An Investigation of Addressing, Openings and Closings in Chinese Personal Letters

Linguistics Department School ofModem Language Studies The University ofNew South Wales

AN INVESTIGATION OF ADDRESSING, OPENINGS AND CLOSINGS IN CHINESE PERSONAL LETTERS

Xianghong Cao

Thesis submitted in fulfilment of the requirements for the Degree of DOCTOR OF PHILOSOPHY

April, 2004 PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Project Report Sheet

Surname or Family name: C tt..O

First name: X I~M'J Other name/s: Abbreviation for degree as given in the University calendar: (11,-_ "_S) school: /Vlodem L~l/l."cJe ~t-tcf,es Faculty: Ar::Cs Title: Art ~ves-tJ~'tr/\. of ~s-s/~ / tpe/u(Js cu-.o( clos/(Js .t.A. cA.;..... e.S""e per

Abstract 350 words maximum: (PLEASE TYPE)

Investigations into address forms, openings and closings in the sociolinguistics and discourse fields have largely focused on oral communication. This study has set out to examine linguistic and pragmatic features of addressing, opening and closing behaviours in relation to the effects of social factors and the cultural constraints in actual language use in Chinese personal letters. The aim of this research is to provide a better understanding of the forms, functions, sequences and uses of utterances in these three routinized speech behaviours in written interaction and intercultural communication in the Chinese context. Using a database of 259 personal letters written by Chinese in Xinjiang Autonomous Regions, an analysis and discussion are presented, using both qualitative and some simple quantitative techniques.

Findings reveal that there is a variety of address forms that Chinese writers use in personal letters, which is a reflection of the diversity and elaborateness of this practice in Chinese. In terms of openings and closings, this study suggests that in Chinese personal written interaction, these two behaviours are highly ritualized and closely structured, consisting of a series of phatic and routine speech acts to achieve the establishment and the consolidation of interpersonal relationships between writer and recipient. With respect to the relationship between language variation and social variables, the study shows that the choices of language patterns or sequences vary with generation, gender, the level of authority, the level of intimacy and social distance. This research also reflects that the influence of Confucian ethics in the Chinese cultural context has a considerable impact on Chinese people's linguistic performance in terms of addressing practice, openings and closings in personal letters.

Findings from this study highlight three issues. First, it is suggested that there is much to be gained in pragmatic studies of phatic communion, speech acts and politeness by shifting the focus from the single unit to the discourse domain. Second, the influences of social variables in language use are extremely complex, and analysis needs to take this into account. Third, there is a need to incorporate the larger cultural context as a determinant factor in the study of speech acts and behaviours in communication.

Declaration relating to disposition of project reporVthesis CERTIFICATE OF ORIGINALITY

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in this thesis. Any contribution made to the research by others, with whom I have worked at UNSWor elsewhere, is explicitly acknowledged in this thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance fromot hers in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

(Signed) ACKNOWLEDGEMENT

I would like to gratefully acknowledge all those who have contributed to the completion of this thesis. First, and foremost, I would like to express my deepest thanks to my supervisor, Dr Rod Gardner, for his great assistance and inspiration. His extraordinary patience, enthusiasm, and invaluable advice and guidance helped me overcome many difficulties, and had a great influence on this thesis.

I would also like to express my gratitude to Professor Peter Collins and Dr. Louise J. Ravelli for their valuable suggestions, as well as to Professor John Brotherton for his generous help with the formatting of the very final version of the text. Special acknowledgment must go to Professor Zhu Wanjing, who offered good advice on the concept of politeness during the time of my external research in Beijing.

Particular thanks are given to my students, colleagues and friends in Xinjiang Normal University who provided me with the personal letters that comprised the source of data for this research. Without their co-operation this thesis would not have been possible. My sincere thanks also go to Kang Yuanfei, Xing Mingang, Ben Hamblin, Wang Li, Maria Oujo, Emilia Djonov, Emmi Mikedakis, Gavin Farrell, Ke Wanli, Li Haihua, Gunnar Mollenbeck, Cui Jiangying, Zhang Shaosan, Zhang Erjie and Xu Shaorong for the immense moral support, encouragement, help, and kindness they gave me throughout much of my research.

Special thanks must also go to the Applied Linguistics Department of Auckland University for kindly providing me with access to their library and their computers during the time of the pre-final draft preparation.

Finally, my heartfelt thanks are given to my husband, Chen Peizhong, and my daughter, Chen Bai, for their remarkable understanding, patience, and spiritual and practical support. Particular appreciation is given to my mother for her complete devotion to the care of my family. I hope my whole family will share my happiness as I submit this thesis. ABSTRACT

Findings reveal that there is a variety of address forms that Chinese writers use in personal letters, which is a reflection of the diversity and elaborateness of this practice in Chinese. In terms of openings and closings, this study suggests that in Chinese personal written interaction, these two behaviours are highly ritualized and closely structured, consisting o_f a series of phatic and routine speech acts to achieve the establishment and the consolidation of interpersonal relationships between writer and recipient. With respect to the relationship between language variation and social variables, the study shows that the choices of language patterns or sequences vary with generation, gender, the level of authority, the level of intimacy and social distance. This research also reflects that the influence of Confucian ethics in the Chinese cultural context has a considerable impact on Chinese people's linguistic performance in terms of addressing practice, openings and closings in personal letters.

Findings from this study highlight three issues. First, it is suggested that there is much to be gained in pragmatic studies of phatic communion, speech acts and politeness by shifting the focus from the single unit to the discourse domain. Second, the influences of social variables in language use are extremely complex, and analysis nee.ds to take this into account. Third, there is a need to incorporate the larger cultural context as a determinant factor in the study of speech acts and behaviours in communication.

ii TABLE OF CONTENTS

CHAPTER ONE: INTRODUCTION

1.1 The nature of this study 1 1.2 Background 5 1.3 Research questions 6 1.4 Research objectives 7 1.5 Organization of the thesis 7

CHAPTER TWO: LITERATURE REVIEW

2.0 Introduction 9 2.1 Conceptual framework 9 2.1.1 Phatic communion and the relationship aspect of communication 9 2.1.2 Speech act theory 13 2.1.3 Grice's Cooperative Principle 18 2.1.4 Concept ofpoliteness 23 2.1.4.1 Brown and Levinson's model ofpoliteness 24 2.1.4.2 Studies inspired by Brown and Levinson's model of politeness 25 2.1.5 Language and social variables 28 2.1.5.1 Power and solidarity 28 2.1.5.2 Language and gender 30 2.1.6 Summary 33 2.2 Studies of address forms, openings and closings in interaction 34 2.2.1 Address forms 34 2.2.1.1 Concept and functions 34 2.2.1.2 Previous studies of address forms 35 2.2.2 Openings 42 2.2.2.1 Concept of conversational openings 42 2.2.2.2 Types of conversational openings 43 2.2.2.3 Functions of conversational openings 44 2.2.2.4 Structures of conversational openings 48 2.2.3 Closings 54 2.2.3.1 Concept of conversational closings 55

lll 2.2.3.2 Functions of conversational closings 55 2.2.3.3 Structures of conversational closings 58 2.3 Unresolved issues and research opportunities 63 2.4 Concluding remarks 66

CHAPTER THREE: METHODOLOGY

3.0 Introduction 67 3.1 Data collection 67 3.1.1 Sampling 67 3.1.1.1 Place 68 3.1.1.2 Data 69 3 .1.1.2.1 Chinese personal letters 69 3 .1.1.2.2 Reasons for choosing naturally occurring personal letters 71 3.1.1.2.3 Writers and recipients 73 3 .1.2 Methods for data collection 75 3.1.3 Transliteration 76 3.2 Data analysis 77 3.2.1 Framework for analysis 77 3 .2.1.1 Identification of sections of personal letters 78 3.2.1.2 Identification of linguistic forms/components/sequences 79 3 .2.1.2.1 Identification of linguistic forms in the addressing section 79 3 .2.1.2.2 Identification of components in the opening and closing section 80 3 .2.1.2.3 Identification of sequences in the opening and closing section 86 3 .2.1.3 Identification of social values 88 3.2.1.3.1 Seniority/juniority 88 3.2.1.3.2 Formality/informality 88 3.2.1.3.3 Distance/intimacy 89 3.2.1.3.4 Respect/consideration 90 3.2.2 Data analysis 91

IV 3 .2.2.1 Quantitative vs. qualitative analysis 91 3.2.2.2 Combining statistical analysis and in-depth interpretation 93 3.3 Concluding remarks 94

CHAPTER FOUR: ADDRESS FORMS IN CHINESE PERSONAL LETTERS

4.0 Introduction 95 4.1 Address forms and functions in Chinese personal letters 95 4.1.1 The name-alone strategy 96 4.1.2 The compound name strategy 99 4.1.3 The no-name strategy 102 4.1.4 Discussion 104 4.2 Address forms and social variables 107 4.2.1 Address forms and social roles 107 4.2.2 Discussion 111 4.2.3 Address forms and age/generation 116 4.2.4 Discussion 118 4.2.5 Address forms and gender 120 4.2.6 Discussion 123 4.3 Concluding remarks 126

CHAPTER FIVE: OPENINGS IN CHINESE PERSONAL LETTERS

5.0 Introduction 128 5.1 Opening strategies 128 5 .1.1 Greetings 131 5.1.2 Inquiries 135 5 .1.3 Acknowledgements of receipt of a previous letter 137 5.1.4 Apologies 140 5.1.5 Expressions of good wishes 142 5.1.6 Expressions ofmissing 143 5 .1. 7 Expressions of concern 144 5.1.8 Expressions ofthanks 145 5.1.9 Reassurances about oneself 146

v 5 .1.1 0 Expressions of congratulations 146 5.1.11 Expressions ofhope 147 5.1.12 Sending regards to others 148 5 .1.13 Discussion 149 5.2 Opening types 152 5.2.1 Single openings 153 5.2.2 Combined openings 154 5.2.2.1 In initial letters 155 5.2.2.2 In reply letters 162 5.2.3 Discussion 169 5.3 Openings and social variables 174 5.3.1 Openings and social roles 174 5.3.1.1 Single openings and social roles 175 5.3.1.2. Combined openings and social roles 175 5.3.1.2.1 In initial letters 176 5.3.1.2.2 In reply letters 178 5.3 .1.3 Discussion 179 5.3.2 Openings and age/generation 182 5.3 .2.1 Opening types and age/ generation 182 5.3.2.2 Combined openings and age/generation 183 5.3 .2.2.1 In initial letters 183 5.3.2.2.2 In reply letters 184 5.3.2.3 Discussion 185 5.3.3 Openings and gender 187 5.3.3.1 Differences in the choice of opening types 187 5.3.3.2 Differences in the choice ofthe strategy of single openings 188 5.3.3.3 Differences in the choice of combination patterns of openings 188 5.3.3.3.1 In initial letters 189 5.3.3.3.2 In reply letters 190 5.3.3.4 Discussion 191 5.4 Concluding remarks 193

VI CHAPTER SIX: CLOSINGS IN CHINESE PERSONAL LETTERS

6.0 Introduction 195 6.1 Closing strategies 195 6.1.1 Expressions of good wishes 197 6.1.2 Pre-closings 199 6.1.3 Longing for future contact 200 6.1.4 Sign-off phrases 202 6.1.5 Expressions of concern 202 6.1.6 Reassurances about oneself 203 6.1. 7 Farewell messages 204 6.1.8 Sending regards to others 205 6.1.9 Offering help 205 6.1.1 0 Expressions of thanks 206 6.1.11 Apologies 206 · 6.1.12 Discussion 207 6.2 Closing types 211 6.2.1 Single closings 212 6.2.2 Combined closings 214 6.2.3 Discussion 229 6.3 Combined closings and letter types 233 6.4 Closings and social variables 234 6.4.1 Closings and social roles 234 6.4.1.1 Single closings and social roles 235 6.4.1.2 Combined closings and social roles 236 6.4.1.3 Discussion 239 6.4.2 Closings and age/generation 242 6.4.2.1 Closing types and age/generation 242 6.4.2.2 Combined closings and age/generation 243 6.4.2.3 Discussion 244 6.4.3 Closings and gender 246 6.4.3 .1 Differences in the choice of closing types 246 6.4.3 .2 Differences in the choice of the strategy of single openings 24 7 6.4.3.3 Differences in the choice of combination patterns of closings 247

vii 6.4.3.4 Discussion 248 6.5 Concluding remarks 250

CHAPTER SEVEN: CONCLUSION

7.0 Introduction 252 7.1 Major findings of the study 252 7.1.1 Address forms as multiple indicators 252 7 .1.2 Openings as a means of opening up a channel and establishing the relationship 254 7 .1.3 Closings as a means of ending the contact but continuing the relationship 25 8 7.2 Significance of this study 261 7.2.1 Implications for sociolinguistic theory 262 7.2.2 Implications for learning and teaching language and cross-cultural communication 266 7.3 Limitations of this study 268 7.4 Directions for future research 269

REFERENCES 272

List of tables 291 List of figures 293

Vlll CHAPTER ONE INTRODUCTION

1.1 The nature of this study Personal letter writing is regarded as "one of the most widespread uses of writing among different social classes and cultural groups" (Kalman 1996: 193), despite the use of advanced technologies of communication becoming more widespread nowadays. In China, personal letter writing has had a particularly long and significant history (Chen 1999), and still retains its essential status in modem times. This is due especially to the increases in the literacy rate (82%, based on 1997 census), the mobility of the work force, the enrolments in tertiary schooling (3.41 million students in universities and colleges, based on 1998 census), and overseas migration.

Personal letter writing in China is an important communicative activity for three major reasons. First, it is seen as an economically interactive practice that has been readily adopted by many people in their communicative repertoires. Second, it is an effective way of exchanging information, thoughts, feelings and attitudes, of educating younger persons and of showing respect to older people (Nin & Liu 1993). Third, letter writing is also viewed as a kind of activity in which a great deal of attention is paid to politeness, conveying a high degree of respect and solidarity (Maier 1992; Nevalainen & Raumolin-Brunberg 1995). Considering their economic, social, emotional and politeness implications, personal letters have thus become one of the most useful sources of material for studying sociolinguistics and pragmatics.

Researchers are interested in how sections of letters are realized and in how letters can be discussed and explained in pragmatic terms. This study has attempted to explore these characteristics in three behaviours of modem Chinese personal letters: addressing practices, openings and closings. The purpose of this exploration is to discover the "knowledge, beliefs and values concemmg literacy [and to discover] how it is accomplished" (Kalman 1996: 192).

There are three respects in which conducting investigations into the addressing, opening and closing behaviours is significant. The first is that of academic consideration. As is

1 known, there are strict rules covering the composition of personal letters, such as the rules of addressing practices, openings and closings. These issues have been central to the production of personal letters. For instance, addressing practices can be seen as one of most important communicative activities "which can best illustrate the close bond of language and society" (Shih 1986: 98). The opening behaviour in personal letters is a crucial step of opening up the communicative channel (Hopper 1992), of identifying and affirming the membership of writer and recipient in a social group, and of establishing and negotiating the relationships between them (Laver 1975, 1981; Yahya Othman 1995). The performance of closings in personal letters is also viewed as a significant behaviour, because in it the writer tries to mitigate the face threat to the addressee, which may arise from the initiation of an end of contact, and also tries to consolidate and reinforce the continuation of the relationship into the future (Laver 1981; Placencia 1997).

Taken together, it is believed that studying these three crucial speech behaviours in personal letters will provide some information about the linguistic and pragmatic features of address forms, openings and closings in written interaction in the Chinese context. Such information will contribute to a better understanding of the rules of address, openings and closings. These rules reflect Chinese people's principles of social interaction, thought patterns, values and beliefs. This information will also show how social relationships, emotions, cultural values and norms, personal interests and histories figure in written language (Dittmer 1991). Thus, the findings obtained from the investigations of letter writing will shed light on the study of certain routine speech behaviours in various types of communications.

