modality shifts” to “modality shifts” and prefer “addition” to “omission” in the reconstruction of modality. The unsymmetrical power relations between the Chinese government officials, the interpreters and the question-asking journalists, their different degree of familiarity with the GPC venue and the different physical distance between them explain why the degree of “strengthening” shifts is significantly higher in the interpretation of Chinese officials’ speeches than journalists’ speeches. The dual goal of the GPC interpreters to translate the Chinese government officials’ speeches into English and to effectively publicize China’s policies to the world, the temporal setting of the interpreting data in this study, and the direction of “Chinese-to-English” interpreting coupled with the different communicative conventions between the two languages explain why interpreters frequently allow modality shifts especially shifts of “weakening”, “subjectivization” and “de-obligation” instead of always resorting to literal translation in the reconstruction of modality. Moreover, the “classic consecutive” working mode implies the possibility of unintentional modality shifts, which contributes to the high frequency of modality shifts, and the political alignment between the interpreters and the Chinese government officials as ensured by the “A-to-B” direction of GPC interpreting also explains why interpreters prefer the shifts of “strengthening” and “objectivization” in rendering existing Chinese modality markers to make the Chinese officials sound more confident, sincere, powerful or authoritative on important political issues (Research Question 2b). Since the present study aims to explore the ST-TT relations concerning modality in interpreting rather than investigate the linguistic features of the interpreted texts, comparative analysis has only been conducted between the Chinese source texts and interpreted English texts, different from the mainstream product-oriented CTS projects involving the comparison of translated texts and non-translated texts in the same genre. However, viewing the patterns of modality shifts in Chinese-English GPC interpreting together with the contrastive patterns of modality between the interpreted texts and the original GPC texts discussed in Li and Hu (2013) (reviewed in Sect. 2.2.2), we can see that on one hand, GPC interpreters tend to overuse higher-value and obligational modality markers under the influence of the Chinese ST, on the other hand, GPC interpreters tend to opt for “weakening” and “de-obligation” shifts in their reconstruction of modality under the influence of the TT-norms. Therefore, GPC interpreters’ actual practice in reconstructing modality

7.1 Major Findings

137

generally reflects a “dynamic equilibrium” (Monacelli 2005, p. 5) between the ST-oriented norm of fidelity and the TT-oriented norm of acceptability with a bias towards the ST, which is also indicated in the meta-discourse. The close relationship between the regular patterns of modality shifts in GPC interpreting and its meta-discourse and typological features discussed above reveals both the nature of GPC interpreting as socially situated practice and the interpreters’ active role in mediating the communication between Chinese government speakers and foreign journalists.

7.2

Contributions and Implications

As a corpus-based study of modality shifts in Chinese-English GPC interpreting with the relationship between interpreting choices and the sociocultural context discussed, the present study has both theoretical/methodological contributions and practical implications. First, since CIS is still a “cottage industry” compared with “the relatively organized and coordinated landscape of CTS” (Setton 2011, p. 34), this study has attempted a comprehensive analysis of modality shifts in Chinese-English GPC interpreting based on the specialized interpreting corpus CECIC 4.0 and has revealed interesting regularities in the interpreters’ choices. Thus this study contributes to the development of CIS as an “offshoot” (Shlesinger 1998, p. 1) of CTS and a promising empirical approach of IS. Second, by integrating the broad framework of DTS, the corpus-based methods, and the context analysis of interpreting, this study offers a descriptive-explanatory model for socially-oriented interpreting studies, which proves to compensate for the limitation of purely descriptive corpus-based methods. Since “corpora findings may tell us how translators translate, but not why they translate the way they do” (Straniero Sergio and Falbo 2012, p. 22), this study illustrates a feasible approach for CIS to go beyond pure description of corpus findings and dig for explanations in the sociocultural context of interpreting. Third, against the background that the bulk of literature on IS so far focuses on the cognitive processes of interpreting with the geopolitical “bias towards prestigious forms of interpreting practice in developed countries” (Cronin 2002, p. 387), this study chooses to analyze the linguistic choices of GPC interpreting in China in relation to its the sociocultural context. The analyses of modality shifts, the meta-discourse and the typological features contribute to our understanding of the nature of interpreting as a socially situated practice and the active role of the interpreters in mediating the communication between the ST speakers and TT listeners. Fourth, although various translation shifts have been explored in previous studies on GPC interpreting (Wang 2009; Liu 2010; Hu and Tao 2012; Sun 2013), which lead to the agreement that the GPC interpreters do not always follow the principle of “faithfulness” and “neutrality” and interpreting is not a mechanical cross-lingual activity, they were mostly analyzed on the macro-level (i.e., the

138

7 Conclusion

omission or addition of certain linguistic items/devices). The relative frequency of translation shifts as compared with “zero-shifts” has not been discussed. Therefore, the present study contributes to the understanding of ST-TT relations in GPC interpreting by examining the degree and pattern of modality shifts at the semantic level with the multi-dimensional view of modality from SFG. Last but not least, the present study has practical implications in terms of the treatment of modality in interpreting. As modality markers are used frequently in human communication to realize important interpersonal functions and their use involve different communicative norms in different languages and cultures, how to translate and use them properly is an important question in interpreting practices. The patterns of modality shifts with contextualized examples in the present study suggest that in addition to the use of “stock equivalents” between the source and target languages, various strategies including “weakening”, “strengthening”, “de-obligation”, “subjectivization” and “objectivization” can be adopted according to the specific context for achieving a better communicative effect in interpreting.

7.3

Limitations and Suggestions for Future Research

With the focus on modality shifts, the present study has demonstrated how the multi-dimensional analysis of modality from SFG can be applied to corpus-based translation and interpreting studies, but due to the limited time and scope for a single research project, there is still a lot of room for extending and improving the present study for us to have a more profound understanding of modality in GPC interpreting. The first limitation lies in the directionality of the interpreting corpus. Although this study draws on a corpus of 21 transcribed GPC texts, which is much larger in size than the corpora used in previous book-length studies on GPC interpreting (Wang 2009; Liu 2010; Sun 2013), only Chinese-to-English interpreting data are available for analysis because the English-to-Chinese interpreting data were deleted during the compiling stage of the corpus CECIC for the sake of homogeneity and convenience in text alignment. Despite the limited size, it is desirable to build an additional sub-corpus containing the English-to-Chinese interpreting data and compare the patterns of modality shifts between the two directions, which may reveal important evidence for the effect of directionality on interpreting choices or the role of participation framework in GPC interpreting. The second limitation is the lack of information about the GPC interpreters’ reflections on their own linguistic behaviors concerning modality. Due to practical difficulties, I have not communicated directly with any of the 16 interpreters involved in my corpus data, so it is not clear to what extent the modality shifts revealed in the present study are the desirable results of conscious strategies instead of unintended shifts caused by the cognitive load. Although the textual and contextual analysis supports my assumption that most of these shifts are the interpreters’ active choices for achieving a better communicative effect, it is desirable

7.3 Limitations and Suggestions for Future Research

139

and meaningful to interview the interpreters themselves, which would shed more light on the nature of these modality shifts. Finally, it would also be meaningful to compare the professional GPC interpreters’ choices in the reconstruction of modality with the interpreting students’. An experiment can be conducted by selecting some ST samples from CECIC and asking a group of interpreting students to do the same “classic consecutive” interpreting. Then two comparable sub-corpora can be built and analyzed in terms of modality shifts. The corpora may be limited in size, but the findings of the comparative analysis will definitely be beneficial for the teaching and practice of interpreting since GPC interpreters represent the top-level of professional interpreting competence in China. To conclude, with the aim to conduct a descriptive-explanatory study of the reconstruction of modality in Chinese-English GPC interpreting and with the employment of the multi-dimensional analysis of modality from SFG and corpus methods, the present study has revealed interesting patterns in modality shifts in Chinese-English GPC interpreting and the close relationship between the interpreting choices and the sociocultural context. Despite a number of limitations discussed above, it has original contributions to corpus-based interpreting studies, socially-oriented interpreting studies and studies of translation shifts as well as practical implications to the teaching and practice of interpreting. It is hoped that further research projects mentioned above can be conducted in the near future so that we will achieve a more profound understanding of interpreters’ choices concerning modality.

References Cronin, Michael. 2002. The empire talks back: Orality, heteronomy and the cultural turn in interpreting studies. In The interpreting studies reader, ed. Franz Pöchhacker, and Miriam Shlesinger, 386–397. London: Routledge. Diriker, Ebru. 2004. De-/re-contextualizing conference interpreting. Amsterdam/Philadelphia: John Benjamins. Hu, Kaibao 胡开宝, and Qing Tao 陶庆. 2012. 记者招待会汉英口译句法操作规范研究 [Syntactic operational norms of press conference interpreting (Chinese-English)]. 外语教学与 研究 [Foreign Language Teaching and Research] (5): 738–750, 801. Li, Xin 李鑫, and Kaibao Hu 胡开宝. 2013. 基于语料库的记者招待会汉英口译中情态动词的 应用研究 [A Corpus-based Study of Modal Verbs in Chinese-English Government Press Conference Interpretation]. 外语电化教学 [Computer-Assisted Foreign Language Education] (3): 26–32, 74. Liu, Hui. 2010. Audience design in interpreted press conferences (Chinese-English): Face management and information management. Ph.D. Thesis, Heriot-Watt University. Monacelli, Claudia. 2005. Surviving the role: A corpus-based study of self-regulation in simultaneous interpreting as perceived through participation framework and interactional politeness. Ph.D. Thesis, Heriot-Watt University. Setton, Robin. 2011. Corpus-based interpreting studies (CIS): Overview and prospects. In Corpus-based translation studies: Research and applications, ed. Alet Kruger, Kim Wallmach, and Jeremy Munday, 33–75. London/New York: Continuum International.

140

7 Conclusion

Shlesinger, Miriam. 1998. Corpus-based interpreting studies as an offshoot of corpus-based translation studies. Meta 43 (4): 486–493. Straniero Sergio, F., and Caterina Falbo. 2012. Studying interpreting through corpora. An introduction. In Breaking ground in corpus-based interpreting studies, ed. Francesco Straniero Sergio, and Caterina Falbo, 9–52. Bern: Peter Lang. Sun, Tingting. 2013. Interpreting China: Interpreters’ mediation of Government Press Conferences in China. Beijing: Foreign Language Teaching and Research Press. Wang, Binhua 王斌华. 2009. 口译规范描写及其应用——一项基于中国总理“两会”记者招待 会交传语料的研究 [Description of norms in interpreting and its application—A study based on the corpus of consecutive interpreting in Chinese premier press conferences]. Ph.D. Thesis, Guangdong University of Foreign Studies.

Appendix A

Top 350 Wordlist of the English TTs

Rank

Word

Frequency

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

THE AND OF TO I IN THAT WE A IS THIS HAVE CHINA WILL FOR ON # YOU WITH BE ARE AS LI OUR ALSO PEOPLE HAS IT SO

9565 5058 4340 4053 3349 2965 1949 1843 1824 1817 1323 1281 1250 1157 1120 912 911 886 848 785 783 762 668 666 659 628 625 619 572 (continued)

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2

141

142

Appendix A: Top 350 Wordlist of the English TTs

(continued) Rank

Word

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

GOVERNMENT NOT CHINESE COUNTRIES FROM YEAR BY ZHU AT WHAT ALL TWO DEVELOPMENT QUESTION CAN THERE BEEN WOULD SOME VERY CHINA’S DO BETWEEN US TAIWAN ECONOMIC BUT WORK SHOULD MY YANG MORE COOPERATION WORLD THEIR OR ONE THINK NOW HONG

Frequency 529 515 508 484 471 471 457 456 405 405 397 396 389 389 385 381 380 367 364 359 347 325 324 321 314 313 311 310 300 294 294 292 288 286 285 283 282 281 278 276 (continued)

Appendix A: Top 350 Wordlist of the English TTs

143

(continued) Rank

Word

70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109

AN THEY ABOUT YOUR RELATIONS REFORM INTERNATIONAL KONG THESE WAS ITS FOREIGN OTHER COUNTRY TIME MAKE LIKE LAST PREMIER YEARS C WEN MADE I’M NEED UNITED SUCH NEW STATES NO BELIEVE STATE TAKE WELL IF GOOD ME UP IMPORTANT OVER

Frequency 271 267 266 264 261 257 255 255 245 240 236 232 228 227 223 219 218 217 215 214 213 212 209 208 208 206 201 200 194 187 183 181 181 180 179 178 178 176 174 174 (continued)

144

Appendix A: Top 350 Wordlist of the English TTs

(continued) Rank

Word

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149

ACTUALLY FINANCIAL HOW POLICY ER BECAUSE THOSE ANY KNOW WHICH ONLY ISSUE MANY EFFORTS HOPE FIRST PEOPLE’S BOTH THANK GOING VISIT OUT WHEN JAPAN MARKET PARTY STILL SAY UH LONG REGARD SIDES HE SEE CENTRAL SAID MEASURES GROWTH CONTINUE RELATIONSHIP

Frequency 166 166 163 161 157 155 155 154 154 154 153 150 150 146 146 145 145 143 143 141 141 140 140 139 139 139 139 135 134 133 133 132 131 131 130 130 129 127 126 126 (continued)

Appendix A: Top 350 Wordlist of the English TTs

145

(continued) Rank

Word

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189

INTERESTS SYSTEM JUST PRESIDENT WANT TRADE ECONOMY PAST NATIONAL FURTHER QUESTIONS THREE THROUGH MR MUST POLITICAL SITUATION THEN PEACE REPORT WHO ALREADY DEVELOPING INTO SAME MEETING THEM PRESS THAN WAY SUPPORT MATTER PEACEFUL UNDER EVEN LEVEL LAW PART AFTER COURSE

Frequency 125 124 123 123 122 121 120 119 117 115 114 112 112 111 111 110 110 109 107 107 106 104 104 100 100 99 99 98 96 96 95 94 94 94 93 93 92 92 91 90 (continued)

146

Appendix A: Top 350 Wordlist of the English TTs

(continued) Rank

Word

190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228

FUTURE STRAITS GIVE MOST PROBLEMS VIEW SOCIAL AREAS ASK BILLION HOWEVER PROBLEM SECOND COUNCIL MUTUAL MAJOR HAD INCREASE MINISTER STABILITY HERE CONCERNING DON’T DURING EACH HIGH PUT RUSSIA WERE INVESTMENT SINCE DIPLOMACY PROGRESS CRISIS CONGRESS LEADERS PLAN ROLE CURRENT

Frequency 90 90 89 89 89 89 87 86 86 85 85 85 85 84 84 83 82 82 82 82 81 80 80 80 80 80 80 80 80 79 79 78 78 77 76 76 76 76 75 (continued)

Appendix A: Top 350 Wordlist of the English TTs

147

(continued) Rank

Word

229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267

ENSURE ENTERPRISES ABLE ISSUES MC SIDE EXCHANGE MAY SECURITY POSITION BIG HIS MAINLAND PUBLIC RMB CONFERENCE FULL THEREFORE COMMON NEWS ORDER DEVELOP FACT INCLUDING MUCH USE LET SET ALWAYS TALKS NEXT REGION AGAINST OWN SESSION VARIOUS ASIA BEFORE COME

Frequency 75 75 74 74 74 74 73 73 73 72 71 71 71 71 71 69 69 69 68 68 68 67 67 67 67 67 66 66 65 65 64 64 63 63 63 63 62 62 62 (continued)

148

Appendix A: Top 350 Wordlist of the English TTs

(continued) Rank

Word

268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306

MAKING PROMOTE ANSWER GREAT IMPROVE JAPANESE EVERY TERM TOGETHER BEING CONCERNED END INTEREST WHETHER CLEAR COULD FRIENDS MENTIONED NUMBER PLACE POINT RIGHTS STRATEGIC WORKING FIVE GO RURAL TIMES DIFFERENT OPEN POWER ANOTHER BETTER CONDITIONS MAINTAIN PARTIES QUITE STRENGTHEN STRONG

Frequency 62 62 61 61 61 61 60 60 60 59 59 59 59 59 58 58 58 58 58 58 58 58 58 58 57 57 57 57 56 56 56 55 55 55 55 55 55 55 55 (continued)

Appendix A: Top 350 Wordlist of the English TTs

149

(continued) Rank

Word

307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345

TAKEN AGREEMENT ATTENTION DIFFERENCES EDUCATION FEW FORWARD DIFFICULTIES PLAY DIPLOMATIC EXAMPLE EXCHANGES TERMS WHOLE AM CORRUPTION GET JOINT KIND LARGE OPPORTUNITY SOUND VICE ACCORDING ACTIVITIES AFRICAN ASKED ENVIRONMENT MILITARY OPENING PARTICULAR POLICIES POSSIBLE PRESENT PURSUE BECOME CORRESPONDENT ENERGY FRIENDLY

Frequency 55 54 54 54 54 54 54 53 53 52 52 52 52 52 51 51 51 51 51 51 51 51 51 50 50 50 50 50 50 50 50 50 50 50 50 49 49 49 49 (continued)

150

Appendix A: Top 350 Wordlist of the English TTs

(continued) Rank

Word

346 347 348 349 350

GENERAL HELP LOT RATE RELEVANT

Frequency 49 49 49 49 49

Appendix B

Top 350 Wordlist of the Chinese STs

Rank

Word

Frequency

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

的 是 我 我们 在 中国 了 啊 和 这个 问题 # 也 李 有 不 要 这 呃 一个 呢 发展 中 朱 关系 一 对 会 想

7868 2044 1699 1547 1350 1285 1183 1155 1003 983 909 782 726 702 684 660 625 602 557 556 540 527 516 473 465 461 418 404 399 (continued)

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2

151

152

Appendix B: Top 350 Wordlist of the Chinese STs

(continued) Rank

Word

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

来 政府 地 说 谢谢 人民 都 国家 总理 上 就是 两 你 记者 经济 合作 杨 年 您 就 现在 已经 很 但是 改革 到 个 国际 好 那么 没有 将 大 一些 国 方面 为 温 多 进行

Frequency 388 385 373 356 354 349 345 341 339 332 328 321 317 315 315 313 310 300 300 299 297 291 280 274 267 265 261 258 251 248 247 243 242 241 238 235 230 225 224 222 (continued)

Appendix B: Top 350 Wordlist of the Chinese STs

153

(continued) Rank

Word

70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109

还 香港 C 工作 认为 台湾 可以 能够 世界 新 什么 解决 和平 讲 今年 美国 人 更 希望 大家 外交 所 同 镕基 社会 他们 它 应该 从 政策 把 那 去年 他 重要 金融 如果 最 请问 关于

Frequency 218 217 215 215 215 214 203 193 192 187 184 183 178 177 173 170 170 167 167 164 163 162 160 159 159 158 158 158 156 152 148 148 148 148 148 145 141 141 140 138 (continued)

154

Appendix B: Top 350 Wordlist of the Chinese STs

(continued) Rank

Word

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148

加强 做 美 共同 努力 还是 这些 过 与 继续 稳定 而且 日 时候 以 地区 三 得 还有 下 双方 当然 访问 通过 非常 日本 再 次 建设 使 市场 向 这样 去 措施 这种 企业 政治 对于

Frequency 134 133 131 130 130 127 127 126 126 125 122 121 121 121 121 118 118 117 114 114 113 112 112 111 110 110 110 109 109 109 109 109 109 108 107 106 105 105 104 (continued)

Appendix B: Top 350 Wordlist of the Chinese STs

155

(continued) Rank

Word

149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187

有关 MC 实现 利益 所以 推进 进一步 特别 需要 跟 请 之间 第二 因为 看 先生 因此 知道 包括 报告 可能 而 那个 同时 愿意 采取 两岸 回答 情况 着 事情 不能 会议 支持 给 几 问 主席 自己

Frequency 104 103 103 102 102 102 101 101 100 99 99 99 98 97 96 96 95 95 94 93 93 92 92 92 92 91 90 89 89 89 88 87 87 87 85 85 85 85 85 (continued)

156

Appendix B: Top 350 Wordlist of the Chinese STs

(continued) Rank

Word

188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226

并 领导人 们 最近 很多 积极 困难 您好 目标 取得 任何 提出 第一 友好 发生 让 时间 影响 原则 增加 的话 刚才 非洲 当中 或者 怎么 保持 财政 推动 基础 一下 能 实行 危机 一定 这次 投资 战略 比较

Frequency 84 84 84 84 83 83 83 83 82 82 82 82 81 80 79 78 78 78 78 78 77 76 75 74 74 74 73 73 73 72 72 71 70 70 70 70 69 69 68 (continued)

Appendix B: Top 350 Wordlist of the Chinese STs

157

(continued) Rank

Word

227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265

开放 领域 贸易 如何 促进 得到 等 历史 提 制度 至于 位 相信 举行 你们 又 表示 国务院 伙伴 今天 提问 以后 主要 副 最后 作为 存在 看到 朋友 用 中方 作用 今后 一点 处理 建立 许多 增长 地方

Frequency 68 68 68 68 67 67 67 67 67 67 66 65 65 64 64 64 63 63 63 63 63 63 62 61 61 61 60 60 60 60 60 60 59 59 58 58 58 58 57 (continued)

158

Appendix B: Top 350 Wordlist of the Chinese STs

(continued) Rank

Word

266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304

坚持 全面 收入 以及 种 并且 外长 维护 出现 第三 里面 钱 体制 亚洲 正在 总统 发挥 关注 提高 同志 银行 会谈 十 安全 反对 还要 另外 各国 过去 全国 不断 俄 各 计划 农民 是否 谈 完全 于

Frequency 57 57 57 57 57 56 56 56 55 55 55 55 55 55 55 55 54 54 54 54 54 53 53 52 52 52 52 51 51 51 50 50 50 50 50 50 50 50 50 (continued)

Appendix B: Top 350 Wordlist of the Chinese STs

159

(continued) Rank

Word

305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343

变化 腐败 各方 件 民主 中央 保障 必须 代表 发展中国家 核 化 环境 交流 教育 农村 群众 十分 未来 意见 注意 改善 活动 具体 目前 哪些 确实 相互 成功 其他 形势 由于 只要 重大 当前 领导 农业 为了 但

Frequency 49 49 49 49 49 49 48 48 48 48 48 48 48 48 48 48 48 48 48 48 48 47 47 47 47 47 47 47 46 46 46 46 46 46 45 45 45 45 44 (continued)

160

Appendix B: Top 350 Wordlist of the Chinese STs

(continued) Rank

Word

344 345 346 347 348 349 350

方 高 经贸 起来 唐 北京 出

Frequency 44 44 44 44 44 43 43

Appendix C

Coding Sample of Modality Shifts: I/We think1

1

See Table 3.6 for the coding scheme.