The second point of significance is that of providing native speakers of Chinese with information about the conventions, knowledge and cultural values embedded in these behaviours. Writing letters is a goal-oriented social action. When people are engaged in writing, they normally have one or more communicative goals in mind: to convey information, to express emotions and affections, to identify the social status of the recipients and themselves, to identify the interpersonal relationships between them and to establish or reinforce the relationships. In order to ensure that these goals are achieved, native speakers need to manage their language in an appropriate way,

2 especially in these three sections of letters. However, according to the author's experiences, it seems that not all individuals of the native language are competent in writing the addressing, opening and closing sections of personal letters. Some writers do not know how to use formulaic and routinized expressions, which are "appropriate to their social status (and rank)" (Duranti 1992: 93), and do not know how to "connect linguistic units with social meaning" (Farr 1993: 3). Their inadequate knowledge of the beliefs and values involved in these behaviours may lead to inappropriate communication between native speakers of Chinese. For example, a university professor was upset being addressed by full name without any title by one of his university students (Chen 1999).

There are several reasons why some Chinese people fail to use appropriate expressions in these sections of personal written communication. One reason is that even if they do know the basic requirements for propositions of letter writing (i.e. addressing, opening, main body, closing, signature), they ignore the matters of style and content in each of these parts (e.g. how to use the honorific term of address towards seniors). The other reason is that while Chinese people have obtained linguistic knowledge of the structuring of these three sections, they lack information about the cultural values and norms which underpin these behaviours (e.g. the use of particular formulaic expressions in particular contexts). With such a shortage of "pragmatically relevant declarative knowledge" (Faerch & Kasper 1984), it is certain that native speakers of Chinese will experience some particular problems with the use of certain formulaic and conventional language in certain sections of personal letters. Therefore, it is a worthwhile endeavour to investigate addressing, opening and closing behaviours in Chinese personal letters. The empirical evidence gained from the research will be helpful for the development of both the literacy ability (Akinasso 1991) and communicative competence (Cook Gumperz 1993; Shethar 1993) ofnative speakers.

This investigation also has implications for the teaching Chinese as a foreign language and for cross-cultural communication. With the implementation of the open-door policy in China, especially with China's entry into the World Trade Organization (WTO), there has been an ever growing number of contacts between Chinese and people from other linguistic and cultural backgrounds. This communication takes place within the areas of

3 business, science, education, and also within personal contexts. However, with respect to the 'leamability' of languages, it is widely known that the speech in the first language cannot be directly translated into a second or foreign language. Therefore, people of one cultural group may "unknowingly violate another culture's ways or rules of speech" (Mayumi 2002: 3). This can happen in personal letters, as such speech behaviours as addressing practices, openings and closings are all open to misinterpretation and misunderstanding on the part of language learners and on the part of native speakers (Wolfson 1983). For example, seniors or elder friends of Chinese may feel uneasy when being addressed by their given names by American English people, since there are rigorous hierarchies in the addressing practices of Chinese society. However, it has been a trend for some time in the United States to address seniors or elder friends by their given names as a way of expressing solidarity (Hook 1984: 186). The striking contrast between Chinese and Americans in addressing behaviours may create intercultural communication obstacles, which lead to "shock, disappointment and even grievous insult" (Wolfson 1983: 62).

These sorts of misinterpretation and misunderstanding result from speakers transferring the rules of their own speech community into those of a new speech community (Wolfson 1983). They also result from speakers' lack of shared contextual knowledge when communicating with other persons. These problems are similar to those experienced by native speakers of Chinese when writing to each other.

To fully understand the complexities of cross-cultural communication one must understand both the forms and functions of communication and the embedded cultural values. To do this it is necessary to conduct systematic empirical analyses of the everyday routine speech behaviours of native speakers, such as addressing forms, openings and closings in various types of communication. Such explorations will help people of different cultures to be more aware of the other's ways of communication and will enable them to obtain access to the other's culture, thus helping eliminate cross cultural miscommunication and conflicts.

4 1.2 Background In the last few decades, a series of significant articles have contributed to a growth in the literature attempting to examine the speech behaviours of human interactions. This literature can be divided into two important categories. The first category comprises sociolinguistic studies focusing on speech-act level routines, such as greeting (e.g. Coupland, Coupland & Robinson 1992; Goffman 1971; Yahya-Othman 1995), thanks (e.g. Aijmer 1996; Coulmas 1981), apologies (e.g. Cohen & Olshtain1981; Holmes 1990), and wishes and hopes (e.g. Fong 2000; Verschueren 1981). The second category of literature deals with speech-behaviour level routines, such as terms of address (e.g. Blum 1997; Braun 1988; Brown & Gilman 1960; Chao 1956, 1976; Li 1997; Oyetade 1995; Zhu 1997), openings (e.g. Hopper 1992; Houtkoop-Steenstra 1991; Hutchby 1999; Omar 1992; Schegloff 1968, 1986; Schiffrin 1977) and closings (e.g. Albert & Kessler 1978; Clark & French 1981; Placencia 1997; Schegloff & Sacks 1973; Takami 2002). However, most studies concerned with the above aspects have focused on spoken language (e.g. face-to-face interaction, telephone talk). Limited attention has been given to the investigation of these speech-act and speech-behaviour routines in written communication. Even among the few studies that have been conducted, the majority are restricted to the business domain (Pilegaard 1997; Wei 1996; Yeung 1997). As a result of this, there is a lack of evidence on how people manage their language in particular speech behaviours and certain speech acts in written communication. Especially neglected is the personal field, which has been identified as a crucial area for the development of language and communicative competence, both for native and non native speakers.

Moreover, many previous studies have apparently been set to examine the influence of social variables (e.g. roles, age, gender) on the language management of speech behaviours and acts. Few studies have actually integrated social variables into examining the relationship between linguistic choices and social factors. As a result, there is a lack of knowledge about how the forms and content of addressing, opening and closing behaviours are influenced by the role status, the specific context and the social relationship between writers and recipients. In addition, while a great deal of research has focused on the influence of social factors, few studies have paid much attention to the effect of the large culturally contextual factor on the manifestation of

5 language in these behaviours. Thus, the existing literature falls far short of providing a general understanding of how addressing, opening and closings are shaped by particular cultural values and beliefs, which is claimed as "a whole complex network not only of meanings and symbols but also of event scripts" (Inchaurralde 1997: 55).

Furthermore, the majority of studies have been concerned mainly with American English or some European languages. There are few such studies concerned with the Chinese language. This fact shows that although Chinese is spoken by over one billion people, including native speakers of Chinese and speakers of various other language backgrounds, research on certain crucial speech behaviours in personal written communication is still in its infancy.

The point of all this is that the studies of some important speech behaviours available so far have not revealed the complexities of speech behaviours in various types of communication and in various languages. Thus, this study is an attempt to provide empirical linguistic and pragmatic evidence in relation to the addressing, openings and closings in personal letters in the Chinese context.

1.3 Research questions The present study has been motivated by recent sociolinguistic perspectives on speech behaviours, and by the need to extend people's understanding of issues concerning both intra-cultural and inter-cultural communication. In particular, it is intended to examine the speech behaviours of addressings, openings and closings in Chinese personal letters, with the purpose of discovering the linguistic, pragmatic and cultural phenomena embedded in these three speech behaviours. It is expected to add interesting dimensions to studies of speech behaviours in people's daily communication. Specifically, it will seek answers to the following basic research questions:

• What are the forms and functions of address, openings and closings in Chinese personal letters?

• To what extent do social variables (i.e. social roles, age, and gender) affect the use of address forms, openings and closings by the Chinese in personal letters?

6 • To what extent do the Chinese cultural values and norms influence the Chinese performance of these three behaviours in personal letters?

1.4 Research objectives The purpose of this study is to examine how Chinese people formulate their language in addressing, opening and closing behaviours in personal letters according to their status and the relationships with the recipients. By using authentic discourse as data and by implementing qualitative and to some extents quantitative approaches, this study is intended to discover and explain phenomena from the standpoints of lexis, semantics, syntax, discourse and pragmatics. Such a comprehensive analysis could render an overall picture of the forms and functions of the three speech behaviours in Chinese personal written interaction. It should also explicate how the occurrences of linguistic and pragmatic phenomena are subject to social factors and to the larger cultural constraint of a given situation. Generally, it is one of the initial studies whose findings should be significant. This is because it clarifies some important aspects of the manifestation of certain routine speech behaviours in written communication, thus contributing to the growing body of work on language-specific speech behaviours. Also it describes the genuine use of language in the Chinese context, and hence has a bearing on the wider fields of language pedagogy and cross-cultural communication.

1.5 Organization of the thesis The thesis is organized into the following seven chapters which delineate the research procedures and findings.

Chapter 2 reviews the literature on addressing forms, openings and closings. It also reviews other relevant issues, including the theoretical frameworks on phatic communion, speech act theory, Grice's Cooperative Principle, the concept of politeness, and the relationship between language and social variables. All the previous studies related to these three speech behaviours are regarded as both theoretical background and the basis of further discussion.

Chapter 3 presents the methodology used in the data analysis (i.e. ethnographical approach) and addresses the reasons for this methodology. In addition, the features of

7 data (authentic personal letters with different dimensions), the social characteristics of writers and recipients and the translation of the examples are all described in this chapter. Also defined and explained in this chapter are the analytical frameworks for analyzing the data from the levels of the lexical form, speech act and sequence. Finally, the reasons for combining quantitative and qualitative methods are explained.

Chapters 4, 5, and 6 constitute the main analysis and research findings portion of the work. The lexical and semantic-pragmatic analysis of address forms is presented in Chapter 4. In this chapter, attention is paid to the characteristics of written address forms on the lexical and semantic level and to social factors and underlying cultural meanings. Chapters 5 and 6 deal respectively with openings and closings focusing on the sequential and pragmatic analysis with respect to the basic, dominant and deviant structures in which they occur. Also demonstrated is the relationship between the formulation of ritual expressions and patterning of sequences and the constraints of social factors and cultural contexts.

Chapter 7 is dedicated to summing up the major findings of this study. It discusses the significance of this study in contributing to sociolinguistic theory, language teaching in both domestic and international domains and cross-cultural communication. It also indicates the limitations of this research and provides suggestions for future research.

In conclusion, speech behaviours such as addressings, openings and closings are aspects of language, which can best reflect the close bond between language and society. They play an important role in people's dealing with others. For example, from a simple greeting- 'Ninhao' (a singular and honorific pattern of 'How are you?') to an elaborate expression of concern 'Chunhan douqiao, wang ni duo zhuyi shenti. ' (In this chilly air of the spring period, I hope you take care of your health.), people can give some insights into addresser's relationship with the addressee. Knowing what people are doing when they use language in particular behaviours can aid our understanding of how language is a social action. The goal of this thesis is to take a step in the direction of deepening this understanding.

8 CHAPTER TWO LITERATURE REVIEW

2.0 Introduction This chapter begins by reviewing the literature on the conceptual framework relevant to this study, including phatic communion, speech act theory, Grice's Cooperative Principle, the concept of politeness and the relation of language to certain social variables. Next, it surveys some empirical research bearing on the issues of address forms, openings and closings in human interaction within the theoretical framework. It aims to map out a general picture with sufficient detail in order to guide the present research. After summarizing relevant theories and the findings of linguistic and sociolinguistic studies that have investigated addressing, opening and closing behaviours, the need for further research into several topics will be discussed.

2.1 Conceptual framework 2.1.1 Phatic communion and the relationship aspect of communication In every interaction, people are involved in the establishment and the maintenance of interpersonal relationships with others (Holmes 2000: 33). From this perspective, phatic communion cannot be ignored as a minor or marginal discourse mode, as it is closely linked to aspects ofhuman relationships. The phrase 'phatic communion' was introduced by the anthropologist Malinowski (1923, 1972), who treats phatic communion as "a type of speech in which the ties of union are created by a mere exchange ofwords" (1972: 151), that is, the whole purpose of speech is to establish and maintain the social bonds of the interactants, rather than to convey referential information. As Haberland (1996: 164) emphasizes, for Malinowski, the main contrast is between 'communion' and 'communication':

Are words in phatic communion used primarily to convey meaning, the meaning which is symbolically theirs? Certainly not!... A mere phrase of politeness . . . fulfils a function to which the meaning of its words is almost completely irrelevant. Inquiries about health, comments on weather, affirmations of some supremely obvious state of things - all such are exchanged, not in order to inform, not in this case to connect people in action, certainly not in order to express any thought.... [They serve] to establish bonds of personal union

9 between people through together by the mere need of companionship. (1972: 151)

Malinowski's interpretation ofphatic communion appears to be somewhat contradictory - talk is aimless, uninteresting, sometimes suspect, and even irrelevant, but it is part of a process satisfying the human need for social interaction and for personal connection. But even so, the main concern expressed in his interpretation is quite clear. As Coupland and Coupland put it, phatic communion may be best understood as "some sort of minimalist communicative practice, thought along several possible dimensions" (1992: 210), and "a constellation of interactional goals that are potentially relevant to all contexts ofhuman interchange." (1992: 211).

Inspired by Malinowski's work, a number of studies have appeared which discuss phatic communion or small talk. In his well-known 'channel' model of communication, Jakobson (1960) describes the function of small talk as maintaining an open channel of communication between participants (the 'phatic' function). While examining functions of phatic communion among acquaintances in casual conversation settings, Laver (1975, 1981) views phatic talk, especially at the margin of conversations (e.g. greetings, partings), as means of establishing and maintaining social relationships and means of achieving transitions within and out of more content oriented talk. He argues that these formulaic phrases of greetings or partings can signal a speaker's perception of the formality of a situation and his/her relationship with the hearer (1981: 301 ).

As a notable figure in the post-Malinowski era, Laver's largest contribution is that he applies the concept of phatic communion to the demonstration of the verbal ordering of conversational openings and closings and the explanation of these sequences. He introduces the structures of the linguistic routines of phatic communion in the initial and end phases (although they are simple) and specifies how speakers claim solidarity, intimacy and status relationship through particular choices of formulaic expressions (Laver 1975: 224-225). These insights, as pointed out by Coupland et al. (1992: 212), would seem to shed light on Brown and Levinson's (1978, 1987) politeness model

10 which suggests that conversation reveals cultural conventions and demonstrates respect between participants and the management of relationships.

Schneider (1988) was the first person to do an extensive study on the genre of small talk as phatic communion occurring in several common situations (the chance meeting of strangers). Under the framework of politeness, Schneider provides two basic functions of small talk. One is to address positive face needs: fulfilling a socialization function as in Malinowski's phatic function. The other is to address negative face needs: avoiding the embarrassment or 'danger' of silence with a stranger. The significance of Schneider's work is that he focuses on a description of forms, structures and topics rather than explaining the phatic function carried by the small talk. He notes that typical small talk consists of a series of conversational segments, each of which follows a sequence of moves. The findings of this study need to be understood as particularly important since they have shifted the focus of research from the examination of phatic communion within a single utterance to that within sequential contexts.

Coupland and Coupland (1992) conducted a study of the phatic process in elderly people's responses to the phrase 'how are you' in interviews about their medical experience. They found that the interviewees provided a range of responses from the purely phatic 'Fine, thank you' to the purely task-oriented 'I have asthma'. The authors conclude that small talk is a style of interaction which fulfils the function of 'phaticity', and that the level of 'phaticity' of small talk varies based on the current needs of the addressers to manage their interpersonal relationship with the addressees. Coupland and Coupland's study offers an important demonstration that phatic talk does not exist in discrete frames only - all talk can be characterized by varying levels of phaticity.

In his 1999 study, Bickmore investigates small talk embedded in text-based computational conversation between a real estate agent and a customer. While analysing the processes of the small talk through the 'chatterbot' systems, the researcher finds that text-based casual conversation is conducted with segments and moves as reported in Schneider's study (1988). The conversation moves from the greeting frame to the task frame, and then the farewell frame. Bickmore (1999: 10) insists that

11 small talk plays an important role in management of the channel of communication and interpersonal relationship between interactants. It can be seen as a part of deliberately planned communicative behaviour, which helps achieve the task-oriented goals.