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2

161

Year

2000

2000

2004

2009

No.

1

2

3

4

PMT

PMT

PMT

PMT

Speaker Actually [[I think]] the corruption problem and also the campaign against corruption is a significant question for all the governments in the world ever since governments began to emerge in this world As for when can actual results be seen in our efforts of developing China’s west, actually [[I think]] we are already seeing some effects and some results in building infrastructure facilities in China’s west because the Chinese people, for instance, are very capable of building such projects as building railways, and highways and expressways [[I think]] an im, very important contributing factor is the fact that we have always upheld unity of the party and safeguarded social and political stability in this country Just as you said, and [[I think]] I can also give another line to the effect that after encountering all kinds of difficulties and experiencing all kinds of hardships, at the end of the day, we will see the light at the end of the tunnel

我 想 反 腐败 的 问题 是 有 政 府 以来 各国 政府 所 面临 的 一个 重大 的 问题

有些 事情 正 像 你 所 说 的, 山 重水 复 疑 无 路, 柳暗花明 又 一村

取得 这样 重大 的 成就, 一个 重要 的 原因 就是 我们 坚持 维 护 全党 的 团结 和 统一, 维护 社会 政治 的 稳定

何时 见效 呢, 我 想 基础 设施 建设 现在 已经 在 见效, 中国 人 修 铁路 公路 是 拿手好戏

TT

ST

Z

Z

我想

o

CF

SST-ZC

SST-ZC

ZSST

ZSST

SST

NA

NA

E

E

SMT

SOS

SOS

E

E

SO

(continued)

SV −

SV −

E

E

SV

162 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1989

2012

1999

1999

1998

No.

5

6

7

8

9

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker

我们 中国 的 内阁 都 是, 呃, 不 管 从 哪里 学习 回来 的, 有 什 么 经验, 都 是 中国 的 内阁

我 想 大多数 舆论 都 认为 中国 这种 做法 是 坚持 了 金融 改革 的 原则, 符合 国际 的 惯例

人民币 能否 成为 世界 货币, 这 个 问题 不 应该 由 我 来 回答, 而且 现在 还 言 之 过早

因此 , 在 网上 听到 有 “ 拍 砖 ” 的 声音 , 我 并 不 感到 奇怪 , 我 以为 这 是 正常 的 事情

TT And [[I think]] if these efforts, if we continue with this effort, it is entirely possible for us to put an end to the stagnating situation of agriculture and make some major improvement in this situation I don’t feel surprised at all that there have been some critical comments on the Internet about the performance of the government. And [[I think]] it is only natural for that to happen : The question whether the RMB will become a world currency, in my view, is not a question that should be answered by me and [[I think]] it’s too ear1y now to discuss this question And [[I think]] most of the public opinion hold the view that this approach of China is consistent with the principle of financial reform and also consistent with international practice And [[I think]] no matter where the member of the cabinet once studied or what experience the member of the cabinet used to have, this cabinet is the cabinet of China

ST

我 想 只要 按照 我们 这样 一个 路子 这样 抓 下去 , 呃, 我们 有 完全 有 可能 啊 , 结束 近 几 年 农业 徘徊 的 局面 , 使 我国 农 业 生产 达到 一个 新 的 台阶

Z

我想

Z

SST-ZC

ZSST

SST-ZC

ZSST

我以为

SST ZSST

CF 我想

SMT

NA

E

NA

E

E

SV

SO

SOS

E

SOS

E

E

(continued)

SV −

E

SV −

E

E

Appendix C: Coding Sample of Modality Shifts: I/We think 163

Year

1989

1989

1990

1991

1998

2009

1989

No.

10

11

12

13

14

15

16

(continued)

PMT

PMT

PMT

PMT

PMT

PMT

PMT

Speaker

我 想 这 就是 问题 其中 之一

因为 这 两者 是 相互 联系 的

最好 的 办法 就是 逐步 实现 “三通” , 直接 的 航运, 直 航

只要 有 人民 的 支持 , 有 党 的 支持 , 我 想 我 能够 把 这个 工 作 做 得 好 , 做好 我 想 十四大 将 如期 召开, 不 会 提前

这个, 这 几 句 话 很 清楚 地 表 示 老百姓 对 民主 过程 的 一 种 怀疑

TT In the present circumstances, we especially need a stable situation and [[I think]] such a stable situation is in the interest of Chinese people And [[I think]] such a doggerel is a clear indication of the doubts on the part of the Chinese people concerning the democratic process in China And [[I think]] that I can surely do a good job in this respect with the support of the Chinese people and the support of the party And [[I think]] the 14th Party Congresss will be held on time and will not be moved, upward And [[I think]] the best way to do that is to gradually establish the direct air shipping, and postal and trade links between the two sides. And [[I think]] these 2 aspects are interrelated And [[I think]] this influx of job-seekers from the rural areas is one of them

ST

而且 中国 现在 非常 需要 有 一 个 稳定 的 局势, 这 是 符合 全 体 中国 人民 的 利益 的

CF

SST

ZSST

我想

我想

Z

ZSST

SST-ZC

SST-ZC

ZSST

我想

Z

SST-ZC

SST-ZC

Z

Z

SMT

E

NA

NA

E

E

NA

NA

SV

SO

E

SOS

SOS

E

E

SOS

SOS

(continued)

SV − E

SV −

E

E

SV −

SV −

164 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2002

2007

1989

1998

No.

17

18

19

20

(continued)

PMT

PMT

PMT

PMT

Speaker

至于 参加 WTO 的 问题, 我 认 为 最好 由 李 岚清 副 总理 来 回答

克服 这些 困难 的 关键 在于 全 国 上下 同心同德

台海 两岸 和平 发展 是 大势所 趋, 是 任何人 … 无法 改变 的

TT And [[I think]] this is a subject of discussion among the elites of Hong Kong and we are also having a discussion with Hong Kong SAR officials and people from Hong Kong about this subject And [[I think]] to develop peaceful relations and facilitate the growth on both sides of the Taiwan Straits is a historical trend and no one can block. And this trend can not be changed whether you like it or not And [[I think]] to overcome these difficulties, it is necessary for people across the nation to unite [SOUND LOST] As for the question of China’s accession into the WTO, and [[I think]] Vice Premier Li Lanqing is in the best position to answer that

ST

那么 所有 的 香港 的 精英 都 在 为 这个 问题 在 进行 探讨 . 我们 也 愿意 和 香港特区政府, 和 香港 人民 一 起来 探讨 这个 问题

CF

我认为

Z

Z

Z

SST

ZSST

SST-ZC

SST-ZC

SST-ZC

SMT

E

NA

NA

NA

SV

E

SO

E

SOS

SOS

SOS

(continued)

SV −

SV −

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 165

Year

2007

2001

1991

No.

21

22

23

(continued)

PMT

PMT

PMT

Speaker

我 想 还有, 会 有 别的 记者 还 要 提 这 方面 的 问题, 所以 我 就 借此机会 多 讲 两 句

可见 中国 人民 啊 对 我们 啊 还是 有 信心 的

TT I think democratic system, like whatever truth, needs to be tested by practice and what actually happens. Taking the democratic systems in whatever country or region for example, whether the system is good or bad, [[I think]] we can leave the answer to the reality and to the real practice [[I think]] this is reason, this is, uh, proof enough to show that the Chinese public have confidence in the government Because [[I think]] that maybe there are some other correspondents who will ask the question about this or to this effect, so I would like to take this opportunity to say a few more points

ST

民主 制度 啊, 同 任何 真理 一 样, 同样 要 接受 实践 的 考验, 任何 地区 和 国家, 民主 … 制度 的 状 况, 优劣 , 都 要 以 实践 为 标准

CF

我想

Z

Z

SST

ZSST

SST-ZC

SST-ZC

SMT

E

NA

NA

SV

E

SO

E

SOS

SOS

(continued)

SV −

SV −

166 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2013

1998

1991

2001

No.

24

25

26

27

(continued)

PMT

PMT

PMT

PMT

Speaker In 2011, I made a visit to Hong Kong. During the trip, I announced the central government’s policies for supporting Hong Kong’s prosperity, and deepening cooperation between the mainland and Hong Kong. You described these policies as a large gift box, [[I think]] that it’s a pretty good analogy But [[I think]] any activities conducted by any organization in Hong Kong must be, must be in full compliance with the Basic Law for Hong Kong SAR and also relevant laws and regulations of the Hong Kong Special Administrative Region government But [[I think]] because you are, come from Singapore, so I would break the rule and answer all of your three questions But [[I think]] it will take a fairly long period of time

不过 我 想 你 是 从 台湾 来 的 记者 吧, 新加坡 来 的 记者 . 那 好 吧, 我 就 回, 破 一 次 例, 回答 三 个 问 题 . 怎么样, 可以 吗? 但是 我 想, 总是 会 要 有 一个 相当 长 的 时间 吧

但是 我 想, 香港 任何 组织 的 活动 都 必须 符合 基本法, 符合 香港特区 的 政府 的 法规

TT

ST

我 记得 前年 到 香港 访问, 呃, 宣布 了 中央政府 支持 香港 繁 荣, 推动 两 地 啊 深化 合作 的 一 系列 政策, 你 把 它 比喻 为 “大 礼 包” 倒 是 很 形象

ZSST

ZSST

我想

ZSST

我想

我想

SST-ZC

SST

Z

CF

E

E

E

NA

SMT

E

E

E

E

E

E

SOS

SO

(continued)

SV −

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 167

Year

2000

2002

2000

1991

2009

No.

28

29

30

31

32

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker But [[I think]] it is dramatically different from the point, the main point that I was talking about on Taiwan, so it is as different as earth is with heaven But [[I think]], nevertheless, Hong Kong still enjoys very advantageous position And in terms of the market capacity in China… of China, [[I think]] actually the Chinese telecom market is the second largest in the world But I,but I can assure you that no matter what would happen in the future, the new leadership in China, [[I think]], will remain stable and so will be the political situation in China As far as the possibility and importance of this 8% growth rate goal is concerned, [[I think]] I have already made it clear in my report

关于 实现 8% 的 左右 这个 目 标 的 可能性, 我 在 报告 中 已 经 阐述 了

啊, 但是 一旦, 不论 出现 什么 情况, 我 想 现在 的 领导 集体 是 稳定 的, 中国 的 政治 形势 也 是 稳定 的

我 想 中国 的 移动 电话 的 容 量 已经 是 世界 第二

但是, 我 始终 认为, 香港 具有 非常 优势 的 地位

TT

ST

但是 跟 我 的 讲话 风 马 牛

SST-ZC

ZSST

我想

Z

ZSST

ZSST

SST-ZC

SST

我想

我认为

Z

CF

NA

E

E

E

NA

SMT

SOS

E

E

E

SOS

SO

(continued)

SV −

E

E

E

SV −

SV

168 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2007

2011

2011

2007

No.

33

34

35

36

(continued)

PMT

PMT

PMT

PMT

Speaker

你 说 我 的 文章 讲 100年 不 变 , 说 是 100年 不要 民主 , 这 是 你 误解 了

你 谈到 了 改善 两 国 国民 的 感情 的 事, 我 认为 这 事 很 重 要

我们 愿意 在 这些 领域 呢 加强 沟通, 啊, 这 既 是 “金 砖 国家” 的 人民 的 愿望, 我 也, 我 想 也 是 世界 人民 的, 啊 期待

TT As far as Dalai Lama is concerned, [[I think]] we need to observe what he says, but more importantly we also need to watch what he does therefore to enhance their exchanges and dialogue is not only the shared wish of people of the BRICS countries, [[I think]] it also represents the expectation of people of the world You mentioned how to nurture friendly sentiments between peoples of the two countries. [[I think]] this is indeed a very important thing You just now mentioned that by writing that article, particularly you quoted some of my writings, do I mean that the system in this country will remain unchanged for 100 years, and in 100 years time democracy is not needed in this country? [[I think]] that is a misunderstanding of what I have written

ST

对 达赖喇嘛, 我们 不仅 要 看 他 说 些 什么, 更 要 看 他 做 些 什么

CF

SST

SST-ZC

ZSST

我认为

Z

ZSST

ZSST

我想

我想

SMT

NA

E

E

E

SV

SO

SOS

E

E

E

(continued)

SV −

E

E

E

Appendix C: Coding Sample of Modality Shifts: I/We think 169

Year

2007

1999

1991

No.

37

38

39

(continued)

PMT

PMT

PMT

Speaker

这个 … 宣判 王 丹 呢 是 另外 一 回 事情, 因 为 他 触犯 了 中国 的 刑律, 而 且 对 王 丹 的 审判, 我, 呃 我 刚才 讲 了, 这 是 法院 的 事情

我 觉得 那些 债权 银行, 呃, 以 某些 金融 机构 对于 这个 问题 的 估计 太 悲观 了, 就 认为 中 国 已经 发生 了 金融 危机, 没 有 支付 能力, 不 讲 信用, 过分 了

TT [[I think]] to address this problem, first and foremost, we need to start our efforts in the institutions and systems I think some creditor banks and financial institutions have been too pessimistic in their estimation and evaluation of this incident uh in believing that China has, is experiencing a financial crisis and that China has no payment capabilities and that China is no longer creditworthy. [[I think]] that is going too far As for the, these, such specific cases, for example, the trial of Wang Dan, [[I think]] that is a completely different matter, because he, this person, violated the Chinese criminal law and also as I mentioned just now, the Chinese judiciary is independent

ST

解决 这个 问题, 首先, 还 得 从 制度 上 入手

CF

Z

Z

Z

SST

SST-ZC

SST-ZC

SST-ZC

SMT

NA

NA

NA

SV

SO

SOS

SOS

SOS

(continued)

SV −

SV −

SV −

170 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2007

1999

1993

1991

1990

No.

40

41

42

43

44

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker As for the major problem plaguing, plaguing China’s economy, [[I think]] they are as follows. The unstable, and unbalanced and uncoordinated and unsustainable development, and these are structural tensions in our economy Indeed, China Telecom is reducing the prices. But I don’t think that the reduction is far enough. [[I think]] the price should be reduced continuously by several times [[I think]] this is an important question [[I think]] this problem is also a very important one in today’s China’s economic life [[I think]] these are entirely… different matters and that’s an obvious misunderstanding on the part of some comrades

经济 , 中国 经济 存在 的 最 大 问题 依然 是 不 稳定, 不 平衡, 不 协 调, 不可 持续 的 结构性 的 问 题

所以 这 显然 是 一个 误解

我 想, 这个 问题 就 涉及 到 一 个 很 重大 的 问题 啊 这 在 中国 的 当前 经济 里 面, 啊 是 个 很 重要 的 问题

是 的, 中国 电讯 正在 降价, 降 得 还 不够, 还要 成倍 的 降

TT

ST

CF

Z

Z

我想

Z

Z

SST

SST-ZC

SST-ZC

ZSST

SST-ZC

SST-ZC

SMT

NA

NA

E

NA

NA

SV

SO

SOS

SOS

E

SOS

SOS

(continued)

SV −

SV −

E

SV −

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 171

Year

2007

2013

2013

1999

2009

2007

No.

45

46

47

48

49

50

(continued)

PMT

PMT

PMT

CP

PMT

PMT

Speaker The first and foremost [[I think]] to resolve the issues concerning people’s wellbeing, we need to have systematic guarantee [[I think]], as long as we maintain “the one China” principle and the bond of compatriots between the two sides, then there will be great space and potential for advancing cross straits relations [[I think]] we are running overtime, so we have just one last question On the question of GITIC, [[I think]] this is also a question of interest to many of the press correspondents here [[I think]] we need to approach this issue of an 8% goal for the economic growth from the following three perspectives As for the question that whether China can maintain its current growth momentum of steady and fast economic growth, [[I think]] the conditions are available. The conditions are there

我 曾经 说 过 , 中国 已经 连续 多年 保持 平稳 较 高 的 速度, 还 能 不 能够 继续 保持 下去? 我 说 条件 是 具备 的

我 想, 对于 8% 左右 的 经济 发 展 目标, 可以 从 三 个 方面 来 认识

关于 广信 事件, 我 想 这个 问 题, 你 这个 问题 也 是 大家 所 关心 的, 因此 我 就 多 说 两 句

嗯, 时间 关系, 是 不 是 再 提 一 个 问题?

我 想, 只要 维护 一个 中国, 维 护 同胞 之 情, 两岸 啊 关系 发 展 的 空间 和 潜力 巨大

TT

ST

解决 民生 问题 , 第一 要 有 制 度 的 保障

我说

我想

我想

Z

我想

Z

CF

SST-C

ZSST

ZSST

SST-ZC

ZSST

SST-ZC

SST

E

E

E

NA

E

NA

SMT

SOS

E

E

SOS

E

SOS

SO

(continued)

SV −

E

E

SV −

E

SV −

SV

172 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2013

2009

2002

2011

No.

51

52

53

54

(continued)

PMT

PMT

PMT

PMT

Speaker [[I think]] we should not make groundless accusations against each other and spend more time doing practical things that will contribute to cyber security [[I think]] we need to step up our cooperation among the member states in energy, resources, transport and other major infrastructure projects and in agriculture and high-tech fields Although this problem is giving me headache, [[I think]] the problem of the slow increase of farmers’ incomce will resolve in the end China and the EU are two major economies in the world, and we are each at a crucial stage of, in our respective development. We should pursue cooperation through mutual help. [[I think]] this is a good thing for both sides to do. So why not go ahead with it?

那么 中国 跟 欧洲 啊, 是 两 个 世界 上 非常 重要 的, 啊, 经济 体, 又 分别 处于 各自 发展 的 关键 时期, 我们 相互 帮助, 啊, 这个 合作 共 赢, 这 是 何乐而 不为

想 虽然 头痛, 它 还是 会 得到 解决 的

我 想 我们 可以 在 能源, 资源, 交通 等 基础 设施, 农业, 高新 技术 等 方面 加强 合作

TT

ST

我 想 我们 还 是, 呃, 少 做 些 啊 没有 根据 的 相互 指责, 多 做 维护 网络 安全 的 实 事儿

SST-ZC

ZSST

我想

Z

ZSST

ZSST

SST

我想

我想

CF

NA

E

E

E

SMT

SOS

E

E

E

SO

(continued)

SV −

E

E

E

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 173

Year

2004

1999

1989

2000

2013

No.

55

56

57

58

59

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker And so here [[I think]] I would select a couple of questions to address here as a token of my respect Just now I … cited two underestimations, but here [[I think]] there is an overestimation And if some regions and cities along the coast where the economy, finance and commerce are fairly developed want to experiment with or draw on some of the experiences in Konghong or to reach the same level of prosperity in Hongkong, [[I think]] this is understandable However [[I think]] maybe I should emphasize and use the word “strict” more times than that There is some talk about urbanization and some concerns expressed about it. [[I think]] people generally want us to make steady progress in urbanization

这些 议论 、 担心, 我 想 呢 都 是 为 使 城镇 化 的 路 啊 走 得 更好

讲 得 还 不够

如果 说 某些 地区 某些 城市 从 经济 , 从 商业 , 从 金融 方面 借 鉴 香港 的 某些 经验 或者 是 达到 香港 的 一些 繁荣 的 程 度, 如果 指 这 方面 来讲 , 那 将 是 可以 理解 的

我 刚才 讲 了, 美国 有 两 个 过 低 估计, 我 看 这里 又 来 了 一 个 过 高 估计

TT

ST

我 还是 想 借 这 机会 回答 群 众 的 一个 问题, 以 表示 我 对 大家 的 敬意

我想

Z



我看

我想=I think I would

CF

ZSST

SST-ZC

SST-C

ZSST

SST-C

SST

E

NA

E

E

E

SMT

E

E

SOS

SOS

E

SOS

SO

(continued)

SV −

SV −

E

SV −

SV

174 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1999

1989

2007

No.

60

61

62

(continued)

PMT

PMT

PMT

Speaker President Jiang Zemin paid a state visit to Japan and the two sides agreed to take history as a guide and to work to establish a friendly relationship and co-operation for peace and development between China and Japan, [[I think]] this serves the fundamental interests of both China and Japan You know [[I think]] we would not have problems if we are not in such a, we were no, if we were not faced with the current economic difficulties and if various localities had not put up a lot of unnecessary buildings, hotels and auditoriums On the Dalai Lama, [[I think]] our policy has been very clear-cut and consistent. That is, as long as the Dalai Lama recognizes that Tibet is an inalienable part of China’s territory and that Taiwan is also an inalienable part of China’s territory, as long as the Dalai Lama gives up his efforts to split

我们 对 达赖喇嘛 的 政策 是 明 确 的 和 一贯 的, 就是说, 啊, 只 要 达赖喇嘛 … 承认 … 西藏 是 中国 不可分 割 的 领土 的 一部分, 台湾 是 中国 不可分割 领土 的 一部分 , 放弃 分裂 活动 , 那么 我们 就

如果 我们 不 是 遇到 现在 的 这个 经济 上 的 这些 困难, 如 果 不 像, 各地 不 是 像, 以前 那样, 像 前 几 年 那样 盖 了 那么 多 的 某 些 不 必要 的 楼堂馆所, 我 想 我们 支付 亚运会 这笔 钱 应该 说 是 没有 什么 问题 的

TT

ST

江 泽民 主席 访问 了 日本, 双 方 以 史 为鉴 呢, 共同, 这个, 同, 同 意 建立 这种 致力 于 和平 与 发展 的 友好 合作 关系, 我 想 这 是 得到, 这 是 符合 中 日 两 国 人民 的 根本 利益 的

Z

我想

我想

CF

SST-ZC

ZSST

ZSST

SST

NA

E

E

SMT

SOS

E

E

SO

(continued)

SV −

E

E

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 175

Year

2007

2003

1991

No.