A more recent study is presented in Pavlidou's contrastive analysis of the use ofphatic talk in Greek and German telephone calls between relatives and friends (2000). The author finds evidence that supports Laver's (1975, 1981) proposal: Phatic utterances, such as ritual questions (e.g. 'How are you?'), ritual expressions of good wishes (e.g. 'Happy birthday!') and a certain phatic particle (e.g. 'Na'), occur more in the talk where people call each other primarily for social rather than transactional purposes. He also points out the limitation of Laver's (and others') conception ofphatic communion, as it is restricted to a certain "set of stereotyped phrases of greeting, parting, commonplace remark about the weather, and small talk" (Laver 1975: 218). Pavlidou (2000: 138) claims that in real life other means such as the phatic particle 'Na' in German, and a playful use of the 'V' (vour) form and the repetition of simple greeting formulae also have phatic function. The results of this study are interesting because they include some additional linguistic means as phatic devices for facilitating the management of interpersonal relationships.

From the literature surveyed above, phatic communion can be seen as "a constellation of interactional goals that are potentially relevant to all contexts of human interchange" (Coupland and Coupland 1992: 211). It has significant social functions of"helping to smooth interaction and connecting people" (Tracy & Naughton 2000: 63). This functional approach, which emphasizes "relational goals and a relatively low level of participants' attendance to the seriousness or factuality of talk" (Jaworski 2000: 113), is important as a way of investigating some linguistic practices mainly for "doing sociability" (Coupland & Coupland 1992: 212) in oral conversation. It is also useful to explore people's routinized behaviours in written communication (e.g. personal letters). That is because personal written communication, for example, comprises a set of routinized expressions (e.g. in-group terms of address, openings or closings), which can be seen as small talk (Coulmas 1981). The writers' relational goals may be achieved via these routines, that is, maintaining rapport between writer and recipient (Coates 1988), and "establishing norms of acceptable moral conduct of its members" (Jaworski

12 2000: 112). These are important aspects of phatic communion embodied in certain routines of written communication. A survey of the literature, however, may suggest that there is little exploration into how phatic aspects are achieved through certain routines in written communication, specifically in the personal area. Considering this, research in relation to this question is required.

2.1.2 Speech act theory Another important notion in the study of language is speech act theory. Speech act theory addresses questions about the nature and the use of language, such as "What kind of actions do we perform in language?" and "How does an utterance have meaning?" (Winograd & Flores 1986: 32). These questions and theories of speech acts are relevant to this study, because the use of some expressions (e.g. greetings, longing for future contact) in openings or closings of personal letters have the function of performing speech acts, and the meanings of the routine expressions directly convey the writer's intention and the social functions of those acts. Using Bach's words, the concept of speech acts should be regarded as a general term for any type of language use, spoken or written. That is because speech acts, no matter what medium they are using, "fall under the broad category of intentional action, with which they share certain general features" (1998: 82). Therefore, this study will use speech act theory as one of the theoretical frameworks.

Originating in the philosophy of language, speech act theory emphasizes the communicative function of speech. As Searle (1969: 16) claims, "speech acts are the basic or minimal units of linguistic communication." For the present purpose, the important aspects of this theory include the notion of illocutionary acts, the intentionality of speech acts, the expressive illocutionary acts and the indirect act.

According to Austin's (1962) speech act theory, the illocutionary act refers to the conventional force uttered by words in a particular context. Making statements, asking questions, apologising, thanking, congratulating and so forth are regarded as illocutionary acts (Searle, Kiefer & Bierwisch 1980: vii). Central to the illocutionary acts is the notion of intentionality. Mey (200 1: 93) notes that speech acts are not

13 produced through the philosopher's thinking, but in its use in a real situation. As Searle aptly states:

When I make a noise or a mark on a piece of paper to be an instance of linguistic communication, as a message, one of the things I must assume is that the noise or mark was produced by a being or beings more or less like myself and produced with certain kinds of intentions. (Searle 1969: 16)

The significance of Searle's approach is that he provides an account of language use as an intentional action, that is, when speakers are performing an illocutionary act, they are trying to do things with their words. For example, when one expresses concern to the recipient in a personal letter, such as 'Chunhan douqiao, wang duo zhuyi shenti' (In the chilly air of early spring, I hope you take more care of your health), he bears the goal in mind that he hopes the recipient will look after himself/herself well. Why is intention crucial in speech acts? It is crucial because illocutionary acts are initiated by speakers, and their social functions are mostly exhibited through the speakers' intention (Norrick 1978: 280). Taking the preceding case for example again, by expressing concern, the writer intends to convey considerateness to the recipient. The perlocutionary effect (Austin 1962) of the considerateness would be getting the recipient to believe the writer's positive feelings and attitudes. Generally speaking, the social function of an illocutionary act necessarily includes its perlocutionary effect (Norrick 1978: 280). A speaker's intention and its effect on the hearer cannot be completely separated: Some illocutionary acts can be regarded as a means of achieving a particular goal (Holtgraves 2002: 17), or "may go beyond it" (Norrick 1978: 280).

However, intentionality is not just a simple matter of intention belonging to a particular speaker. For Searle, the main issues for this aspect involve how to establish the conditions for a particular act (e.g. sincerity, felicity, etc.), and how a speech act functions in a context. The latter is very much relevant to the present study.

According to Searle (1969), all speech acts are situated acts. They rely on the situation in which they are performed. From this point of view, a situated speech act is entirely associated with what has been called a speech event in ethnographic and anthropological studies (Bauman & Sherzer 1974). Speech is embodied in an

14 institutionalized social activity of a certain kind, such as addressing practice, openings and closings in interaction. In such situations, according to Mey (200 1: 219), certain utterances can be expected and will then be accepted. It is the acceptance of the speaker and the hearer's utterance that may establish and reaffirm the social events (behaviour) in which the utterances are delivered. Mey's approach emphasizes that "the individual speech acts [e.g. greetings] make sense only in the events [e.g. openings]" (2001: 219). Hank (1996: 266) holds a similar view that "meaning arises out of the interaction between language and circumstances, rather than being encapsulated in the language itself."

For its success, a speech act is dependent not only on a condition and a particular situation, but also on the larger cultural context. This implies that while looking at speech acts, one should consider the given cultural conditions, "which allow, and afford, a particular act of speaking" (Mey 2001: 94). For example, in Chinese culture, when people initiate letter writing, they, especially intimates, often make inquiries about the recipients' state of health and of other personal affairs. Such inquiries, viewed as a display of one's kindness, have highly social-cultural values (Zhan 1992: 9). Thus, the choices of speech acts embodied in individual languages not only reflect 'natural logic,' but also demonstrate what Gumperz refers to as 'cultural logic,' which is equivalent to "social norms as extralinguistic forces which merely determine how and under what circumstances such meaning units are used" (1982: 185-186).

Speech act theory makes it clear that people are doing something with words (phrases, sentences) in a particular context, and in society. It also develops a classification of illocutionary acts into certain basic categories. According to Searle (1976), there are five basic and primitive illocutionary points: 1) directives, attempts to get the hearer to do something (e.g. requesting, advising); 2) assertives, attempts to commit the speaker to something being the case (e.g. suggesting, swearing); 3) commisives, attempts to commit the speaker to doing something in the future (e.g. promising, planning); 4) declaratives, attempts to change the state of the world in an immediate way (e.g. you are fired); 5) expressives, attempts to express how the speaker feels about the situation, (e.g. thanking, apologizing). The fifth category - expressives can be applied to the speech acts to be investigated in this study.

15 The expressive act, according to Searle (1976: 12), refers to an attempt "to express the psychological state specified in the sincerity condition about a state of affairs specified in the propositional content." Under these defining properties of expressives, Searle lists a range of expressive acts: apologizing, thanking, congratulating, condoling, deploring, and welcoming. Norrick argues that Searle's list cannot be considered exhaustive. He claims that what the speaker intends to achieve by performing an expressive act depends on its illocutionary point of expressing an emotion (1978: 278). Based on this assumption, Norrick adds three categories - lamenting, forgiving and boasting into this type of acts. In addition, he discusses various conditions on expressive acts, such as the factive, the value judgement and the role identification condition, by which people can differentiate the various expressive acts. It is important to note that this provides the basis of an analysis which is extended to the discussion of the social functions of expressive acts.

While examining the role of the emotional message in arguments, Gilbert (1999) treats expressive acts as emotional message acts. He states that an emotional message act involves an expression of emotion "that is identifiable to the recipient or observer" (1999: 5). According to him, in real life there are many emotions such as anger and boredom, that can be identified through language, and these emotions vary widely with different cultures.

One important characteristic of the emotional message (expressives) is its actual demonstration of emotional content. From this point of view, emotional message acts can be of several types (Gilbert 1999: 5). If applying this approach to speech acts in personal written communication, it might be that an expression of self-reassurance or a phrase of longing for future contact can be an emotional act. That is because both acts reflect the writer's particular feelings and emotions (e.g. concern, solidary). One significant point in Gilbert's study is that he emphasizes that judgement of an emotional message act is dependent on what kind of emotion is expressed, and also on what the goal of the addresser is by using an emotional act (e.g. making the recipient relaxed or happy).

16 Another important aspect of Searle's theory is the indirect act. According to Searle (1975), in ordinary conversation a certain set of speech acts is typically performed in indirect ways, such as requesting and offering. Searle also states that these forms have become "conventionally established as the standard idiomatic forms for indirect speech acts," that is, "while keeping their literal meanings, they will acquire conventional uses as, e.g. polite forms for requests" (Searle1975: 76). One of the most interesting aspects of this approach is that it links the mechanism for performing indirect acts with a manifestation of conventionality and politeness. For Searle, indirection is often used as a means of showing politeness and a device which helps people resort to conversational implicature. Such views contribute to working out the speaker's intentions from what s/he actually says and why the speaker has recourse to this device. This is useful for accounting for a wide variety of indirect speech acts, such as those involved in this study (e.g. pre-closing phrases).

Searle (1975: 76) also claims that the production of indirect speech acts may vary from one language to another, although he takes such differences as idiomatic due to accidental variation. Searle's approach provides insights for characterising certain indirect acts in particular contexts, and in cross-cultural situations in which the concept of face and the management of politeness are different. It leads to the appearance of empirical studies such as the Cross Cultural Speech Act Realization Project (CCSARP) (Blum-Kulka, House, and Kasper 1989). This project investigated and confirmed cultural differences and strategies for request and apologizing in German, Hebrew, Danish, Canadian French, Argentinean Spanish, British, American and Australian English. For instance, both American and Hebrews generally preferred to use on record H-oriented request rather than off-record ones. But the second preference for them was different: Hebrew liked to use S dominant direct request whereas Americans favoured the off-record strategy. In addition, while interacting with familiars, Slavs, Hebrews and Germans used direct on-record request mitigated with modal particles and diminutives, while English lacked such devices and therefore used fewer S dominant strategies. Explaining indirect acts differently from different languages, Wierzbicka (1991: 62) notes, in connection with Searle's view, specific differences in indirect speech acts between languages result from variations in "cultural norms and cultural assumptions."

17 In sum, speech act theory develops a framework for viewing language use as action (Holtgraves 2002). It explicitly identifies communicative actions and accounts for how the actions are performed in conversational interactions. What it does not offer is an explicit identification of speech acts in naturally occurring written interaction, such as personal written discourse. In particular, there is little information about what actions people are performing in some routine sections, or how these speech acts reveal the writer's intentions and what the perlocutionary effect is on the recipients. This seems to be an interesting point that researchers need to consider while investigating language use in a variety of human interactions. Specifically, researchers need to present a methodology for structuring and defining written speech acts.

2.1.3 Grice's Cooperative Principle The above review suggests that people use language to do things. To ensure a correct use of interactional language, people need to follow an accepted way of speaking, to follow what Grice (1975) calls the Cooperative Principle (CP). Grice (1989: 26) claims that when speakers intend to contribute to a conversation, they make cooperative efforts in which each participant recognizes a common purpose, or at least a mutually accepted direction. Grice's principle is based on the specific assumption that the purpose of conversation is a "maximally effective exchange of information" (1975: 49). Apparently, this principle can be applied to written communication of the type typically found in personal letters, because in this type of text, the communication, especially in a certain speech behaviour (i.e. openings), works in a sense with maximal social functions such as clarity and relevance. As argued by Holtgraves (2002: 25), "the CP and corresponding maxims are assumed to underlie rational interaction of any sort." According to Grice (1975: 47), there are four criteria in applying the Cooperative Principle:

Maxim of Quantity: Make your contribution as informative as required, but do not offer more information than is required

Maxim of Quality: Try to make your contribution true

Maxim ofRelation: Be relevant

Maxim ofManner: Avoid obscure expressions and ambiguity. Be brief and orderly

18 Grice's (1975) maxims explicate that in a conversation, one's utterances should be clear (Manner), truthful (Quality), relevant for the topic (Relation), and the right size (Quantity). It is reasonable for people to follow such maxims with the hope to achieve their interactional goals. Grice's maxims can be seen as an important contribution to speech act theory. They provide a framework for how speakers convey and hearers recognize the force of illocutionary acts (Holtgraves 2002). For the purpose of the present study, two important conversational maxims that speakers pay attention to will be looked at closely: the maxims of Quantity and ofRelation.

Following Grice, Horn (1984) introduces two functional economies of Zipf and Martinet 'The Principle ofLeast efforts' and 'The Principle of Sufficient Efforts' which he identifies with Grice's maxim of Quantity as follows:

The Q Principle (Hearer-based): Make your contribution sufficient and say as much as you can.

The R Principle (Speaker-based): Make your contribution necessary and say no more than you must.

Horn (1984: 12) maintains that the maxim of Quantity is important in conversation. It emphasizes the need for the speaker to convey his message fully. Horn also notes that when defining Quantity, Grice builds in the maxim of Relation: "what would make a contribution more informative than required, except the inclusion of materials not strictly relevant to and needed for the matter at hand?" On the basis of these considerations, the maxim of Quantity can turn out to be: Let the message that speakers (or writers) mean to convey as sufficient and related as needed. Horn's approach to the principle of Sufficiency derived from Grice's 'Relation' and 'Manner' can be applied to the analysis of the data, particularly to some combined openings for this present study.

In the combined opemngs provided here, one father places an inquiry before an expression of hope while writing to his son: 'The final examinations are approaching.

19 Have you been busy with study? I hope you are studying harder and trying to get good marks'. Generally speaking, the inquiry in this combination appears to function as a preface (e.g. pointing out the situation), which is used to invite the writer's next expressive act: hope, the main concern of the writer. Constructing openings like this, on the one hand, may enable the expressions to be of sufficient length and clarity in terms of content. On the other hand, it may display the writer's reasoning and performance of speech acts to "transmit their mental states to other agents" (Moulin & Rousseau 2002: 177). In other words, it reflects the rhetoric of a language. According to Kirkpatrick (1991: 184), rhetoric is concerned with "how language is used to persuade people." In the present case, by using an inquiry the father indexes why he expresses hope for his son, here laying the ground for the emergence of the expression of hope.

In fact, organizing the opening with more than one strategy also reflects the influence of Confucian philosophy upon speech, that is, the need to be clear by saying more, a principle encoded in Grice's maxim of Quantity and Hom's principle of sufficiency. While giving advice on the style of formality, Confucius states:

If the speeches are too long, they sound artificial. If they are too short, they fail to convey the speaker's meaning. The perfection of the speaking art is to make speeches that convey the speaker's meaning and not more (Oliver 1971: 92)

Viewing this statement, it seems that Confucian traditional beliefs within the past history of China place great emphasis on clarity and relevance of verbal communication. This preference, according to Kirkpatrick (1991: 184), reflects the influence of the ways "in which a culture perceives the notions of politeness, power, and relevance." It is this clarity and relevance through which this linguistic behaviour is achieved by the writer that one can find the writer's intent in affecting meanings of combined strategies. Using Held's (1989) words, such clarity and relevance can be seen as a form of maximization of intent of the writer. From this view, the maxim of Quantity (also Relevance) is thus linked to the intentional force as well as the constitutional components of illocutionary acts for a certain type of linguistic behaviour in communication.