63

64

65

(continued)

PMT

PMT

PMT

Speaker the country, we will be in a position and we are willing to have consultations and dialogues with the Dalai Lama on his personal future And I can tell you in a very clear manner that, in the final analysis, by developing social democracy, we mean that we want to make people the masters, masters of their own house. To do this, …[[I think]] we need to further develop the socialist system and continue our efforts to develop the political and democratic system as well With regard to the basic law article 23 legislation, [[I think]] it is common practice for all countries to make law to prohibit criminal activities which endanger national security and unity [[I think]] every people who are not biased and who show respect for the objective facts can reach a common point of understanding with us on this issue

只要 是 没有 偏见 的 人 、 尊 重 客观 事实 的 人, 我 想 我们 都 应该 在 这个 问题 上 找到 共同点

[SOUND LOST ] 基本法 第 23 条 立法 的 问题, 我 认为, 立法 禁止 危害 国家 安全 和 统一 的 犯罪 行径, 这 是 世界 各国 刑 事 立法 的 通例

我 可以 明确 地 说, 社会主义 民主 … 归根 结底 是 , 让 人民 当家作主,… 就是 要 实行 依法 治国, 完善 法制 , 建设 法治 国家

TT

ST

可以 就 他 个人 的 前途 问题 同 他 进行 协商 和 对话

ZSST

ZSST

我认为

我想

SST-ZC

SST

Z

CF

E

E

NA

SMT

E

E

E

E

SOS

SO

(continued)

SV −

SV

176 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1991

1991

1989

1991

No.

66

67

68

69

(continued)

PMT

PMT

PMT

PMT

Speaker [[I think]] if you have not forgotten, then you can certainly recollect and remember that last year, around October, round August and September, the three factions of the resistance forces led by Prince Norodom Sihanouk, including Khmer Rouge, already accepted the document worked out by the five permanent members of the United Nations Security Council on the settlement of the Cambodian problem [[I think]] indeed there are some reasons behind what you have said [[I think]] it has something to do with our efforts to scale back the capital construction [[I think]] it is known to all that China and the Soviet Union are close neighbors sharing a border of 7000 km

大家 都 知道, 中国 和 苏联 是 有着 7000 公里 边界 的 邻国

我 想 这 与 我们 压缩 基本建设 规模 有 一定 的 关系

我 想 … 您 刚 才 讲 的 是 有 一些 道理 的

TT

ST

我 想 如果 你 不 是 健忘 的话, 早 在 去年 的 8, 9月份 的 时候, 这个 抵抗 力 方, 抵 抗 力量, 西哈努克 领导 的 抵抗 力量 三 方, 包括 红 色 高 棉 在内, 都 已经 接受 了 联合国安理会 五 国 的 关于 解 决 柬埔寨 问题 的 这个 文件

CF

Z

我想

我想

Z

SST

SST-ZC

ZSST

ZSST

SST-ZC

SMT

NA

E

E

NA

SV

SO

SOS

E

E

SOS

(continued)

SV −

E

E

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 177

Year

1991

1990

1989

2013

No.

70

71

72

73

(continued)

PMT

PMT

PMT

PMT

Speaker

这 两 种 < repetition > 疑问, 疑 问 哪 , 我 认为 都 可以 打消

我 想 我们 这次 中 苏 之间 的 最 高级 会晤 将 在 两 国 关系 正常化 方面 取得 最 大 的 成效

我们 的 新闻 报道 相反 是 比较 克制 的

TT [I think]] it is known to all that this 10-year Programme and the Eighth 5-year Plan were formulated by the Communist Party of China and the State Council of the People’s Republic of China. And the proposal for the formulation of the programme and the plan was adopted at the 7th Plenary Session of the 13th Communist Party Central Committee [[I think]] that in reporting the news in Taiwan, our news agencies have been quite restrained [[I think]] the biggest significance for this upcoming summit meeting would be in the normalization of the relation, state-relations between our two countries [[I think]] these two concerns are really unnecessary

ST

呃, 大家 都 知道 这 一 次 的 十 年 规划 和 八五 计划 纲要 是 由, 呃 中共中央 和 国务院 制定 的, 呃 七 中 全会 啊 通过 了 建 议

CF

SST

ZSST

ZSST

我想

我认为

SST-ZC

SST-ZC

Z

Z

SMT

E

E

NA

NA

SV

E

E

SO

E

E

SOS

SOS

(continued)

SV −

SV −

178 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1991

1989

1991

2004

2007

No.

74

75

76

77

78

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker [[I think]] these are issues, two issues, of entirely different natures, because for the disputes between states we have always called for a solution through peaceful consultations [[I think]] this is a very important question, because for our nation and national economy to develop, it is impossible for us to proceed smoothly without the enthusiastic support of the workers and farmers [[I think]] you are right in saying that right now there’s much talk in many countries in the world and also in the mass media about the establishment of new world order [[I think]] that is also the common will of all the Chinese people, including our Taiwan compatriots As for your second question about the abduction of Japanese nationals, [[I think]] we have all along expressed our sympathy and understanding over this matter

于 你 提 的 第二 个 问题 , 我们 已经 多次 表示 , 对于 日本 公 民 遭受 绑架 一 事 表示 同情 和 理解

你 说 得 不错, 现在 呃 … 许多 国家 啊 … 和 世界 的 舆论界 都 在 谈论 建立 这个 国 际, 啊 或者 是 世界 新 秩序 的 问题 这 是 包括 台湾 人民 在内 的 整个 中华 儿女 的 共同 愿望

恩, 你 提 这个 问题 很 重要, 我 们 国家 要 发展, 经济 要 发展 , 嗯, 没有 工人 农民 的, 他们 的 积极性, 这个, 要 得到 顺利 的 发展 是 不 可能 的

TT

ST

我 想 这 是 完全 不同 的 两 个 性质 的 问题 . 国 与 国 之间 的 争端 我们 一贯 主张 通过 和平 协商 的 办法 来 加以 解决

SST-ZC

SST-ZC

Z

Z

SST-ZC

SST-ZC

ZSST

SST

Z

Z

我想

CF

NA

NA

NA

NA

E

SMT

SOS

SOS

SOS

SOS

E

SO

(continued)

SV −

SV −

SV −

SV −

E

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 179

Year

1990

2007

1993

1989

No.

79

80

81

82

(continued)

PMT

PMT

PMT

PMT

Speaker [[I think]] about the June 4th Event, Comrade Chen Xitong who’s the State counselor and Mayor of Beijing gave a very detailed report and account about the June 4th Event in his report released last year at a standing committee meeting of the National People’s Congress [[I think]] the core of your question is that what is socialist democracy As for how the situation in Russia will develop, now [[I think]] it would be difficult for anyone in the world to make a, an accurate prediction As for what measures will take to attain our objective, [[I think]] Premier Li Peng has already given a detailed explanation in his report

那么 至于 说 采取 什么 措施 能 够 达到 这 一点, 我 看 李 鹏 总 理 在 政府 工作 报告 里面 已经 完全 都 讲 了

至于 … 那里 的 情况 会 向 什么 方向 发展, 现在 我 看, 呃, 世界 上 很 难 有 人 能够 做出 非常 准确 的 预测

于 你 提到 一个 核心 问题 , 就 是 社会主义 民主 是 什么?

TT

ST

我们 国务委员 兼 北京市 市长 陈 希同 同志 已经 在 去年 人民, 人大 常委会 的 会议 上 就 六四 事件 做 了 非常 详尽 的 报告

我看

我看

Z

Z

CF

ZSST

ZSST

SST-ZC

SST-ZC

SST

E

E

NA

NA

SMT

E

E

E

E

SOS

SOS

SO

(continued)

SV −

SV −

SV

180 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2007

2001

2011

1998

No.

83

84

85

86

(continued)

PMT

PMT

PMT

PMT

Speaker

我 觉得 有 三 个 积极 的 “趋向” 是 值得 关注 的 但是 , 我 觉得 也 不能 忽视 另 外 一点 , 作为 世界贸易组织 , 那么 缺乏 中国 这样 一个, 我们 现在 贸易, 在 贸易 国 的 排序 上 已经 达到 第十 位 了, 没有 中国 的 参加 , 我 看 它 也 很 难 发挥 作为 一个 世界贸易组织 的 应有 的 作用

那 这 不 是 说明 中国 有 言论 自由 吗 ?

TT From now on, [[I think]] we need to ensure that all the examination and approval conducted by the government, particularly when it comes, when it involves the interests of the ordinary people, need to be done in a open, fair, just and transparent manner [[I think]] this is evidence enough that people in China enjoy freedom of speech [[I think]] there are three positive trends that merit our attention However, on the other hand, the WTO as a world trade organization, [[I think]], would be hard to play its due role as a world trade organization without the participation of China, already the 10th largest trading nation in the world

ST

今后, 啊, 凡 属 审批 事项, 特别 是 涉及 人民 群众 利益 的, 都 要 实行 公开, 公正 和 透明

CF

SST

SST-C SST-C

我觉得

SST-ZC

SST-ZC

我觉得

Z

Z

SMT

E

E

NA

NA

SV

SO

E

E

SOS

SOS

(continued)

SV + SV +

SV −

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 181

Year

2013

2009

1990

2009

No.

87

88

89

90

(continued)

PMT

PMT

PMT

PMT

Speaker So let me just say a few more words. Recently I’ve been following very closely news reports about China, er, including the issues discussed by foreign media organizations. [[I think]] there’re two main concerns. First, whether the Chinese economy will continue to grow sustainably? And secondly, whether a stronger China will become more assertive and even a hegemony? [[I think]] this is a conscience and responsibility that every journalist should have here, this is also the expectation of people far and wide [[I think]] it is only a prediction of yourself and you are not a Chinese To attain our goal in this plan, [[I think]], first and foremost, we need to have very strong confidence

实现 这个 计划, 我 依然 认为, 首要 的 还是 要 坚定 信心

恐怕 只 能够 代表 你 自己 , 你 也 不 是 中国 人

我 想 这 应该 是 每 位 记者 的 良知 和 责任, 也 是 人们 的 期 望

TT

ST

呃, 这 一 段 时间, 我 一直 在 看 有关 中国 问题 的 报道, 当然 主要 是 国际 舆论 所 关注 的 一些 问题, 大概 主要 是 两 方 面: 一个 是 担心 中国 经济 能 不 能够 持续 发展 ; 还有 一个 呢 是 担心 啊 中国 发展 起来 会 不 会 恃 强 称霸

CF

我认为

恐怕

我想

Z

SST

ZSST

SST-C

ZSST

SST-ZC

SMT

E

E

E

NA

SV

E

SO

E

SOS

E

SOS

(continued)

SV +

E

SV −

182 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2013

2007

1991

1998

No.

91

92

93

94

(continued)

PMT

PMT

PMT

PMT

Speaker

当然 , 在 政府 的 机构 进行 民 主 的 选举 , 包括 选举 总理 和 主席 , 这 是 一个 政治 体制 改 革 的 问题

我 想, 我 不 回答, 回 避 这个 问题 不好, 我 就 借此机会 讲 一 讲

中国 购买 美元 资产 是 互利 的, 中国 组建 外汇 储备 的 投 资 公司, 不 会 影响 美元 资产

TT You asked whether this is an ideal plan. [[I think]], to succeed in doing anything, one has to strike a proper balance between ideal and reality [[I think]] China buys US dollar denominated assets on the basis of mutual benefit. And I can assure you that by instituting such a foreign exchange reserve investment company, it will not have any adverse impact on the US dollar denominated asset [[I think]] it is not good for me to deviate from this question, to parry out this question. So I would like to take this opportunity to make a few more remarks Er, with regard to the democratic elections for the governmental officials, such as president and premier. [[I think]] that’s a question involving political restructuring

ST

你 谈 到 理想, 凡事 我 想 啊 要 做成 的 话, 总是 要 在 理想 和 现实 之间 啊 做出 可能 的 选择

CF

当然

我想

Z

我想

SST

SST-C

ZSST

SST-ZC

ZSST

SMT

E

E

NA

E

SV

SO

SOS

E

SOS

E

(continued)

SV −

E

SV −

E

Appendix C: Coding Sample of Modality Shifts: I/We think 183

Year

1990

1990

2004

2009

2009

No.

95

96

97

98

99

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker

关于 第一 个 问题, 其实 我 在 英国 回答 《 金融 时报 》 记者 问题 时 已经 谈到 了, 但是 我 还 想 说明 一下

这种 接 谈 还 可以 继续 进行 下去, 关键 在于 达赖喇嘛 要 有 诚意, 使 接 谈 取得 实质性 的 成果

根本 的 原因 在 机制, 在 体制

那么 至于 说, 是不是 靠 这个 药方 子 就 解决 了 市场 疲软 问题, 现在 没有 试验, 刚刚 开 始

TT In my view, as for our prescription, [[I think]] it is both effective and prudent As for whether we can solve the problem by relying on this new prescription [[I think]], so we have just started our work in this regard [[I think]] the fundamental reasons for this lie in institutions and mechanisms But for such contacts and the consultations to make progress, [[I think]] what is important is for the Dalai Lama to have sincerity, otherwise no substantive results can be made. Such contacts and consultations may continue As for your first question, [[I think]] I’ve already given an answer to a question raised by FT in the United Kingdom. And still, I would like to use this opportunity to make further clarification

ST

我 的 看法, 这个 药方 子, 第一 是 有效 的 , 第二, 也 是 谨慎 的

CF

其实

Z

Z

Z

我的看 法

SST

SST-C

SST-ZC

SST-ZC

SST-ZC

ZSST

SMT

NA

NA

NA

NA

E

SV

SO

SOS

SOS

SOS

SOS

E

(continued)

SV −

SV −

SV −

SV −

E

184 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2000

1989

2008

2008

2007

No.

100

101

102

103

104

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker [[I think]] the reason is that there are always some people in a certain country in the world who are always there standing against China [[I think]] this, on the hand reflects that there are still problems in terms of the decision making. Secondly, it also means that the question of separation of power between the government and the party as put set forth at the thirteenth Party Congress has not been handled very well [[I think]] China’s relations with Singapore are good [[I think]] we should focus our efforts on the following priority areas. First, high level exchanges and contact mechanism. We should maintain the momentum of high level exchange of visits and contact You’ve also raised the question on the foreign exchange reserve of China, and how to use the huge amount of the foreign exchange reserve. [[I think]] this is a very big question and it’s also a major problem

你 谈到 中国 的 外汇 储备 如何 使用, 这 确实 是 我们 面临 的 一个 大 的 问题

我 看 我们 跟 新加坡 的 关系 就 发展 得 不错 嘛 我 想 呢, 呃, 双方 努力 的 重点 恐怕 是 如下 几点: 一个 就是 要 进一步 加强 高层 接触 交流 的 机制, 保持 双方 高层 会晤 啊 互访 的 这个 势头

这 一方面 说明 咱们 的 决策 有 问题 , 另外 就是 还 很, 很 明显 的 说明 这个 十三大 提出 的 党 政 分开 这个 问题 还 没 得到 好好 的 解决

TT

ST

那 就是 因为 在 美, 在 某 一个 国家, 有 一些 从来 就是 反对 中国 的 人士

CF

SST

ZSST ZSST

SST-C

我想

确实

SST-ZC

SST-ZC

我看

Z

Z

SMT

NA

E

E

NA

NA

SO

SOS

E

E

SOS

SOS

(continued)

SV −

E

E

SV −

SV −

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 185

Year

1989

2001

2001

2009

No.

105

106

107

108

(continued)

PMT

PMT

PMT

PMT

Speaker

20 国 集团 金融 峰会 一定 要 把 关心 和 帮助 发展中国家, 特 别 是 最 不 发达国家, 作为 重 要 议题

至于 我 什么 时候 退休 , 退休 以后 干什么 , 那么 我 可以 说 , 我 在 1998 年 的 时候 我 就 讲 了, 我 将 勇往直前, 义无反顾, 鞠躬尽瘁, 死而后已

我 想 这些 政府 机构 的 改革 , 对于 提高 政府 的 效率 和 职能 的 转变 , 是 有 很 大 的 好处 的

TT As for which department of ministry is specifically responsible, [[I think]] as a large and comprehensive body, the state education commission, the state education commission has responsibility for it And as a result, [[I think]] the governments have improved their work efficiency and they have improved their institutional functions As for when I’m going to retire, what I’m going to do after retirement, [[I think]] I already made the answer very clear in 1998 when I said I would press ahead without any hesitation and devote all that I have to the country, to the people till the last minute of my life [[I think]] at the G20 Financial Summit, we must see to it that how to share concern, show concern for the developing countries and help the developing countries, the least developed ones in particular become an important topic on the agenda of the meeting

ST

至于 具体 负责 的 部门, 我 想 国家教委 是 个 大 的 综合 教育 部门 , 我 想 它, 负有, 应该 负有 这 方面 的 责任

CF

Z

我可以 说

我想

我想

SST

SST-ZC

SST-C

ZSST

ZSST

SMT

NA

E

E

E

SV

SO

SOS

SOS

E

E

(continued)

SV −

E

E

E

186 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2007

2009

1991

2002

No.

109

110

111

112

(continued)

PMT

PMT

PMT

PMT

Speaker [[I think]] these words are not only addressing the government, ordinary government officials, and also addressing those high-ranking government officials [[I think]] if we take a closer look at what we put down in the report, we can see that concerning the signing of the agreement and establishment of the mechanisms, we need to follow three alignments But on that occasion, back then, because the National People’s Congress was not in session, so it was inconvenient for me to reveal this news to…to him, to the outside. So [[I think]] I can have your understanding at this point So [[I think]] in the same fashion, there is no need to make the problems in some individual branches of Chinese banks into sensational stories

同样, 中国 的 个别 的 分支 机 构 发生 一些 违法 的 事件, 也 值 不得 大家 渲 演, 渲染

在 当时 那个 情况 下, 因为 还 没有 召开 人大, 所以 我 当然 不 便于 透露, 这 一点 大家 是 会 谅解 的

我 讲 的 这个 协议 和 这个 机 制 如果 深 一 步 来讲, 它 应该 包括 “三 个 适应”

TT

ST

我 想 这话 不仅 是 对 政府 工 作 人员 说 的, 更是 对 领导 干 部说的

Z

Z

Z

我想

CF

SST-ZC

SST-ZC

SST-ZC

ZSST

SST

NA

NA

NA

E

SMT

SOS

SOS

SOS

E

SO

(continued)

SV −

SV −

SV −

E

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 187

Year

1991

2002

2002

1990

1999

1999

No.

113

114

115

116

117

118

(continued)

PMT

PMT

PMT

PMT

PMT

PMT

Speaker

所以, 所谓 中国 盗窃 美国 的 军事 机密 的 问题 啊, 可以 认 为 是 一 种 天方夜谭

我 想 这样 文章 的 出现 也 反 映 着 现在 美国 的 那 一 股 潮 流, 实际上 是 由于 美国 内部 的 斗争 把 中 美 关系 作为 牺 牲品 了

而 这 件 事情 我们 不 能够 离 开 当时 的 历史 条件 来 看待 这个 问题

我 相信, 在 一个 时期 以后, 这 种 贫富 差距 扩大 的 现象 最终 会 得到 解决 它 将来 的 地位 是 不可限量 的

TT So [[I think]] maybe some day in the near future the two sides can find a common point and find a solution So [[I think]] that this problem will be tackled after a certain period of time So [[I think]] that Hong Kong would have no limit in further developing its role in the future So [[I think]] that you can’t regard this question by divorcing yourself from the pervading circumstances So [[I think]] the emergence of such kind of article has reflected that the anti-China wave that is emerged in the United States actually is a reflection that the China-US relations have been victimized by the internal struggle in the United States So [[I think]] the question of so-called China’s theft of military secrecy from the United States is sheer, is really a fallacy

ST

我 想 也许 在 不久 的 将来 我 们 可以 找到 一个 共同点, 找到 解决 的 办法

CF

可以认 为

我想

Z

Z

我相信

我想

SST

SST-C

ZSST

SST-ZC

SST-ZC

SST-C

ZSST

SMT

E

E

NA

NA

E

E

SV

E

E

SO

SOS

E

SOS

SOS

E

E

(continued)

SV −

SV −

SV −

E

188 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2002

1998

1999

2002

1991

1991

No.

119

120

121

122

123

124

(continued)

PMT

PMT

PMT

PMT

PMT

PMT

Speaker

啊 因此 我们 进一步 准备 啊 在 各个 领域, 在 政治 领域, 经济 领域 啊, 科学 教育 各 方面, 都 和 东南亚 国家 保持 良好 的 关 系, 这 是 您 刚才 说 的, 这 是 对的

我 想, 该 提 的 问题 都 差不多 了吧

我 认为 本届 政府 实现 了 它 对 全国 人民, 全国 人民 代表大 会 所 做出 的 诺言

这 是 不 容易 的

这种 三 大 民主 , 我 认为 也 是 非常 好 的 一个 方向

TT So [[I think]] the problem is not whether or not the country has a fiscal deficit So [[I think]] these are all very good ways and also good directions for development So [[I think]] this is really not easy So [[I think]] this government has made good its promise and delivered what we have promised before the people of this country, before the deputies of the National People’s Congress So [[I think]] you have almost exhausted all the questions that you want to ask And China also stands ready to further develop relations with these countries in all the aspects, including in the political, economic, scientific and educational fields. And China would like to maintain good relations with all of them in all these fields. So [[I think]] you’re right in saying that just now

ST

因此, 问题 不 在于 你 的 财政 有 没有 赤字

CF

Z

我想

我认为

Z

我想

Z

SST

SST-ZC

ZSST

ZSST

SST-ZC

ZSST

SST-ZC

SMT

NA

E

E

NA

E

NA

SV

SO

SOS

E

E

SOS

E

SOS

(continued)

SV −

E

SV − E

E

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 189

Year

2008

2004

2007

No.