20 Apart from proposing effective ways of using language through the maxims, another significant feature of Grice's theory is its assumption of a mutually shared contextual background and information between the participants in an interaction. According to Grice (1975), an implicated meaning cannot be conveyed successfully if both the speaker and hearer do not share a common background other than a common knowledge of the conventional linguistic means. Contextual background knowledge refers to the selection and the use of appropriate socio-cultural knowledge as part of communicative competence (Lee-Wong 2000: 13). In this regard, Faerch and Kasper's (1984) comprehensive framework of what they term pragmatically relevant declarative knowledge is quite explicit:

Linguistic knowledge: Rules and items of the language

Speech act knowledge: Knowledge about what are possible verbal acts within a socio-cultural community and its institutions

Discourse knowledge: Knowledge pertaining to structural patterning m interactions

Socio-cultural knowledge: Knowledge about social values, norms, institutions, relevant relations between individuals, patterns of behaviour in a given society, e.g. interactional principles such as face-saving

Content knowledge: Knowledge about which of the features present in a given situation are relevant as context-determining factors

(1984: 215-216)

From this, it is assumed that contextual background knowledge includes conventions of both linguistic forms and social meanings. It deals with beliefs and values of the culture. It is also concerned with the matter of communicative competence by selecting relevant features in a specific context. From a socio-psychological point of view, to be a competent conversationalist, a speaker should take into account the relevant features which he/she thinks will influence the result of his/her utterances (Lee Wong 2000: 14). For example, initiating an end of written interaction is distinctly a goal-driven act, namely, it is the writer's desire to terminate the contact with the recipient. Which of the contextual features does a writer need to consider or

21 highlight? In Chinese culture, for instance, if a writer holds a high social position and is distant in relationship to the recipient (e.g. in a teacher-student relationship), would it not be right that this writer should convey a certain level of politeness to the recipient by using the indirect form of initiation (e.g. offering reasons for finishing the written contact) towards the low status recipient (e.g. student)? Or if the writer and recipient are socially equals and intimate in relationship (i.e. friend-friend), does the writer need to show the recipient as socially superior or junior to him/her while employing the initiating-ending statement? These are essentially questions of what the writer considers as relevant, since they will enable the writer to adopt a certain pattern or expression appropriate to the situation, and thus make the communication undertaken successfully.

Based on this, it is evident that relevant declarative knowledge plays an important role in meaning (social function) processing and the resultant implication. This notion is implicated in Sperber and Wilson's (1986) discussion of relevance- the relationship between the linguistic and the pragmatic aspect of utterance meaning. According to the authors, the principle of relevance has conditions that "an assumption is relevant in a context if and only if it has some contextual effect in that context" (1986: 122) and that "an assumption is relevant to an individual to the extent that the contextual effects achieved when it is optimally processed are large" (1986: 145). This, as Andersen (1988: 604) argues, "predicts correctly that even if communicator and audience know the same syntactic and semantic rules, miscommunication will occur insofar as they have different background knowledge and conventions of speech."

In summary, Grice's Cooperative Principle provides a mechanism by which the speakers convey and the hearers recognize illocutionary meanings. The four maxims and the participants' mutually assumed contextual knowledge are crucial in determining the participants' successful communication in conversations. It is with these features that Grice's framework turns out to be "a social psychological model for communication" (Holtgraves 2002). However, over the years, scholars have observed several weaknesses with Grice's theory. One obvious weakness is that Grice only provides an informal and partial account of these maxims in oral conversation and he does not justify them in other types of communication, such as written interaction.

22 Moreover, as argued by Vanderveken (2002: 14), Grice states his maxims by focusing on the exchange of information, that is, assertive utterances made in a special kind of interaction. He does not take account into phatic communion (Malinowski 1972), linguistic rituals (Firth 1972), or routines (Laver 1981 ), whose main social and affective goals are to maintain the relationship between the participants. There is little discussion about whether an adequate account of those phatic expressions or routines will require reference to the more information-oriented maxims derived from Grice's Cooperative Principle. Specifically, people know little about whether and how the maxims of Relevance and Quantity work within the phatic expressions embodied in certain linguistic behaviours in written discourse.

2.1.4 Concept of politeness Speech act theory provides a conceptual framework of language as social action: people do things with their words. However, as noted by Holtgraves (2002: 37), as interactants, they cannot simply perform speech acts without considering the interpersonal implications. To announce the end of contact or to apologize, for example, not only performs an expressive act, but also conducts an action with a clear implication for the identities of the participants. That is why one's verbal interactions with the other are not always direct, clear and maximally efficient, which violates Grice's four maxims (1975). Instead, people speak or write indirectly and gently in particular contexts. All these issues are related to and can be answered with reference to another motivator of communication in people's life: the need to be polite (Laver 1981: 289).

Politeness is crucially a universal phenomenon in the course of social interaction between people. It refers to "the various ways in which polite behaviour is perceived and talked about by members of socio-cultural groups" (Watts, Ide & Ehlich 1992: 3). It concerns the relation between forms of language usage and the processes of socialisation and consequent social behaviour. It is interpreted as one of the constraints on the use of language (Gu 1990: 220). In the following discussion, some key politeness theories or principles which are relevant to this study will be briefly reviewed.

23 2.1.4.1 Brown and Levinson's model of politeness Among the studies of politeness theory thus far, Brown and Levinson's politeness model is considered the most comprehensive and influential (see Fraser 1990; Thomas 1995). What makes their model so influential is that, according to Mey (2002: 11), they focus on the function of linguistic expressions rather than looking at the politeness of linguistic expressions. They have extended the concept of politeness by emphasizing the function of utterances in relation to human relationships, which leaves space for contrasting many languages that differ grammatically. In addition, Brown and Levinson's model provides a verbal way to "relieve the interpersonal tension, which arises from communicative intentions that conflict with social needs and statuses" (Mayumi 2002: 11).

A key notion in Brown and Levinson's politeness theory is that of 'face', which can be traced back to Gof:finan's (1967, 1972) work on deference and his notion of face. According to Brown and Levinson (1978, 1987), people possess two kinds of face: positive face- the want that one's own wants be desirable by others; negative face- the want that one's own action not to be impinged upon by others (1978: 62). Within Brown and Levinson's (1987) approach, politeness is perceived as an instrumental act. It is used to address the interactants' need to maintain or remove threats to positive or negative face. This view implies that most speech acts inherently threaten either the hearer's or the speaker's face-wants, and politeness is involved in redressing those face threats. Brown and Levinson (1987) distinguish four main strategies for performing speech acts as follows:

Bald-on-record: the direct way of saying things

Positive politeness: the expression of solidarity, attending to the hearer's positive face-wants (e.g. claiming common ground)

Negative politeness: the expression of restraint, attending to the hearer's negative face-wants (e.g. being conventionally indirect)

Off-record politeness: the avoidance ofunequivocal imposition (e.g. using hints)

Brown and Levinson's typology of four superstrategies is an attempt to reveal some differences of ways in which people show politeness by performing an act. These

24 superstrategies are ordered along a continuum of politeness according to the extent to which face-work is encoded in the utterance. For their politeness formula, Brown and Levinson (1987) postulate three social variables: P (the perceived power difference between hearer and speaker), D (the perceived social distance between them) and R (the cultural ranking of the speech act - how threatening or dangerous it is perceived in a given culture). In Brown and Levinson's view, each of these has an individual effect on the choices of politeness strategies by participants. While keeping these three variables as very general social dimensions, Brown and Levinson (1987) also emphasise that their main purpose is not to specify what factors are involved in estimating these complex parameters because such factors are certainly culturally-specific. For example, the degree of imposition of an act can vary through settings and most importantly, across cultures. What may be regarded as relatively imposing in one culture (e.g. playing with food in the United States) tends to be regarded as unimposing in another culture (e.g. China) (Chen 1996).

Brown and Levinson (1987) discuss several syntactic strategies to achieve positive and negative politeness. Positive politeness strategies are defined as an attempt to meet positive face wants, such as noticing or attending to the addressee's interest, using in group identity markers, or claiming common ground, etc. (Brown & Levinson 1987). Negative politeness strategies are defined as an attempt to meet negative face wants, such as using indirect speech acts, giving deference, apologizing (Brown & Levinson 1987). The concept of positive and negative politeness strategies is one of the important aspects of this theory of politeness. It provides a comprehensive framework for researchers to investigate complex politeness phenomena in different contexts and different cultures.

2.1.4.2 Studies inspired by Brown and Levinson's model of politeness Brown and Levinson's politeness model, as evaluated by Brown (1990), can be regarded as an important intellectual work that has been empirically supported by researchers. However, there are a number of criticisms that have been made of their theory. One common criticism concerns the universality of the concepts of 'face' and politeness as face-saving strategies. Many scholars have pointed out that Brown and Levinson's politeness model lacks consideration of cultural variations (e.g. Blurn-Kulka

25 1990; Chen 1993; Fitch & Sanders 1994; Fraser 1990; Gu 1990; Ide 1989; Kasper 1990, Koo 1995; Mao 1994; Matsumoto 1988). For example, Matsumoto (1988) finds that in Japanese culture, participants pay more attention to their relationships instead of emphasizing individual rights. Thus, negative face want is not important in their culture. Gu (1990) maintains that the Chinese notion of face is not one in relation to psychological wants, but rather to societal norms. Hence, offering, inviting and promising which are regarded as threatening the hearer's face by Brown and Levinson are not seen as impeding the hearer's freedom in Chinese context. Gu (1990: 243-245) claims that the Chinese concept of politeness is characterized as moralizing to some extent. Therefore, the politeness principle is a "sanctioned belief that an individual's social behaviour ought to live up to the expectation of respectfulness, modesty, attitudinal warmth and refinement." Gu also stresses that since ancient times, politeness in Chinese culture aims to "lay down rules of conduct", and the "rules for politeness are moral maxims, the breach of which will incur social sanctions" (1990: 239-240).

A second frequently raised challenge is related to the determinants of politeness. Some researchers have argued that other dimensions need to be considered. For example, they have proposed that the distance factor should include several independent factors, such as affect and familiarity (Blum-Kulka 1990; Brown & Gilman 1989; Forgas 1999; Fraser 1990; Koo 1995; Tracy 1990). In the study of politeness in four Shakespearean tragedies, Brown and Gilman (1989) find that affect strongly influences politeness. The increase in affect results in the increase of politeness. While exam1rung parental politeness in family discourse among Israeli, American and American immigrant families, Blum-Kulka (1990: 267) notices that three key notions (power, informality and affect) interact in setting the degree of politeness. The factor of affect which is often revealed through the use of mitigation, for example, plays an important role in family discourse, serving to confirm in-group membership and to stress emotional bonds between speaker and hearer.

Another argument deals with Brown and Levinson's lack of examination of politeness in some 'ordinary conversation', in which there seems to be lack of a face-threatening situation (Mayumi 2002: 27). Matsumoto (1989: 208) argues that English speakers

26 can say 'Today is Saturday' in the same form to anyone, while Japanese people cannot use the same form of the sentence in all situations. They must choose one in accordance with the given context, that is, although there is no a face-threatening-act (FTA) in uttering one sentence, Japanese speakers are still careful in the choice of the linguistic pattern. This, as noted by Shih (1986), predicts that politeness should not be viewed merely as a conflict avoidance strategy (face-saving), but also positively as a harmony and friendship enforcement strategy. Shih (1986: 22) states that some speech acts, such as making an offer or complimenting can be seen as a face-satisfying act (FSA), thus making the addressee happy. She also points out that if one performs an FSA sincerely (e.g. to inquire after the addressee's personal states), he will be regarded as behaving politely. The absence of doing an FSA may be thought of as impolite behaviour. According to Shih, it is the combination of the concept of face-satisfying act with face-threatening-act that makes the whole politeness framework more balanced (1986: 24).

To sum up, politeness theories have provided a basic framework for examining how people attempt to convey their identities and relationships with the other through the manipulation of politeness (Holtgraves 2002: 62). However, so far most sociolinguistic studies of politeness phenomena focus on exploring politeness features within a single speech act or a move (e.g. apology, greeting). There is limited investigation of politeness within the domain of discourse. In other words, little work has been done in relation to how the level of politeness is normatively and strategically expressed through a series of speech acts involved in certain speech behaviours in various types of interactions (e.g. personal written interaction). Thus, people have little information about how the participant conveys politeness by performing certain routine behaviours (e.g. openings, closings), which are achieved through a series of speech acts and about the ways in which these acts are related to each other under the politeness framework. For better understanding of the complexity of politeness phenomena in human interaction, people need to consider the ways in which politeness is manifested through the interactional discourse (e.g. combined speech acts in routine behaviours), which the later chapters intend to focus on.

27 2.1.5 Language and social variables 2.1.5.1 Power and solidarity Besides considering how language helps people do something and how language is managed effectively and politely in ordinary life, another aspect to consider is the relation between language and society (Hudson 1996), that is how certain linguistic variables (e.g. forms) are related to certain social variables (e.g. age, social rank, group membership). For the present purpose, two important variables are power and solidarity (although this study uses variants of these, e.g. seniority, closeness). Power, as termed by Brown and Gilman ( 1960), refers to authority or superiority of one person over another. It is fostered by social factors such as age or social rank. It is therefore non-reciprocal. In this study, those with higher social position are called superiors (e.g. teachers), and those with lower social position are called 'inferiors'. Those who are older in age are called 'seniors,' and those who are younger in age are called 'juniors.' Solidarity, by contrast, is "invoked between equals, people who are close or have a certain level of intimacy" (Oyetade 1995: 516). It deals with the "desire to be friendly and create rapport" (Tannen 1986). It is therefore inherently reciprocal. According to Hudson, power is "self-explanatory," but solidarity is concerned with social distance between participants, which is related to "how much experience they have shared; how many social characteristics they share (e.g. religion, sex, age, region of origin, race, occupation, interest, etc.) and how far they are prepared to share intimacies, and other factors" (1996: 122).

In their classic paper on linguistic markers of social relations, Brown and Gilman (1960) assume that power is associated with asymmetrical relationships. In an interaction, "one person may be said to have power over another in the degree that he is able to control the behaviour of the other" (1960: 255). Therefore, according to Brown and Gilman (1960: 255), power is "non-reciprocal in the sense that both cannot have power in the same area of behaviour." For example, superiors give T ('tu', meaning familiar/singular 'you') to inferiors and receive V ('vous', meaning respectfuVplural 'you') from them. Between equals, the pronominal address is reciprocal where the individual gives and receive the same address form. Further, Brown (1965) describes these two social variables in the invariant norm of address in English, which formulates the universal relationship between status and solidarity. He finds that the forms of

28 address used by superiors in an asymmetrical relationship can also be used between intimate equals, whereas the forms of address used by the inferiors of an unequal dyad are also the forms reciprocally used between distant equals. This result reflects the interdependence of status and solidarity and suggests the ways in which the difference and the change of status and interpersonal relationship are spelt out.

Following Brown and Gilman, a great number of theorists and researchers have examined more closely the influence of power and solidarity on both address forms and speech act realisation modes in different languages. For example, Slobin, Miller and Porter (1968) consider the association of forms of address in American English to status, familiarity, intimacy and personality in an organizational context. In studies of the request act, the role of power and solidarity is also reported in Ervin-Tripp (197 6), Scollon and Scollon (1983), Blum-Kulka, Danet and Gerson (1985) and Lee-Wong (2000). Their findings demonstrate that requests in American English differ with social distance: Members within a family and friends are more direct than strangers (Ervin-Tripp 1976), and indirectness is found to increase with social distance and decrease with social power (Scollon & Scollon 1983). Also, Wolfson (1992) attributes a high frequency of complimenting which occurred amongst status-equal friends, co workers, and acquaintances in middle class urban American society to the reflection of social relationships in that society. All these explorations thus clearly reflect the interdependence of social status and interpersonal relationship with linguistic variations in frequency and means of speech performance.