125

126

127

(continued)

PMT

PMT

PMT

Speaker

TT [[I think]] it is undesirable to set a deadline to the consultations [[I think]] I’m somewhat troubled by the last two aspects. But there is no alternative I think to resolve the people’s well-being, this issue, actually involves efforts concerning people’s clothing, food, housing and transportation. At this stage, [[I think]] what is most important is to ensure that people have equal opportunities in education. We need to continue to pursue a proactive employment policy. We need to gradually reduce the gap in terms of income distribution and we need to put in place a rural and urban social security system that covers all people

ST

我们 认为 为 磋商 人为 地 设定 时限 的 做法 是 不 明智 的 在 这 两 个 方面, 我 确实 也 不 太 放心, 但是 没有 其他 的 办 法了 这 就 需要 解决 民生 问题 . 民 生 问题 涉及 人们 的 衣食住行, 但 当前, 最 重要 的 是 促进 教 育 机会 的 均等, 啊, 这个, 实施 积极 的 就业 政策, 逐步 缩小 收入 分配 的 差距 , 建立 覆盖 城乡 的 社会 保障 体系

CF

Z

我们认 为 Z

SST

SST-ZC

SST-ZC

SST-C

SMT

NA

NA

E

SV

SO

SOS

SOS

SOS

(continued)

SV −

SV −

E

190 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1989

2008

2003

No.

128

129

130

(continued)

PMT

PMT

PMT

Speaker

我 觉得 你 现在 提出 这个 问题 啊, 为 时 还 稍许 早 了 一点

关于 个人 的 风格, 我 想 中国 的 历任 外长 各 人 都 有 各 人 的 这个, 呃, 风格 , 但是 有 一点 是 共同 的, 那 就是 忠于 国家, 忠于 人民, 广 交 天下 的 朋友, 促进 和平 发展 事业

TT The law of a enterprises has provided for the role and the, how to, the, the role and the importance of workers in such enterprises, and with the economic, further economic development and the further deepening of the reform and the, you know structure, [[I think]] we shall make some supplementary measures and policies in this regard Talking about personal style, [[I think]] each of the successive foreign ministers of China has his own personal style, but they all share a commonality, that is, loyalty to the country, loyalty to the people, and a commitment to make as many friends as possible in order to promote world peace and development. [[I think]] it is a bit too early for you to raise this question at this point

ST

关于 工人 的 作用 以及 工人 的 积极性 的 重要性 在 《企业 法》 里 也 做 了 表述 . 我 想 在 今后 的, 随着 经济 建设 的 发 展, 随着 经济 改革 体制 改革 地 深入, 在 这 方面 将 有 一些 补充 地 规定 和 措施

CF

我觉得

我想

Z

SST

SST-C

ZSST

SST-ZC

SMT

E

E

NA

SV

SO

E

E

SOS

(continued)

SV +

E

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 191

Year

1989

1998

2001

2000

2011

2011

No.

131

132

133

134

135

136

(continued)

PMT

PMT

PMT

PMT

PMT

PMT

Speaker

所以, 中国 在 尽 自己 的 国际 的 责任 义务 的 同时, 把 中国 自己 的 事情 办 好, 也 就是 为 创造 一个 更加 美好 的 世界 的 未来 在 做出 贡献

我 想 这 位 记者 女士 啊, 呃, 讲 得 不错

我 觉得 克林顿 总统 这 一句话 应该 改 两 个 字 才 比较 确切

我 把 这个 叫做 他 的 角色 错 位

我 目前 我 的 思想 非常 单纯

TT Therefore, [[I think]] some errors made in decision-making should be a, should be attributed to a collective responsibility So as for my thinking, [[I think]] at the present that’s very simple I think what I would describe is they have, uh, [[I think]] they have played, uh, they have not played their role properly However in my view, [[I think]] revision needs to be made concerning two words in this sentence made by President Clinton so that it can be more accurate Well, [[I think]] I agree with you, this lady journalist in what you said While fulfilling its international responsibilities, China is working hard to run its own affairs well. [[I think]] that in itself is China’s contribution to building a better future for the world

ST

因此 我 认为 我们 决策 上 的 某些 失误 应该 是 集体 的 责任

CF

Z

我想

我觉得

Z

Z

我认为

SST

SST-ZC

ZSST

SST-C

SST-ZC

SST-ZC

ZSST

SMT

NA

E

E

NA

NA

E

SV

SO

SOS

E

E

SOS

SOS

E

(continued)

SV −

E

SV +

SV − SV −

E

192 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

1990

2007

2004

2007

No.

137

138

139

140

(continued)

PMT

PMT

PMT

PMT

Speaker

我们 改革 和 建设 的 最终 目的 , 是 满足 人们 日益 增长 的 物质 和 文化 的 需求

去年 就 在 这里 我 曾经 把 社 会主义 比作 大海, 海 不 辞 水, 故能成其大

民主 制度 啊, 同 任何 真理 一 样, 同样 要 接受 实践 的 考验, 任何 地区 和 国家, 民主 … 制度 的 状 况, 优劣, 都 要 以 实践 为 标准

TT Well, [[I think]] there is no need to talk about this question now because in giving news reports you focus, your focus is on news that is something new [[I think]] democratic system, like whatever truth, needs to be tested by practice and what actually happens. Taking the democratic systems in whatever country or region for example, whether the system is good or bad, I think we can leave the answer to the reality and to the real practice [[I think]] it was in this room last year that I compared socialism with a big ocean. The ocean never turns away streams, so it becomes wide and deep [[I think]] the ultimate goal of our reform and development is to meet the increasing material and cultural needs of our people

ST

新闻 新闻 吗 应该 是 NEWS, 这 个 问题 已经 , 在 现在 再 来 问 这个 问题 没有 什么 回答 的 必 要了

CF

Z

Z

Z

Z

SST

SST-ZC

SST-ZC

SST-ZC

SST-ZC

SMT

NA

NA

NA

NA

SV

SO

SOS

SOS

SOS

SOS

(continued)

SV −

SV −

SV −

SV −

Appendix C: Coding Sample of Modality Shifts: I/We think 193

Year

2007

2009

2004

2009

2007

No.

141

142

143

144

145

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker

关键 在 维护 台 海 的 和平 与 稳定

你 的 问题 很 重要

你 违反 规定, 问 了 两 个 问题

你 对 “机遇” 的 了解 和 我们 是 一致 的

TT [[I think]] to resolve the people’s well-being, this issue, actually involves efforts concerning people’s clothing, food, housing and transportation. At this stage, I think what is most important is to ensure that people have equal opportunities in education. We need to continue to pursue a proactive employment policy. We need to gradually reduce the gap in terms of income distribution and we need to put in place a rural and urban social security system that covers all people [[I think]] you share the same interpretation of the word “crisis” with us [[I think]] you violated the rules of this press conference by asking two questions [[I think]] you have raised a very important question [[I think]] the key to the importance of these two years lies in safeguarding peace and stability in the Taiwan Straits

ST

这 就 需要 解决 民生 问题 . 民 生 问题 涉及 人们 的 衣食住行, 但 当前, 最 重要 的 是 促进 教 育 机会 的 均等, 啊, 这个, 实施 积极 的 就业 政策, 逐步 缩小 收入 分配 的 差距, 建立 覆盖 城乡 的 社会 保障 体系

CF

Z

Z

Z

Z

Z

SST

SST-ZC

SST-ZC

SST-ZC

SST-ZC

SST-ZC

SMT

NA

NA

NA

NA

NA

SV

SO

SOS

SOS

SOS

SOS

SOS

(continued)

SV − SV −

SV −

SV −

SV −

194 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2000

2011

2008

2008

1989

1990

No.

146

147

148

149

150

151

(continued)

PMT

PMT

PMT

PMT

PMT

PMT

Speaker As to my own job and my own work, [[I think]] it is always limited. On China’s position in the world, [[I think]] what has happened shows that China is still a developing country [[I think]] any fair-minded people with an objective view would admit that the support Chinese people have demonstrated to the Olympic Games is absolutely sincere and from the bottom of the Chinese people’s hearts [[I think]] the first question is easier to answer than the second part [[I think]] now we can stop arguing about this project for a while [[I think]] the problem of a sluggish market in China hasn’t been completely solved

想 只要 公正 的, 客观 的 人 会 认识 到, 中国 人民 是 发 内心 地 对 奥运 盛会 的 巨 地, 衷心 地, 热烈 地 支持

市场 疲软 现象 现在 还 没有 完 全 扭转

所以 我 建议 这个 问题 现在 可 以 不必 去 争论

前面 一半 的 问题 比 后面 一半 的 问题 好像 好 回答 一点

我 都 自 大

关于 中国 的 国际 定位 问题, 我 想 事实 明确 地 表明, 中国 仍然 是 发展 中 的 国家

TT

ST

我 自己 所 做 的 工作 是 有限 的

Z

我建议

好像

我想

我想

Z

CF

SST-ZC

ZSST

SST-C

ZSST

ZSST

SST-ZC

SST

NA

E

E

E

E

NA

SMT

SOS

E

SOS

E

E

SOS

SO

(continued)

SV −

E

SV +

E

E

SV −

SV

Appendix C: Coding Sample of Modality Shifts: I/We think 195

Year

2007

1989

2002

1999

2000

No.

152

153

154

155

156

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker As for your question on the fact that the Chinese government did not meet the goal in environmental protection last year, [[I think]] I have already made a very full explanation at the NPC session on this matter. And this year we are going to take major measures on eight fronts, which I will not repeat here To markedly lower the price rises in this year than last year, [[I think]] we’ll certainly be able to do so if we firmly implement the measures covered by Premier Li Peng in his report [[I think]] it’s still… it’s still too early for you to pose such a question, because even I myself don’t have an answer to your question [[I think]] on this question, some people in the United States have made two mistakes, due to the underestimation [[I think]] on the Taiwan question. I’ve already made myself more than clear

我 认为, 在 这个 问题 上 美国 方面 的 人士 犯 了 两 个, 犯 了 两 个 “过 低 估计” 的 错误 关于 台湾 问题 .我 认为 我 刚 才 已经 讲 得 很 清楚 了

我 觉得 你 提 这个 问题 的 时 机 恐怕 是 太 早 了 一点, 因为 你 所 提 的 问题 的 答案, 连 我 都 不 知道

明显 低于 今年, 这 一 条, 我, 我 认为 如果 我们 坚决 贯彻 执行 李 鹏 总理 的 政府 工作 报告 的话 肯定 是 可以 做到

TT

ST

关于 … 我们 去年 没有 完成 环保 的 目标, 我 在 人代会 上 已经 作 了 充 分 的 说明, 今后 采取 的 措施 也 提出 了 八 个 方面 的 工作. 我 想 不再 重复 了

SST-C

ZSST

ZSST

我认为

我想

ZSST

SST-ZC

SST

我觉得

我认为

Z

CF

E

E

E

E

NA

SMT

E

E

E

E

E

E

SOS

SO

(continued)

SV +

E

SV −

SV

196 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2001

1999

1998

1999

2002

No.