As studies in this field have grown in numbers, so has the exploration of the interaction between power and solidarity. As argued by Spencer-Oatey (1996), it is hard to draw a clear-cut division between these two concepts. Using Tannen's words, "power and solidarity are in paradoxical relation to each other. That is, although power and solidarity, closeness and distance, seem at first to be opposites, each also entails the other" (1994: 22). In his study of naming practice and the power of words in China, Blum (1997: 372) suggests that relative status in China is an important factor in determining which term will be used in naming practice. But he also claims that at times hierarchy does not prevent close attachment being formed. Affection would be displayed through a certain naming pattern by an addresser who has controlling power.

29 According to Blum (1997: 372), when addressing one's closest intimate, the addresser may choose the term that embodies the proper relationship, which is hierarchical, but not distant. Similarly, Blum-Kulka's (1990) study of family interaction suggests that the prevailing style of parental discourse is both highly status-oriented and affection involved. This evidence clearly shows that power and solidarity tend strongly to be expressed by the same form or pattern (Hudson 1996: 124).

Apart from observing the interaction of power and solidarity in human interactions, researchers also find that every language might be expected to have its own ways of singling differences in terms of power or solidarity. As pointed out by Blum-Kulka and House (1989: 137), members of different cultures may differ in their perception of the relative importance attributed to these two dimensions. For example, Wetzel (1993) finds that in western countries the term 'power' is closely linked to dominance and control, thus containing negative meanings. However, for some other cultures, this concept is often associated with different social values. In both China and Japan, for example, due to the influence of Confucian ethics in interpersonal relationships (' Wu Lun '- Five Cardinal Human Relationships), interactants of unequal dyads such as in a teacher-student relationship, may feel mutual ties and responsibility to each other, somewhat similar to that between a parent and child. Superiors in such a context do not simply dominate or control inferiors. Rather, the pairs are bonded together by a role relationship which involves hierarchy and also mutual responsibilities. Thus, it suggests that the performance of interlocutors is also determined by their own cultural values and norms, along with these two basic social variables.

2.1.5.2 Language and gender Gender is another important social variable influencing the linguistic variations. Over the past few decades, there has been a flood of research on gender-based differences in language (e.g. Cameron 1990; Chambers 1992; Goodwin 1998; Holmes 1995; Labov 1972a, 1972b; Lakoff 1973, 1975; Tannen 1984, 1990, 1992; Trudgill 1972; West & Zimmerman 1983; and Zimin 1981 ). These studies suggest that men and women do use language differently and that they have different priorities and purposes in conversation. Of the gender variations in language use, according to researchers, some differences are gender-exclusive, but some are gender-preferential forms. It is

30 commonly accepted that women prefer the more correct and polite forms (Zimin 1981: 35). Tannen's (1990) work provides a typical example of the gender-based differences in language. She maintains that the general difference seems to be that men are more concerned with power and women with solidarity:

For men:

Conversations are negotiations in which people try to achieve and maintain the upper hand if they can, and protect themselves from others' attempts to put them down and push them around. Life, then, is a contest, a struggle to preserve independence and avoid failure.

But for women:

Conversations are negotiations for closeness in which people try to seek and give confirmation and support, and to reach consensus. They try to protect themselves from others' attempts to push them away. Life, then, is a community, a struggle to preserve intimacy and avoid isolation. Though there are hierarchies in this world too, they are hierarchies more of friendship than of power and accomplishment. (Tannen 1990: 24-25)

This general gender-based difference explains a number of differences in linguistic behaviours, though the evidence from empirical studies is still scant. One common gender difference in language is related to speech style. Early research, such as Lakoffs paper "Language and Women's Place" (1975), largely characterized women's speech styles as uncertain, unimportant and weak, and men's speech styles as assertive, clear and adult-like. In later works, these styles of women's language have been described perhaps more accurately as deferential, and as collaborative within the social context in which they are displayed (Zimin 1981: 36). Gender-preferences have also been shown to exist in the use and perception of the lexical items. For example, Lakoff (1975) finds that women use more precise colour descriptions (e.g. mauve, lavender), more 'empty' adjectives (e.g. divine, lovely) and weak expletives (e.g. 'dear me') than men. At the syntactic level, women use more tag questions (e.g. 'John is here, isn't he?') and hedges (e.g. 'sort of, 'you know') than men, signalling uncertainty.

31 These particular aspects of female speech might suggest that in their language use women tend to show more concern about being 'collaborative' towards other females and males. Lakoff maintains that such gender differences in language usage reflect not only different roles, but also inequality of roles and status. She further adds that the inferior status of women in society can be seen as the reason for the observed differences between women's and men's language styles.

Ever since Lakoffs work exploring gender differences in the use oflanguage (1975), a number of scholars have further examined the verbal communication styles of women and men. Quina, Wingard, and Bate (1987) confirm Lakoffs feminine language, assuming that women's speech contains higher social warmth than masculine language. Tannen (1990), too, indicates that women's communication tends to be more supportive and rapport-building, while men typically are more report-giving and informative.

Holmes (1984) investigated gender differences in tag questions based on a New Zealand linguistic corpus. She distinguishes between two types of tags: 'modal tags' which request confirmation of information, and 'affective tags' which indicate the writer's concern for the addressee. Holmes finds that affective tags are used more by women while the modal tags are used more by men, thus suggesting that the preference for affective tags by women reflects their attendance to co-operation in performing speech acts. In her follow-up studies of complimenting and apologizing acts in New Zealand data, Holmes concludes that women's linguistic behaviour can be broadly characterized as "facilitative, affiliative, and cooperative, rather than competitive or control-oriented" (1995: 25).

The examination of the relationship of language and gender is also revealed in Brown's study of women and men in the Mayan community in Mexico (1998). In her paper she tests the hypothesis that women are more polite than men. She focuses her study in the investigation of a class of particles in Tzeltal (the language used in the community), which modifies the illocutionary force of the speech act. Her analysis shows that women and men differ systematically in the selection of strategies and in doing face work: Women prefer to use the extremes of positive and negative politeness strategies,

32 while men like to speak in a more factual-oriented manner. Interpreting these results, Brown (1998: 95) states that women pay more attention to the social power of the addressee and social distance between them than men, hence their overall level of politeness is higher than men.

It is clear that over the past few decades, there has been a rich body of studies in which researchers have either tested Lakoffs claims, or extended her ideas of gender differences in speech by including some features that she did not address. However, despite a great deal of research on the influenye of gender in language styles, people still find many gender-linked questions to be umesolved. As suggested by Simkins Bullock and Wildman (1991: 149), "perhaps the most fundamental inconsistency is the lack of agreement about whether males and females use language differently." McElhinny's study of the interactional styles of female and male police officers in Pittsburgh better illustrates this assumption (1995). In her analysis of police officers responding to domestic violence calls, she suggests that female police officers adopt a style that is similar to that of young male police officers. These results provide evidence for the influence of social practices and activities on the relationship between gender and speech. Such inconsistency in results means that further research should take into account the sub-roles (e.g. social practices and communicative goals), as well as the social status and the social structure in the study of gender differences.

2.1.6 Summary This section has reviewed a range of sociolinguistic and socio-cultural literature pertaining to phatic communion, speech act theory, Grice's Cooperative Principle, the concept of politeness and the relation between language and social variables. The selected literature has shown that language as social action requires a consideration of the speech act performed with an utterance, as well as the interactional and interpersonal implications of performing the speech act in a particular context. In addition to social functions, the literature has also reported how language signals social characteristics of the speaker and hearer (e.g. status, age, gender), and of the relationship between them (e.g. intimate, distant). Obviously the issues surveyed in the literature are of enormous importance and interest as far as human beings' communication is concerned in a general sense. However, there are still a number of

33 potential problems unresolved in those aspects. One of the major weaknesses, as was mentioned previously, is that most of the literature focuses on face-to-face verbal communication, but relatively little deals with written language. Another obvious problem is that much of the literature has tended to regard the single tum of talk. There is little information about how phatic communion, speech acts, politeness strategies work in a series of moves by a particular individual in a particular context. Moreover, none of the conceptual frameworks reviewed above has much to say about how speech acts are structured or the manner in which speech acts are related to each other. It is those unresolved issues that have partially motivated the present study.

2.2 Studies of address forms, openings and closings in interaction This section will review a selection of socio-linguistics studies that has reported on the addressing, and opening and closing behaviour in people's verbal communication. This is to understand the properties, procedures, and the extent to which social factors influence the performance of each aspect and to which socio-cultural values and norms underpin these interactive behaviours. It aims to provide a basis upon which these three behaviours might be observed when they are working in other type of interaction.

2.2.1 Address forms 2.2.1.1 Concept and functions According to Braun (1988), the term address denotes the addresser's linguistic reference to his/her addressee. Using Oyetade's words, "address forms are words or expressions used to designate the person being talked to while talk is in progress" (1995: 515). In most languages address forms are focused on three words classes: pronouns, verbs, and nouns (Braun 1988: 7). According to Braun's terminology (1988), the pronouns of address are pronouns referring to the persons being addressed. In general, these are second person pronouns such as English 'You' or French 'Tu' and 'Vous'. He also treats some other grammatical persons as pronouns of address, particularly when they refer to the communication partner such as German 'Sie' (third person plural) or Danish 'De' (third person plural). Nouns of address refer to substantives and adjectives which designate addressees or refer to them in other ways. This class comprises the most diverse types. One of the commonest types is a person's name, which belongs to the nominal repertory of address in the vast majority of

34 languages. The second frequent kind is terms of kinship, which refers to terms for blood relations. There are also some other nominal variants such as general forms which are not regarded as particular titles but are in common use (e.g. Mr, Mrs), titles which refer to one's bestowed title (e.g. doctor, major), or designate an addressee's profession or function (e.g. teacher, waiter), and words for certain type of relationship (e.g. friend, neighbour).

The study of address forms has always been a popular topic within sociolinguists. Through a functional point of view, address forms have been treated as pragmatic initiators of communication, and as a first step for guaranteeing the interchange in the course of communication (Zhao 1997: 14). They are "one of the areas of sociolinguistics which can best illustrate the close bond between language and society" (Shih 1986: 98). They function to "open communicative acts and set the tone for the interchanges; they establish at least initially, the relative power and distance of speaker and hearer" (Wood & Kroger 1991: 145). They also play an important role in expressing "social attitudes, social structure, and group ideology" in a dyadic exchange (Lee-Wong 1994: 299).

2.2.1.2 Previous studies of address forms The first major work to emerge from the literature is the pioneering study of Brown and Gilman (1960), which focuses on pragmatic aspects of the secondperson pronoun TN ('tu '/'vous '). This study reveals that pronoun usage is governed by two social dimensions: power and solidarity. In particular, the power semantics usually result in the non-reciprocal use of the TN pronouns while the solidarity semantics would lead to reciprocity in language usage. For example, with an increase in subordination there is an increase in the tendency for a subordinate to use the 'V' pronoun in addressing a supenor. Conversely, with an increase in superordination there is an increase in the use of the 'T' pronoun by the superiors. On the other hand, the distance between the speakers also determines the speaker's choice of pronoun, and of all terms of address. The more close or familiar the speakers are, the greater the tendency to use the 'T' reciprocally. As the social distance increases between the speakers, there is an increase in the use of 'V' pronoun.

35 One significant finding of Brown and Gilman's study (1960) is the changing pattern of usage between 'T' and 'V' address forms in German, French, Spanish and Italian, that is, the usage has become liberalized so that solidarity gains weight over power in those languages. Another interesting finding is the correlation of the social structure or ideology and address forms, namely, there are certain styles of address which are typical of social groups or of individual speakers. Brown (1965) also notes that particular weight attached to social characteristics of speakers and hearers and interpersonal relationships between them may vary between different languages and cultures. For example, familiarity and gender differences are given greater weighting in the address systems of India and Japan in comparison with European systems.

Following Brown and Gilman's work, Brown and Ford (1961) investigate American English address system in order to look into what guides the choices that speakers make in addressing different people. Based on a varied collection of data, including modem American plays, interactions in a drafting film, and observation of 56 children, Brown and Ford find that first name (FN) and title plus last name (TLN) are the most common address forms in American English. They also find that FN is reciprocated while TLN is used only at the beginning of acquaintance. Thus, intimacy and distance are two crucial factors determining the selection in symmetrical relationships. On the other hand, non-reciprocity of FN and TLN is caused by differences in age or professional status. One interesting finding shown from Brown and Ford's work is that there is not necessarily a correlation between age and higher occupational status, that is, when a younger person has a higher rank than an older addressee, this younger superior will be addressed by TLN and gives back FN to the older inferior due to the precedence of occupational status over age. In addition to these two major addressing patterns, Brown and Ford also propose other variants: title without name, last name alone, and multiple names, each of which symbolizes a certain level of intimacy between the speaker and hearer.

Ever since Brown and Gilman published their influential paper, a great many works have been devoted to forms of address in the world for the past 30 years like English (e.g. Ervin-Tripp 1972; Wolfson & Manes 1980b), French (Lambert & Tucker 1976), Spanish (Moles 1974), Hungarian (Ostor 1982), Italian (Bates 1975), Slovene (Kess &

36 Juricic 1978), Russian (Friedrich 1966), Yoruba (Oyetade 1995) and Zulu (Suzman 1994). Furthermore, a number of multilingual studies have been carried out to compare two or three languages in the use of pronouns (Braun 1988; Martiny 1996; Kroger & Wood 1992). Generally speaking, these studies are mainly embedded within a description of cultural-specific systems of address forms, or of cross-cultural features of politeness that appear in address forms in both spoken language and written discourse.

One recent example of this paradigm is Nevalainen and Raumolin-Brunberg's study of the pragmatics of address forms in early English correspondence (1995). Under Brown and Levinson's politeness framework (1987), the authors suggest that there is an increase in the tendency for writers to use positive politeness strategy in address form in letters. The choice of a term with a positive orientation (e.g. nickname, dear+ title) may serve as a "social accelerator", signalling that the writer wants to "come closer" to the recipient (Brown & Levinson 1987: 103). Nevalainen and Raumolin-Brunberg (1995: 588) also point out that the address forms with negative orientation of giving deference are simplified not only structurally but also socially. For example, the use of 'sir' may neutralise power distinction among people with hierarchical ranks. A particularly interesting finding of this study is that the results show that power is regularly reflected in address forms not only between people of unequal status, but also between equals. The writers with higher status usually choose a positive politeness strategy in address forms, such as the use of an in-group term which is not used by the recipients with lower statues, thus conveying asymmetries. In contrast, the reciprocal use of positive terms assumes "equal social status, close distance (D) and, presumably, mutual affection" (Nevalainen and Raumolin-Brunberg 1995: 588). Based on these, the authors suggest that in expressing politeness, the forms of address play a critical role.

Another recent investigation is reported in Oyetade's study of the entire system of address forms in Yoruba, a Defoid language of the Niger-Congo phylum (1995). With data from shortwave radio, TV plays, and unobtrusive observation of actual usage, the author assumes that address forms in Y oruba share a certain level of universality with those in other languages like French, English, German, Italian, Spanish, Nuer, and

37 Hindi. Nevertheless there exist unique features that reflect the Yoruba socio-cultural milieu. For example, in the Yoruba addressing system, age supersedes rank in dyadic relationships, and the dichotomy of power and solidarity proposed by Brown and Gilman (1960) becomes unclear with respect to Yoruba kinship terms of address. Thus solidarity does not necessarily imply equality among the Y oruba. As far as the relationship of the address usage and contextual factors is concerned, the choice of the form to address another Y oruba person depends on the relationship between interlocutors, on the norms of behaviour appropriate to the situation, and on "what the speaker wants to emphasize in the relationship with the addressee" (Oyetate 1995: 533).

As with the countries discussed above, there have been a number of studies of addressing phenomena in China during the past 30 years (Blum1997; Chao 1956, 1976; Chen 1991; Hong 1985; Li 1997; Shih 1986; Tian 1996; Wu 1990; Xu & Zhang 1987; Zhu 1997). Hong (1985) examines Chinese addressing phenomena in face-to-face verbal interaction from a politeness point of view. He states that since ancient times the Chinese have had politeness rules for addressing practices: Addressing others requires the use of an honorific term, while referring to oneself, the speaker needs to use a humble term. Every use of address form can be seen as confirmation of the addressee's status. Hong (1985: 206) also argues that choosing which address form to use is governed by human relationships dictated by the Confucian principle of reciprocity: fu-zi (father-son), jun-chen (emperor-officials), fu-fu (husband-wife), zhang-you (older-younger), xiong-di (older brother-younger brother). Hong's reciprocity means that address forms are linked to the roles people assume when they are being addressed in a particular way. Hong also notes that one typical characteristic of the usage of address terms in Chinese society is the extension of kinship terms to non-kin persons. He interprets this finding by the fact that extending the family relationships to outsiders thus fulfils the goal of Tianxia yijia (One family under the sky).

A detailed investigation of addressing is reported by Shih (1986) in her empirical research of address forms and functions. She compares the addressing behaviours conducted by Chinese in Taiwan with those by Americans in both Taiwan and the United States based on a various sources of data (i.e. questionnaires, interview and the

38 author's observations). A close examination of the data shows that Chinese and Americans do share some similarities in the classification of address forms and the linguistic patterns of addressing. However, there are obvious differences between these two cultures. The first striking difference is that in Chinese society a student never addresses his or her teacher by given name. The second one is that in American society people of the same generation make no distinction in seniority and they use reciprocal address forms. The third difference is that Chinese use kinship terms to address non-kins, while most Americans, in contrast, keep the kinship terms only for those with real kinship relations. Her interpretation of such similarities and differences is that rules of address are universally determined by the social variables of relative power, degree of solidarity and the formality of a particular social context. But the degree of impact of these social variables differs from one society to another. As stated by Shih (1986: 100), in American society intimacy takes precedence over deference and formality, while in the Chinese context deference and formality carry greater weight. A Chinese person may show intimacy to a senior or superior, but he or she must express deference by some devices along with conveying intimacy.

Chen (1991) studies greeting expressiOns m China and comes up with a four-fold taxonomy, in which address forms include the use of names, kinship terms and titles and they are grouped under the category of appellatives. Chen (1991: 55) notes that in general these address forms are used to express different attitudes and styles which are suitable for various addressees and situations. Chen also argues that these terms, as with other greeting devices, constitute an important part of the polite languages. For instance, by using a title preceded by a surname, the addresser not only shows respect to the addressee, but also implies a degree of distance. In the study of naming practices and the power of words in China, Blum (1997) examines the use of kinship terms, pronouns and proper names in Chinese face-to-face interaction. Her findings again confirm that the use of kinship terms is quite common in Chinese society. Blum explains that kinship terms are ultimate expressions of solidarity. As far as the influence of social factors is concerned, Blum (1997: 371) suggests that relative status is an important factor in determining which term will be used, but considerations of affection sometimes override those of power. This finding is socially interesting since it implies the importance of finding a rich set of categories

39 of social variables as determinants of addressing practices. Blum also observes that naming practices in China can be seen as a reflection of "giving, withholding and exchanging face" in terms of solidarity as well as authority (Blum 1997: 372). This study again confirms a close link between address forms and the use of politeness in addressing practices.

Investigating the influence of the social environment in the use and change over time of address forms, Li (1997) examines a newly arising phenomenon of address, in which western-style names are used by Chinese Hongkongers to communicate with westerners or among themselves. Li claims that the adoption of a western name to address the other person is mainly motivated by an intention to achieve the "involvement strategy," and to communicate with this person on "a more egalitarian basis" (Scollon & Scollon 1995: 512). The author assumes that using a western address form may enable the speaker to quickly develop friendship and intimacy, which is important in "getting acquainted, both in inter- and intra-cultural encounters" (Li 1997: 489).

Another recent examination is presented in Zhu's (1997) study of address forms occurring in Chinese personal letters in relation to politeness phenomena. Based on the authentic data, the researcher finds that the address forms employed in letters differ across different social relationship: 100% of family members address the other with the marked form (full name is taken as an unmarked form), while 86% of non-family members use such a type of addressing. The researcher also finds that among the marked forms, family members preferred to use the intimacy-oriented form (e.g. endearing term plus name), while non-family members favour the use of a more polite one (e.g. respectful term plus name). Using Gu's (1990) model of Chinese politeness, Zhu explains that intimacy takes precedence over social rank in the address forms of family letters, but familiarity rather than social position is emphasised more among friends and lovers (1997: 19). Discovering the different orientations reflected in the address forms used by family and non-family people, Zhu also finds that the addressing structures are more complex in non-family letters than in family letters. Zhu points out that the choice of address forms is determined by the degree of intimacy, familiarity and social rank. The degree of influence of each of these factors varies with different contexts. Zhu attributes all these results to Chinese people's manifestation of

40 politeness under the category of 'the Address Maxim' proposed by Gu (1990) in his conceptualization of politeness in modem China. These observations provide further evidence of the effect of solidarity and status upon address practices in particular situations, and highlight a reason to investigate the address form in relation to politeness (Blum 1997; Kroger &Wood 1992).

In 1999, Zhu explored the current trends in Chinese business communication in a study which focused on examining linguistic forms of salutations and other formulaic moves, such as headings, greetings and expressing good wishes. The most important contribution of this study is that it provides interesting information about the dynamic nature of sales genres, the relationship between the use of genres and communicative purposes, and the relationship between communicative purposes and social contexts. Zhu (1999: 187) finds that the differences in communicative purpose result in differences at moves and steps. For example, sales invitation and sales 'tongzhi' (circulars) require a more formal style (i.e. using a heading) than sales letters due to the purpose of inviting and giving an implicit order. Sales letters, which primarily aim to establish personal relationships with the reader, prefer to use greetings like 'ninhao' (how are you).

Another significant contribution of this research is the author's attempt to investigate sales genres in relation to politeness strategies, and the effect of social factors, such as social rank, on these strategies in business communication. For example, Zhu (1999: 188) finds that honorific forms like 'gui' (you H) are preferred in salutation in both sales letters and sales invitations due to their 'etiquette' nature. According to Zhu, sales letter writers tend to use such a 'shangxing' form, indicating lower social rank so as to show respect to the reader. Zhu attributes this phenomenon to the specific ways of showing politeness in Chinese sales genres. This finding confirms Blum-Kulka and House's (1989) claim that there are different ways to indicate politeness in different languages. Also, Zhu refers this to Chinese cultural values, that is, each person's linguistic behaviour should be in relation to other people, and the person of lower social rank should show respect to the person of higher rank.

41 In summary, these previous studies conclude that forms of address are seen as "ideally expressing the speaker's innermost thoughts", and as "exemplifying the speaker's knowledge of what is expected on certain occasions and in certain contexts" (Blum 1997: 372). Forms of address are usually a source of social information about the identities of speakers, gender, age, status, occupation and ethnic identity (Li 1994). Thus, these studies constitute impressive support for the exploration of rules of address forms, and represent a certain degree of confirmation with empirical cross-cultural data, which seems to be unmatched for any other sociolinguistic phenomenon. However, the literature review has also demonstrated that most of the research that has been reported on address forms has focused on oral interaction, whilst the study of addressing in written communication (e.g. personal letters), has received very little attentions (see Nevalainen & Raurnolin-Brunberg 1995; Zhu 1997). In addition, there is inadequate knowledge about the extent to which the factors of age, gender and their interactions with other factors affect the use of address forms in this kind of interaction. Recognizing this, it is possible to suggest that further studies of address forms and examining the influence of possible social factors to a more comprehensive extent within written discourse are required.

2.2.2 Openings The study of openings in human interaction has attracted researchers' attention since 1960s. The majority of investigations into this aspect has focused on oral communication (e.g. Goffrnan 1967, 1971; Goody 1972; Hopper 1992; Hutchby 1999; Irvine 1974; Kasper 1989; Kendon & Ferber 1973; Omar 1992; Pavlidou 2000; Schegloff 1968, 1986; Schiffrin 1977). In all the major studies of openings, there appears to be a general agreement that in human communication people must have ways to show that "communication is about to begin and then begins" (Hatch 1992: 8). Another common view from these studies is that the ways in which openings are accomplished vary according to the channel (e.g. telephone talk, face-to-face interaction, letter writing), the mode and contexts.

2.2.2.1 Concept of conversational openings Schegloff refers to conversational openings (in this case, openings in a telephone talk) as "a place where the type of conversation being opened can be preferred, displayed,

42 accepted, rejected, modified - in short, incipiently constituted by the parties to it" (1968: 25). According to Omar (1992: 15), conversational openings can be seen as "a section of talk between two people at the beginning of an interaction". It starts when the two participants· verbally or nonverbally recognise each other and ends when the main topic of the conversation is introduced in the case of intended encounters (e.g. telephone conversations, and planned face to face encounters), or to the time when the participants are ready to take leave of each other in the case of chance encounters such as those in the streets." In brief, openings mark a transition between a period of absence and a period of increased access (Goffman 1971). As Goffman explains:

The expectation that an access ritual will be performed by a certain person at a certain moment establishes a time-person slot such that anything issuing from him at that moment can be very closely and imaginatively read for a functional equivalent of an access ritual. ( 1971: 80)

According to Schiffrin (1977), this access ritual is socially organised despite its · cognition in nature. It is "symbolically conveying a social identity and membership in a relationship" (Schiffrin 1977: 681 ). It enables participants to "present to one another in a special way" (Callan 1972), and put them in a social context as specific social entities (Firth 1972). For this reason, openings are regarded as ritual, rather than technical (Schiffrin 1977), and they imply the speaker's willingness to establish a social relationship with the hearer (Firth 1972).

2.2.2.2 Types of conversational openings There are two types of conversational openings, which have been discussed in the literature: a brief conversational opening and an elaborate conversational opening. The brief opening refers to "a far distance greeting" by Krivonos and Knapp (1975:119), "distance salutation" by Kendon and Ferber (1973: 598), or "passing greeting" by Schiffrin (1977: 688) in English conversational openings. It occurs when there is a considerable physical distance between two participants, and it does not accomplish all stages of an opening as seen in example (1):

43 Example (1). A brief conversational opening (based on Omar 1992: 17) A:Hi,Max!

Following Schiffrin (1977: 688), the brief opening may suggest a different modification of the definition of encounters, since the passing greetings "constitute the only focus of their participants' attention, but others occur during a more ongoing and dominant involvement." Schiffrin also claims that although the brief opening like the passing greeting terminates the minimal access few seconds only after their initiation, they still "leave open possibilities for future contact" (1977: 680).

The elaborate openings refer to a "close distance greeting" by Krivonos and Knapp (1975: 121), a "close salutation" by Kendon and Ferber (1973: 599), or "opening sequences" by Schiffrin (1977: 680). As noted by Schiffrin, opening sequence are ritual interchanges between two acquainted participants, which serve to accompany "an initial increase in their potential for mutual accessibility" (1977: 681). They can be read as "a formal depiction of either the structure of opening sequence or production rules for participants representing possibilities for action within normative limits" (Schiffrin 1977: 681). In opening sequences, the interaction occurs within "a distinct spatial and orientational frame" (Omar 1992: 17). This presents the methodical procedures by which the participants display and achieve identification and recognition of each other (Schegloff 1979: 24). Because of its relevance to the Chinese written openings for this study, the elaborate openings will be discussed in details in section 2.2.2.4.

2.2.2.3 Functions of conversational openings According to Omar (1992: 8), conversational openmgs are used to "establish relationships between participants as well as reveal information about the state of the relationships." Such a relationship-oriented function can be best understood in light of meanings of greetings comprising a crucial part of the full set of components of conversational openings. Schiffrin (1977: 688) argues that greetings in openings are seen as a way of establishing contact between participants who are about to become involved in a longer interaction (elaborate), or simply of maintaining contact between those who have no intention of undertaking further interaction. Coupland and

44 Coupland (1992: 217) treat greetings 'How are you' as "a 'genuine' effort to elicit an appraisal of health or well-being," which serves to "achieve some minimal threshold of relational engagement .... " Berger and Bradac (1982) hold a similar opinion that greetings are often not used to produce self-revelation, "but rather to signal acknowledgement of the other." Based on these views, greetings tum out to be utterances constituting a central example of phatic communion, because the main purpose of the participants using this move is not to exchange cognitive information about the world, but simply to show that the interactants have recognised each other and that they still share membership in a social group (Malinowski 1972: 151).

Conversational openings are also used to reflect social characteristics of participants such as age, rank, and interpersonal relationships between them (e.g. intimacy, distance). Kendon and Ferber summarize these functions of openings in conversation:

Greetings have an important function in the management of relations between people ... In the way in which the greeting ritual is performed, the greeters signal to each other their respective social status, their degree of familiarity, their degree ofliking one another... (1973: 592)

The reflection of social roles in conversational openings has been discussed in detail in studies of the Wolof (Irvine 1974), the Gonja (Goody 1972), the Kiswahili (Omar 1992), and the Baatombu (Schottman 1995). In Gonja openings, for example, status is established in the way they are initiated, particularly through what kinds of poses are adopted and what terms of address are used. In starting communication, lower status persons often initiate the opening while higher status persons often respond. Moreover, the lower status persons request a favour from the higher status persons while the higher status persons give instructions to the lower status persons. In Wolof openings, the participants clearly take two dissimilar roles: the initiator/questioner and the respondent. Usually, two persons who are engaged in an encounter place themselves in an unequal ranking. They come to an agreement about which party takes the high status role and which one takes the lower status role in the process of openmgs.

45 Omar' s study of openings in Kiswahili society reflects the influence of participants' age in linguistic realizations of openings (1992). The results of her study reveal that the younger person initiates the opening by using a deferential greeting, while the older person responds to the respectful form and then dominates the opening by adopting the roles of questioner. But same age participants, on the other hand, join a conversation with a highly involved manner: Each participant competes to take the role of asking questions; any of the participants can initiate the opening; and none of them uses the respectful greeting.

Schottman (1995) studied the elaborate greeting rituals of the Baatombu, an African tribe. He finds that the openings in Baatombu are composed of a series of greetings uttered by superiors and a short and standardized response from inferiors. Unlike that in Wolof and Kiswahili openings, superiors in Baatombu can use any of the initial greetings followed by an inferior's birth rank name which identifies the inferior as a member of a given generation and social category. They also ask about the inferior's state of health after greetings. By prolonging the greeting exchange, the superiors may show their sincere and respectful concern for the inferiors. However, the inferiors have limited choices of responses. Basically their replies used in the opening procedure are merely an acknowledgment that the greeting has been received and not the answer to a question, because in the Baatombu tradition the inferiors must "take care not to appear pretentious or obtrusive" (Schottman 1995: 506). When two equals are involved in greetings, the exchange is much shorter and their structures are less complex, but much more symmetrical. Relations between equals do not play the same essential role in the functioning of Baatombu openings as do relations between participants of unequal status. Schottman's study reveals that the social structure and cultural values embedded in the structure greatly influence participants' use of greetings in openings. In other words, the hierarchical role relationships that define each member's rights and obligations are greatly reinforced through the greeting routine. Schottman (1995) also states that the existence of the hierarchy in the Baatombu greeting routine is a typical characteristic of a rather large linguistic community which has a social structure of a widespread interdependence, as hypothesized by Goody (1972).