157

158

159

160

161

(continued)

PMT

PMT

PMT

PMT

PMT

Speaker

我 想 我 在 1998年 的 时候 所 提出 来 的 “一个 确保, 三 个 到 位, 五 项 改革”, 在 这 四 年 里 面 就 已经 基本 完成

过去 的 一 年 我 感到 非常 困 难, 这个 困难 超过 我 的 预期 的 程度

我, 对不起, 我 觉得 外国 舆论 对 中国 国有 企业 的 困难 看 得太大了

我 觉得 那些 债权 银行, 呃, 以 某些 金融 机构 对于 这个 问题 的 估计 太 悲观 了, 就 认为 中 国 已经 发生 了 金融 危机, 没 有 支付 能力, 不 讲 信用, 过分 了

TT [[I think]] our efforts in, uh, reforming, in carrying out the government institutional reform have been successful [[I think]] some creditor banks and financial institutions have been too pessimistic in their estimation and evaluation of this incident uh in believing that China has, is experiencing a financial crisis and that China has no payment capabilities and that China is no longer creditworthy. I think that is going too far [[I think]] some foreign media have overestimated or played up the difficulties of China’s state-owned enterprises I/Zhu: [[I think]] the past year was extremely difficult and the difficulties actually were greater than we anticipated [[I think]] what I said in my first press con… conference “one must, three completions and five reforms”, all the tasks have been basically completed over the past 4 years

ST

我 认为 本届 政府 的 机构 改革 是 成功 的

CF

SST

SST-C

ZSST

我想

SST-C

SST-C

ZSST

我感到

我觉得

我觉得

我认为

SMT

E

E

E

E

E

SV

E

SO

E

E

E

E

E

(continued)

SV +

SV +

SV +

E

Appendix C: Coding Sample of Modality Shifts: I/We think 197

Year

2000

1989

2008

1989

No.

162

163

164

165

(continued)

PMT

PMT

PMT

PMT

Speaker [[I think]] you have raised a very vivid and interesting example of the estranged wife when posing your question And [[we think]] that if this, the work of this regard is carried out well, then it will contribute to the economic development and the social stability [[We think]] countries should put each other, put themselves in each other’s shoes, take a fair and objective view, recognize and accommodate the reasonable interests and needs of others because this is the way for countries to deepen mutual trust, expand common ground and promote cooperation We welcome supervision on the part of the National People’s Congress over our work because if [[we think]] the St, State Council really has the power, then it is power granted or given to us by the people

我们 欢迎 人大 对 我们 的 工作 进行 监督 , 因为 我们 这个 国 务院 如果 说 有 权 的话 也 是 人民 给 我们 的 权

我 觉得 国家 之间 啊 一定 要 怀着 这个 换位 思考 的 意识, 抱 着 公正 客观 的 态度, 来 承 认 和 照顾 彼此 的 合理 的 利 益 的 需要, 这样 呢 才 能够 增 进 互信, 加强 合作, 加强 共识

如果 处理 的 好 将 会 促进 经 济 的 发展 , 促进 社会 的 稳定

TT

ST

你 讲 的 这个 例子 很 风趣

Z

我觉得

Z

Z

CF

SST-ZC

SST-C

SST-ZC

SST-ZC

SST

NA

E

NA

NA

SMT

SOS

SOO

SOS

SOS

SO

(continued)

SV −

SV +

SV −

SV −

SV

198 Appendix C: Coding Sample of Modality Shifts: I/We think

Year

2003

2003

1989

2008

No.

166

167

168

169

(continued)

PMT

PMT

PMT

PMT

Speaker

我们 呢, 这个, 啊, 这个 … 非常 重视 在 平等 相互 尊重 的 基础 上, 同 欧盟, 欧洲 国家 就 人权 问题 进行 建 设性 的 对话

今年 嘛 生产 降 到 一个 我们 认为 比较 适合 的 速度 , 这样 也 减少 了 用工

关于 伊拉克 的 投票 的 问题, 伊拉克 问题 的 最后 的 投票 问 题, 对于 新 决议, 我们 认为 没 有 必要 提出, 因为 1441 号 决 议 规定 的 任务 还 没有 完成

TT [[We think]] regarding the Iraq issue, diplomacy at the Security Council is still going on. We are not yet at the end of the road to a political solution With regard to the vote on Iraq and the new resolution, [[we think]] it is not necessary to introduce any new resolution because the tasks spelled out by Resolution 1441 are not completed yet Therefore, this year, when we reduced the growth rate to what [[we think]] is the appropriate level, this has also resulted in some surplus farmers [[We think]] it is important to have constructive dialogue with the European Union and European countries on human rights on the basis of equality and mutual respect

ST

我们 认为 现在 围绕 着 伊拉克 问题, 特别 是 在 安理会 的 外 交 舞台 上, 政治 解决 的 道路 尚未 穷尽

CF

SST

ZSST

ZSST

我们非 常重视 =we think it is important

ZSST

ZSST

我们认 为

我们认 为

我们认 为

SMT

E

E

E

E

SV

E

E

E

E

SO

E

E

E

E

Appendix C: Coding Sample of Modality Shifts: I/We think 199

Appendix D

Telephone Interview with Mr. Chen Mingming

Interviewee: Mr. Chen Mingming, former Director of the Department of Translation and Interpreting, Ministry of Foreign Affairs Interviewer: Li Xin, author of this book Time of interview: 16:00–16:30, April 27, 2015 Working Language: Chinese Note: The following script is drafted based on the interviewer’s notes taken during the telephone interview and has been confirmed by the interviewee as an accurate representation of ideas exchanged during the interview. 1. Q: 您认为, 政府记者会口译的最高标准或原则是什么? 对原文的忠实是否 比译文的表达流利和可接受性更加重要? 或者两者同样重要? A: 最高原则还是忠实, 完整, 准确。 这是第一位的。 在此基础上要尽量让 译文流畅, 容易被听众接受, 不会让听众产生误解。 比如, 在不违反说话者 原意的前提下, 可以对汉语口语中的重复进行删减, 使英语译文更加 presentable, 也可以对原文逻辑不够清晰等情况进行调整, 使说话者的本意更 加清楚地的传递到译文中。 但是总的来说, 对原文进行增减调整的余地不 是很大。 2. Q: 张援远 (2004) 在 《谈谈领导人言论英译的几个问题》 一文中提到 “对 准确性的再理解”, 即准确不等于完全直译, 并指出“现在一个突出的问题 是, 有一些同志为了‘准确’, 更准确地说是为了避免被说成 ‘犯错’, 工作中 习惯于 play safe, 就是喜欢追求译文外部形态上和原文相近, 偏重直译”, 所 以他鼓励译员 “下功夫透彻地理解原义, 追求更加传神的表达。” 您是否同 意他的观点? 他还指出, 外国人经常批评我们的译文“官气”太重, 缺乏 “亲 和力”, 翻译人员虽不能解决这一问题, 但却可以在某种程度上对其进行缓 解。 这是否意味着外交部翻译室鼓励译员在为国家领导人担任口译时, 适 当灵活地调整原文的用词, 情态, 语气等, 以拉近与听众的情感距离? A: 同意要避免硬译, 避免中式英语, 但这并不意味着鼓励译员对原文大胆 地进行调整。 一切取决于原文。 忠实, 准确地传达原文信息是第一位 的。 对情态词的增减和调整也是不鼓励的 (“当然不鼓励”), 因为增减和调 整之后会与原意有出入。 但是针对中文原文枯燥沉重逻辑关系不清晰的 情况,鼓励译员将其灵活处理为更加易懂更加口语化的英语结构。

© Springer Nature Singapore Pte Ltd. 2018 X. Li, The Reconstruction of Modality in Chinese-English Government Press Conference Interpreting, Corpora and Intercultural Studies 1, DOI 10.1007/978-981-10-5169-2

201

202

Appendix D: Telephone Interview with Mr. Chen Mingming

3. Q: 您是否认为记者会口译的 “忠实” 原则意味着在不破坏译文可读性的前 提下, 尽量直译? 具体到情态翻译, 当汉英双语中有情态对等词的情况下, 是否鼓励译员尽量选用这些对等词来译而非灵活调整? A: 是的。 只有当直译会造成逻辑不清, 误解或不通顺的英文时才需要灵 活调整。 4. Q: 施燕华 (2009) 在 《外交翻译60年》 一文中提到“近年来, 外交翻译工作 的一个努力方向, 就是结合中央领导及我部领导对外讲话的场合及听众特 点, 尽量用外国人听得懂, 易理解的语言进行翻译, 在确保忠实原文的基础 上, 使译文更加生动顺畅。 ”您是否同意她的观点? A: 同意, 是口译的方向 。 5. Q: 两会期间的政府记者会一直采用交替传译而非更加节省时间的同声传 译, 主要原因是什么? A: 2015 年两会期间除了外长, 总理这两场记者会采用同传以外, 其他如发 改委, 新闻发言人等记者会都已经改为同传, 为的是提高效率节省时间。 但总理和外长这两场记者会是两会期间最关键, 媒体影响力最大的记者会, 目前看来会一直用交传, 以确保口译的准确性, 并且给领导人以更多的构 思答问的时间。 (同传的失真率要远远高于交传。) 6. Q: 为政府记者招待会担任口译的译员是否都是公务员和党员身份? A: 100% 都是公务员, 大多数都是党员。 7. Q: 政府记者会口译事件的组织者是谁? 发起者是谁? 就汉-英口译而言, 客 户是否仅为在场的外国记者? A: 记者会和口译是由 “人大” 牵头组织的, 邀请总理, 外长来答记者问。 没 有 “客户” 的说法, 只有 “对象”。 译员主要为总理或外长服务, 其口译目的 是将总理或外长的发言信息传达给外国媒体和外国公众。 8. Q: 政府记者会口译质量监督/监控者的角色主要由谁承担? 是否经常会调 查口译译文听众的反馈来保证口译质量? A: 就总理和外长的记者会而言, 口译质量监督/监控的工作由外交部翻译 室来承担。 翻译室的专家领导会听口译现场, 观摩口译录像, 分析现场口 译实践, 然后对译员进行评估和反馈。 听众的反馈对我们来说并不重要, 因为听众的反馈并不是专业的。

Smile Life

When life gives you a hundred reasons to cry, show life that you have a thousand reasons to smile

Get in touch

© Copyright 2015 - 2024 AZPDF.TIPS - All rights reserved.