46 Conversational opemngs can also be used to spell out interpersonal relationships between participants (Hopper 1992; Hopper & Chen 1996; Lindstrom 1994; Schegloff 1979, 1986). Schegloff (1979) finds that in English, for example, participants display intimacy by using as few resources as possible for mutual recognition. Put differently, intimacy constitutes a continuing conversation between participants, and participants convey a state of intimacy between themselves by a quick telephone opening, especially by greetings:

Example (2). UTCL D10.3 (intimate dating partners) (based on Hopper and Chen.1996: 296) A: Hello? C:Hey A:Hi C: Are you ready yet?

This example shows that openings between intimates are conducted in a reduced form: The caller includes identification into greetings, then uses an inquiry 'Are you ready yet?' The "free-standing greeting pair" here conveys an intimate relationship state (Hopper & Chen 1996: 296). In fact, openings between strangers can also be achieved through distinctive reductions (Whalen & Zimmerman 1987). But the difference is that in strangers' telephone openings, the caller often says 'Yes' and then states a reason for the call. Such openings emphasize the self-introduction and the efficient movement toward the transacting business, rather than intimacy.

Houtkoop-Steenstra (1991) finds that Dutch intimates at times use voice-sample identification, although this phenomenon is unusual in the Netherlands. Lindstrom (1994) observes that Halla (Hello), which is not frequent in Sweden, can occur between two intimate participants and Swedish family members. By using this reduced form of openings, similarly, the participants intend to convey a level of intimacy.

Hopper and Chen (1996) specify linguistic resources within telephone openings m Taiwan, which show a greater difference from European languages. According to the authors, telephone partners in Taiwan deploy a variety of greeting phrases to show the state of their interpersonal relationship. For example, between family members the

47 callers may indicate intimacy by a recognitional naming of the answer according to the voice sample. The answer may show intimacy by identifying the caller only by using the given name (for a non-family member) or by showing certainty of identification without using a name (for family members). Using a name to identify the answerer between family members, according to Hopper and Chen (1996: 306), can be seen as a way of avoiding "embarrassing consequences of claiming excessive familiarities." In strangers' telephone openings, in contrast, the caller often uses 'Ninhao' (How are you?). This greeting token serves to show politeness to a stranger and also to assume the social distance. Hopper and Chen (1996: 306) attribute such performances of openings between intimates and strangers to the influence of the Confucian virtue of Li (propriety), which requires members of a society to adapt their talk to a speaker's age, status and intimacy of connection.

All these examples show that linguistic choices and practices within conversational openings, as with addressing practices, can reflect social characteristics of the speaker and the hearer (e.g. age, status), and the state of interpersonal relationships between each other. However, conversational openings are culturally determined. Every speech community conventionalizes a certain format for displaying seniority or juniority and intimacy or distance. Using Hopper and Chen's words, the indication of those features from an opening "get marked with distinctive patterns of language, culture, circumstance, and relationship" (1996: 31 0).

2.2.2.4 Structures of conversational openings Schegloff (1968, 1986) proposes a model of sequence in telephone openings in the United States, which is made of four distinct and staged sequences:

Example (3) Structure of telephone openings (Based on Schegloff 1986: 115)

0. ring Summons/answer sequence A: Hello:: C: H'llo, Clara? Identification/recognition sequences A:Yeh C: Hi. Bernie. Greeting sequence A: Hi Bernie. C: How're you. How are you sequences A: I'm awright,

48 How're you C: Okay::? A: Good. C: Laura there (switchboard request)

This example shows that the summons is the ringing of the telephone at someone's home, which is acknowledged by the answer 'Hello' (Schegloff 1986). According to Schegloff, the summons-answer sequence is "overtly addressed to opening, and confirming the openness of, a channel of communication, and the availability of an attentive ear and a mouth ready to speak" (1986: 117). Besides, the answer also performs the identification work by saying 'Hello', which is overtly addressed to the identification sequence. "If caller recognizes answerer from the voice sample in the answering turn, then the caller should show (or claim) such recognition in next turn, the second turn in the call" (Schegloff 1986: 126). Thus, it appears that telephone partners identify each other from minimal resources, a principle labelled by Schegloff (1979: 50) as "oversuppose and undertell."

The third part is the greeting sequence. Greetings are often the first utterance in the encounter. Normally, the first greeting token makes relevant an immediate return greeting which occurs in the next turn. Such a greeting reciprocity demonstrates a state of mutual participation (Goffman 1963; Kendon & Ferber 1973; Schiffrin 1977) and indicates prior acquaintance (Hopper & Chen 1996). After greetings, the model provides for a slot for brief inquiries about the addressee's state of health. These inquiries are essentially conventional in form and often have positive answers (Sacks 1992). Moreover, the 'How are you?' sequence may serve as greeting substitutes. When inquiries are treated as greeting alternatives, they may receive a response which is not an answer to the inquiry about health (Sacks 1975).

Although Schegloffs work has some limitations (e.g. the core sequences do not cover possibilities that are automatically used by participants, as noted by Lindstrom 1994), his analysis of American telephone conversational openings has provided a powerful basis for studies of opening performance in other situations as well as telephone openings in both English and other language communities. Specifically, scholars have started to examine how openings are sequentially organized in face-to-face interaction

49 (Goody 1972; Irvine 1973; Kasper 1989; Omar 1992), which is distinct from comparisons of telephone openings in different languages (Godard 1977; Hopper & Chen 1996; Hopper & Doany 1989; Houtkoop-Steenstra 1991; Lindstrom 1994; Sifianou 1989). These studies can be used to show similarities and differences in terms of conversational openings between languages and cultures.

Following the study of Schegloff (1968), Kasper (1989) analyses the cross section of conversational openings in English, moving from the initial dyad to the turn when a topic is introduced. Her analysis shows that the English opening in a face-to-face interaction (as in someone's office) is initiated by a greeting (Hello/Hello; Good morning/Good morning) and followed by the other moves: phatic inquiry/phatic response (How are you/Fine Thanks); phatic remarks (Haven't seen you for a while); and topic introduction. This possible English opening can be illustrated in the following example:

Example (4). Structure of English conversational openings (based on Kasper 1989) A: Hello. B: Hello. A: How are you? B: Fine, thanks. I haven't seen you in a while.

Kasper argues that between the initial opening and the tum when the new topic is introduced, English people might have what Schegloff and Sacks (1973) refer to as a "pre-first topic closing." The initiator of a conversation may ask questions like 'Are you busy?' or 'Did I wake you' to see whether the addressee is available to talk. If the initiator gets an affirmative answer, the talk will be closed without introducing a new topic. But if the answer is negative, the initiator is warranted in continuing with conversation.

The sequential organization of openings in face-to-face interaction is explored in depth in Irvine's analysis ofWolofopenings (1974). Based on field observation, Irvine finds that the W olof greeting procedure consists of two stages: Sal = salutation, QP= questions and praising God. Irvine (1974: 170) claims that conventionally people in the Wolof should not greet each other just with the Arabic 'salaam alikum' (Peace be with you, or peace be upon you), since the single greeting would be interpreted as

50 brusque and rude. They should continue their talk by asking 'Do you have peace?' and other questions. The questions asked in a greeting are stereotypes and are often followed by stereotyped responses. The following example illustrates the structure of Wolof conversational openings:

Example (5). Structure ofWolof conversational openings (based on Irvine 1974: 171)

A: Salaam alikurn (And with you peace) ..... Salutation B: Malikum salaam (And with you peace) A: A's name: ..... Naming B: B's name A: B's name: ..... Naming B: A's name A: Na ngga def? (How do you do?): ..... Question/Response B: Maanggi fi rek. (I am here only) A: Ana waa kir gi? (Where /how are ..... Question/Response the people of the household?) B: Nyu-ngga fa. (They are there) A: H'mdillay. (Thanks be to God) ..... Praising God B: Tubarkalla. (Blessed be God)

As stated by Irvine (1974: 173), unlike the American greetings in which the initial statement 'Hello' is reciprocally used by both parties, the Wolof greeting structures are clearly divided into two dissimilar parts: the Initiator-Questioner and the Respondent, and the roles correspond to high and low status. The high status person does not greet a person of low status but the low status person must perform this act to the high status person. This holds true of older and younger brothers, and of men and women. Furthermore, a person who holds low position takes a role of questioner while the high status person remains passive, only responding the phatic inquiry done by the low status person. Irvine refers to these as 'self-lowering and self-elevating,' which she attributes to cultural stereotypes of"noble and griot behaviour" (1974: 173).

One recent study is Omar's examination of openings in Kiswahili society (1992). She argues that Kiswahili openings are elaborate. They consist of recursive patterns which are formulaic and routinely used by participants. According to her analysis, openings in Kiswahili are initiated by a unique opening hodi (knock) and its response karibu (knock), followed by an expression of joy or calling of someone's name, a

51 general salutation, a respectful greeting sequence, and recursive sequence of phatic inquiries and phatic responses. The unique opening hodi and karibu occur only when one visits another person in the house. The following example shows the structure of conversational openings in Kiswahili face-to-face interaction:

Example (6). Structure ofKiswahili conversational openings (based on Omar 1992)

A: Hodi (Knock). Unique opening B: Karibu (Welcome). Reply to unique opening A: Mosi (Name of a person). Calling someone's name B: Iii i ! Expression of joy A: Assalam Aleykum (And with you peace). General salutation B: Wa aleykum salaam (And with you peace). Reply to general salutation A: Shikamoo (Originally I hold your legs). Respectful greeting form B: Maslkheri or Sbalkheri (Good evening or Reply to respectful greeting good morning). form

A: Marahaba. B: Msalkheri. A: Hujambo? (How are you?) ~ Phatic inquiry B: Sijambo (I'm fine). ~ Reply to phatic inquiry A: Watoto hawajambo? (Are the children fine?). ~ Phatic inquiry about the other B: Hawajambo (They are fine). ~ Reply to phatic inquiry

Omar does not explore any assumption that underpins this sequential organization of opemngs. Rather she discusses in detail the influence of age as a major sociolinguistic parameter in people's performance of openings in Kiswahili. She argues that as in the Wolof openings (Irvine 1974), the younger person in Kiswahili adopts a self-lowering role by producing a respectful greeting to the older person. But in contrast to that of the Wolof openings, it is the older person who takes the role of questioner and produces the phatic inquiry to the younger person in Kiswahili. The reasons for this structural similarity and difference cross language groups (Wolof vs. Kiswahili), according to Omar (1992: 85), are linked to the younger person's perception of the need to avoid conflict and face-threats by following sociolinguistic norms.

A research overview of opemng structures suggests that openmgs m oral communication is an example of accomplishing a specific communicative purpose by doing "a restricted range of tasks" (Lindstrom 1994: 232), which is normally achieved

52 through a set of interactional sequence by the use of adjacency pairs. Despite the mechanical and social organizational universals, many differences still remain in the opening structures cross-culturally. These differences seem most reflected in the constitutions of components of openings, in the ways in which identification and recognition are organized (e.g. 'Hello' rather than other objects is used in the speaker's first tum), and in the orientation towards seniority or juniority and intimacy or distance between language communities.

To sum up, openings are an important aspect of behaviour in spoken interaction, in which participants attempt to "establish a channel for communication" [and to] "ascertain the identity of their interlocutor" (So' o & Liddicoat 2000: 105) through ritualized or formulaic sequential language. In addition, the utterance that occur in either an actual telephone opening or an oral conversation "gets marked with distinctive patterns of language, culture, circumstance and relationship" (Hopper & Chen 1996: 31 0). Such findings are useful for examining properties and structures of openings in written communication. That is because there are similar features in the opening process between these two types of interaction, although oral and written discourse differ in the ways in which openings are accomplished and patterned.

Perhaps the most significant similarity is that, following Bhatia's approach to genre analysis, openings in both oral and written communication serve as a "recognizable communicative event," which is identified and mutually understood by members of a certain community in which they occur (1994: 13). Although there are a number of differences in terms of content, form and ways of operating between these two types of interactions, they are "primarily characterized by the communicative purpose(s)" (Bhatia 1994: 13). In other words, participants in both kinds of openings are characteristically engaged in the establishment and maintenance of the relationships with the other party as they move through the opening to the business (i.e. referential exchange) of the encounter.

Another important similarity is that openings m both oral (especially telephone openings) and written discourse can be seen as "a highly structured and conventionalized communicative event" (Bhatia 1994: 14). They involve a restricted

53 range of tasks, normally achieved through a fairly narrow set of sequences. Although in the sequences the constitutive components and structural ways vary between conversational openings and written ones, the underlying sequential framework appear to be similar to some extent. For instance, in the opening process the speaker or writer is inherently engaged in patterning some routines (e.g. greeting then phatic inquiries), which are regularly used according to the rules and conventions for communication. These similarities in the engagement of conventional or routinized communicative behaviour and, consequently, in the choice of certain routine expressions by both speaker and writer may shape a similar underlying internal structure for these two types of openings.

Taken together, it can be assumed that openings in oral and written communication appear to comprise a shared feature: Conventional and routinized sequences are employed to identify the status of participants and to establish the relationship between them. From this point of view, conversational openings may thus represent a basic relational and structural environment for exploring the manipulation of openings in written communication.

2.2.3 Closings Apart from having ways to start communication, people must also have ways to end the communication according to the channels (e.g. telephone talk, face-to-face interaction, letter writing), modes and contexts (Hatch 1992). Recognizing the essential status of closings in people's communication, a number of papers have described the processes, properties and structural features of closings across different languages and cultures in the past thirty years (Albert & Kessler 1978; Clark & French 1981; Laver 1981; Okolo 1995; Omar 1992; Pavlidou 1997, 2000; Placencia 1997; Rostila 1995; Schegloff & Sacks 1973). As with the studies of openings, the extensive literature of closings is centred on investigating this aspect in spoken language. Among the key concerns of these studies is the question of how people close the conversation. What series of moves people make at the end of interaction. How is this speech behaviour related to the management of relationships between participants?

54 2.2.3.1 Concept of conversational closings According to Albert and Kessler (1976: 541), the closing of a conversation refers to "the cessation of all verbal and nonverbal communication between two previously interacting individuals." Schegloff and Sacks (1973) explore this aspect in telephone talk and claim that when closing a conversation, both interlocutors understand the completion of talk and work on finishing the conversation. In brief, conversational closings serve to mark "the end of access and the beginning of a period of absence between participants" (Omar 1992: 32).

2.2.3.2 Functions of conversational closings From an organizational point of view, conversational closings organize the way in which the speakers terminate a conversation. Participants in a conversation come to "a simultaneous agreement that completion of one utterance does not require another and that the silence that follows is not merely a transition between two turns at talk" (Omar 1992: 33).

In addition to the function of organizing the termination of a conversation, conversational closings also perform interactional functions. For example, they are used to confirm and reinforce the relationship between the participants. Several studies (Firth 1972; Goffman1971, 1976; Hopper 1997, 2000; Laver 1975, 1981; Luger 1983; Placencia 1997) have examined the interactional (as opposed to structural) aspects of closing utterances in both face-to-face and telephone interaction. According to these scholars, most expressions that occur in the closing section are not used merely as "exit mechanisms," rather, they perform ritual functions, through which participants "ratify their acquaintance and appreciation of each other" (Placencia 1997: 56). We can better understand what ending a conversation means by considering farewell message. Goffinan refers to the farewell message as "a state of decreased access" (1971: 79). Goffinan also considers variations of the type of rituals used in closings with respect to the acquaintance of the participants, the frequency of the contact the participants have and other social factors, that is, if the leave-taker is simply returning "into the same felt probability of contact from which they came, and if this probability is high, then a phrase such as 'See you' or 'So long' may well be employed" (Goffinan 1971: 82). Furthermore, a relation who has not been seen for a long time will receive a

55 show of more emotion than a neighbour, for example, who has not been seen in a long time. "The more lengthy and absolute the predicted separation, the more expressive the rituals" (Goffrnan 1971: 88).

Like Goffrnan, Firth (1972) highlights the socially phatic functions that closings accomplish. According to Firth (1972: 1), "greeting is the recognition of an encounter as socially acceptable" [while parting] is "the recognition that the encounter has been acceptable."

Laver (1975, 1981) and Luger (1983) view closings as phatic commumon also. According to Laver, closing expressions serve two functions. First, they are used to mitigate "the potential sense of rejection that a participant might feel when his fellow participant initiates the closing phrase" (Laver 1975: 57), which is helpful to reach a consensus regarding the termination of the encounter. Second, they are employed to consolidate the relationship between participants. As Laver claims, "the type of linguistic token chosen by a speaker may reflect his view of the social structuring of the interaction" (1975: 222). According to Luger, phatic formulae "provide the necessary information for defining the situation and signalizing the level of the relationship" (1983: 701).

In addition to confirming and enhancing relationships between the participants, conversational closings are also used to reflect the characteristics of participants and the interpersonal relationship between them. While examining telephone 'goodbye' sequences in American English, Clark and French (1981) suggest that the production of farewells is dependent on the degree of acquaintance between participants. According to Clark and French (1981: 18), the final exchange of 'goodbye' is not simply a means for ending a telephone conversation, rather, it functions "to complete the process of reaffirming acquaintance." The more closely acquainted the participants feel they have become, the more likely they are to reaffirm acquaintance by saying 'goodbye.' Clark and French also claim that the exchange of 'goodbyes' is associated with gender differences: Female callers often use these farewell messages, in order to be more "supportive, polite and expressive" than male callers (1981: 18). According to the

56 authors, the meaning of goodbyes is linked to "socially and culturally defined routines" (Clark & French 1981: 18).

Wolfson (1988, 1989) proposes the bulge theory, which assumes that social distance between participants can affect people's speech behaviour. Wolfson names two extremes on the scale of social distance: intimates and strangers. Interestingly, Wolfson finds that these extremes have similarity in the interactional style due to relative stability in their relationships. In contrast, other relationships such as casual friends and acquaintances become more sensitive in speaking behaviour in order to build more solidarity and better relationships. Wolfson (1988) and other researchers (e.g. Kipers 1983, cited in Wolfson 1988: 34) argue that the bulge theory can be applied to closings in conversations, that is, the sequences of closings among people who have stable and fixed relationships are similar. They further note that it is different for people who have no such relationships such as casual friends and acquaintances.

Recent studies (Pavlidou 1997; Placencia 1997; Rostila 1995) on the closing performance in face-to-face and telephone conversation have highlighted the need to examine the extent to which closings reflect the social role relationships between participants and also the degree of intimacy or distance in particular contexts across different languages. While examining the procedures Ecuadorian Spanish speakers employ to close telephone conversations, Placencia (1997) observes that closings pragmatically serve to express participants' need for appreciation and assurance as to the continuation of their relationship. Placencia also claims that the construction of closings conveys the characteristics of the participants, such as their age and the degree of social distance between them. For example, older participants in Ecuadorian Spanish are likely to produce lengthy closings, which appear to show their orientation toward the formality and deference even with intimates. Also the more elaborate thanks and replies are often used among people who are neither intimate nor distant, which refects the degree of certainty/uncertainty (Wolfson 1988) that participants have about their relationship. Thus, Placencia's findings clearly show that social factors such as age and the degree of intimacy come into play in participants' selection of procedural and linguistic patterns, and thus highlight aspects of conversational closings which are culture-specific.

57 2.2.3.3 Structures of conversational closings Schegloff and Sacks (1973) are regarded as pioneers in the description of closings and their structures through adjacency pair analysis in telephone conversation. According to them, two components remain crucial for any verbal closing: the proper initiation of the closing section and the terminal exchange (Schegloff & Sacks 1973: 318). Thus a proper closing can be demonstrated by the following example:

Example (7). Structure of telephone closings (based on Schegloff & Sacks 1973: 317)

A: O.K. ~ Pre-closing B: O.K. ~ Pre-closing A: Bye Bye. ~ Terminal exchange B: Bye Bye. ~ Terminal exchange

Proper initiation of closing implies that the pre-closing utterances are placed after the completion of a topic and that the pre-closing is accepted by the partner as well. According to Schegloff and Sacks, devices like 'We-ell', 'O.K,' or 'So-oo' normally function as pre-closing phrases, signalling the beginning of closings. They are employed as warrants for the closing to start and proceed if the other participant agrees. This type of devices is called a 'pass' because it operates to indicate that the speaker does not have "anything more or new to say, and also to give a free tum to a next" (1973: 304). By employing the 'pass', the participants understand that they both agree to work to close the conversation.

Schegloff and Sacks (1973: 310-311) also discuss other kinds of pre-closing moves such as 'caller specific', which is related to the interest of the other (e.g. 'Well I'll letchu go. I don't wanna tie up your phone'); 'called specific,' which also appeals to the interest of the called (e.g. 'This is costing you a lot of money'); other pre-closing moves making reference to the materials in the conversation (e.g. 'Well, I'll let you get back to your book'); and other devices which are general in nature (e.g. 'The baby is crying, I gotta go'). Schegloff and Sacks suggest that these kinds of pre-closing moves, as with the devices 'Well' and 'O.K.', represent a warrant for undertaking the closing of a conversation by interrupting a topic and producing a termination of the conversation.

58 The termination exchange is the last exchange in a conversation, and is the part in which "the action of finishing the conversation is actually realized" (Takami 2002: 71 ). According to Schegloff and Sacks (1973), if the pre-closing is properly exchanged, a simple exchange of 'Goodbye' will be followed. Clark and French refer to this part as the 'leaving-taking' subsection and consider the terminal exchange of 'goodbye' or other such expressions to be part of it. In their view, the basic function of the terminal exchange or leaving-taking subsection is the "reaffirmation of acquaintance" (1981: 4). Clark and French (1981: 4) also argue that the leaving-taking subsection is optional and it is only included in a closing when the reaffirmation of acquaintance is required. In addition, Clark and French maintain that the way people close the conversation varies from culture to culture. They find, for example, that "in small close knit societies, in which continuing relations among individuals are taken for granted, people may not need an elaborate form of leaving-taking as in Urban America"(1981: 4). Thus, telephone closings again tum out to be culture-specific: Cultural contexts play a role in the closing process, "along with the type of call, the relationship between telephone partners, etc" (Pavlidou 1997: 147).

Although the pre-closing and the terminal exchange form the essential parts of the closing section, as stated by many researchers (see Schegloff & Sacks 1973), closings in conversations are not a simple matter. Actually they include several moves between pre-closing and terminal exchange. That is because, according to Takami, terminating a conversation can be seen as a FTA and participants therefore have to use some strategies to maintain each other's face and "reassure their relationships" (2002: 72). Schegloff and Sacks (1973: 317) find that in English telephone closings, the moves such as making arrangements are a "reinvocation of certain sorts of materials talked of earlier in the conversation" and belong to this part of closings.

While examining English closings by eliciting data from different status subjects (e.g. acquainted students, non-acquainted students, professor-students) in face-to-face interactions, Knapp, Hart, Friedrich, and Schulman (1973) find different kinds of cues in the closing section. These include reinforcement, which shows agreement to a preceding utterance by a co-participant (e.g. 'yeah', 'ok', 'right', 'sure'); buffing, which is used to change topics (e.g. 'uh', 'er', 'well'); internal legitimizer, which justifies

59 closing by referring to one's own sense that the interaction is completed; (e.g. 'Well, I think that just covers it'); external legitimizer, which justifies closings by making reference to person outside the conversation (e.g. 'I can see that students are waiting to see you. So I'll come later'); welfare concern, which expresses hope for the continued well being of the other party (e.g. 'Take it easy'); continuance statement, which signals that the relationship will continue after the particular encounter (e.g. 'See you later'); terminating phrase, which signals the actual ending of a conversation (e.g. 'Goodbye', 'So long').

Knapp, Hart, Friedrich and Schulman (1973) observe that the choices of the closing cues are dependent on the social characteristics of the participants. For example, when participants are of different status, the participant of lower status tends to use more reinforcement and buffing strategies than the high status participant since these two cues act as a deference mechanism. On the other hand, reinforcement statements occur more often between acquainted pairs than between non-acquainted pairs. Interpreting this result, Knapp et al. suggest that acquainted pairs have a greater motivation for maintaining established friendships than non-acquainted participants for developing new relationships.

A similar investigation of the elaboration of closings is reported in Kasper's study of the procedures of closings conducted by English native speakers and English learners in face-to-face interaction (1983). Comparing the different sequential steps detailed by Schegloff and Sacks (1973), and Knapp et al. (1973) with her data, Kasper (1989: 201) proposes three additional strategies for this behaviour: conclusion statement, which summarizes results from the preceding interaction and ensures common agreements between the participants (e.g. 'Let's do it that way then ok?' 'Ok'.); and appreciation which expresses gratitude for the other participant's concern, or for what the other has done for the speaker (e.g. 'Thanks for looking after the cat') and its response to the thanks (e.g. 'Not at all'). Kasper's closing procedures can be distinguished by the following sequences:

60 Example (8). Structure of conversational closings (based on Kasper 1989)

A. Fine, ok .... Pre-close initiation, B. Right, okay. .... Pre-close acceptance A. Uh, er, well. .... Buffing B. I must be off. You are looking tired. .... Legitimizing A. Let's do it that way then, ok? Ok. .... Conclusion B. All the best; Drive carefully. .... Welfare concern A. Thanks. .... Appreciation B. Not at all. .... Minimizing A. Goodbye; Cheerio; .... Terminating See you later.

Kasper (1989) notes that the occurrence and implementation of these closing structures operate differently between non-educational and educational contexts. She finds that the participants in educational contexts, particularly learners, perform their closings in a reduced form in terms of functions and linguistic implementations, while the participants in a non-educational situation tend to be cooperative and negotiate the closings of an interaction. Kasper (1989: 204) explains that such a functional reduction of closing activities by the language learners is due to their lack of 'pragmatically relevant declarative knowledge' (Faerch & Kasper 1984), which is associated with communicative abilities of interpretation, expression and negotiation. Based on these findings, Kasper (1989: 224) points to the need to develop the foreign language learners' "non-referential aspect of communication, such as the phatic, emotive and relational functions associated with the interactive procedure." While earlier work on the closing structures is often focused on American English in both telephone and face-to-face interaction, some other studies have examined how the closing sequences are managed in other languages and cultures in comparison to those described by Schegloff and Sacks (1973). Omar (1992), for example, examines closings and their relation to age constraints in the Kiswahili community based on the data selected from telephone and face-to-face conversations. Her results indicate that closings in Kiswahili are not as elaborate as openings. One can explicitly initiate the end of contact by the use of a continuation statement, appreciation, phatic inquiry, or the equivalent of 'Ok.' To terminate the conversation, one can use the additional cues of welcome, regards, and farewell message. Based on the evidence, Omar concludes that closings in Kiswahili are not strictly ordered and some components have a dual function of performing both pre-closing and terminating a conversation (1992). What

61 is significance in this study is that some common components which occur in both the opening and closing sections are identified. This provides the evidence that there exists an important parallel between opening and closing behaviours in interactions. As far as the relationship between closings and age constraints is concerned, Omar argues that the age influence, which is an important factor in opening Kiswahili conversation, is not crucial in closings. Thus, openings and closings are governed by different social variables, and social factors (e.g. rank, age) can play a different role in these different behaviours.

Placencia (1997) investigates the closing procedures in Ecuadorian Spanish society by comparing them with those described in Schegloff and Sacks (1973). Her results reveal that the Ecudorian Spanish closings involve some similar procedures to those in American English telephone talk (i.e. the pre-closing statement, terminating exchange). Nevertheless, the results highlight some of the features that appear to be characteristics of Ecuadorian Spanish only. For instance, the pre-closings as warrants for undertaking the closing are often duplicated and restated in Ecuadorian Spanish. And in the case of warrants, there is a range of forms from intimate or neutral to formal or deferential. In relation to these terminating devices, Placencia argues that making arrangement for future contact, expressing thanks, and welfare wishes are common components in Ecuadorian Spanish closings. Placencia also reports that there are some rather lengthy stretches of talk (e.g. expressing thanks, apologies) between the pre-closing and the terminal exchange. Placencia attributes this observation to the participants' desire to offer reassurance as to the continuation of the relationship. Thus, this finding provides further evidence to the claim that "to end an encounter it is necessary to affirm a non ending" (Albert & Kessler 1976: 165).

A survey of the literature on conversational closings reveals that research to date has highlighted complex properties, sequential structures, and operation of sequences of closings and their relationships to social and contextual variables. The findings obtained from these studies demonstrate that conversational closings are a complicated and delicate matter. Within this matter the participant tries to mitigate the face threat arising from the conduct of an end of communication to the other party, and tries to cooperate to reaffirm and consolidate the relationship between them. These studies

62 also show that conversational closings are realized through a range of routinized and formulaic sequences in a tum-taking activity, through which participants attempt to achieve both proper pre-closings and proper termination of a conversation (Schegloff & Sacks 1973). On the other hand, these findings confirm that different languages and cultures differ in the linguistic realization of closings and in the reflection of social characteristics of participants and interpersonal relationships by means of linguistic devices.

Findings from these studies have provided fundamental information that can be used for understanding the phenomena of closings in written communication, because written closings are found to have some similarities to spoken closings. The primary similarity is that, using Takami's words, the closing issue in both types of interaction is "a delicate matter because it can mean a sort of parting that may offend the other interlocutor if not performed appropriately" (2002: 67). For this reason, both speaker and writer need to take into account how to reaffirm and consolidate the relationship with the hearer or the recipient, and how to be polite and maintain mutual face. The underlying communicative purposes are strikingly similar. There are also some similarities in structuring closing patterns between oral and written interaction. For example, in both types of interaction there exist what Schegloff and Sacks (1973) call 'pre-closings' and 'terminal exchange,' although the ways of accomplishment and constitutive components of sequences vary between these two types of communications. Thus, closings in both oral and written interaction share some important properties and are relevant to each other to a certain extent.

2.3 Unresolved issues and research opportunities The first part of this chapter reviewed a range of sociolinguistic studies on phatic communion, speech acts, Grice's Cooperative Principle, politeness theory and the relationships between social and linguistic variables. These studies provide a basic framework to examine some phatic expressions, speech acts and politeness strategies employed in people's interaction, and as such they provide a rich multilayered approach to language use. However, this literature review shows that most of the early literature on these theoretical approaches has focused on a single tum of talk, and that the investigation of a series of moves under these theoretical approaches has not yet been

63 established and clarified. In recent research, early concepts of phatic commumon, speech acts and politeness phenomena have been challenged as a result of reconsidering extending speech acts or phatic expressions to the domain of discourse (Geis 1995, Moeschler 2002). As some sociolinguists have pointed out recently, rather than being performed on an utterance-by-utterance, tum-by-tum basis, speech acts are performed or undertaken over a stretch of talk. On the other hand, the survey of literature reveals that previous examinations are limited to oral communication. There are few accounts of how these pragmatic and sociolinguistic aspects are reflected in written discourse, such as in personal letters. Considering this, it is appropriate to say that phatic communion, speech act theory and the politeness concept should be explored in written communication. In particular, researchers should look into how some written routinized behaviours are structured and the manner in which sequences of speech moves (or acts) are related to each other in relation to these theoretical aspects.

This chapter has also surveyed a range of sociocultural and sociolinguistic literature pertaining to practices of addressing, openings and closings in conversation. The overview of these studies shows that these researchers have made a great theoretical and empirical contribution to the investigation of these 'rituals,' 'conventions' or 'routines' (Coulmas 1981) that are underpinned by sociocultural norms and strategies of communication in people's interaction. However, the preceding literature review reveals that there are numerous questions unanswered and issues that need to be explored. First, most existing studies have examined the performance of addressing, opening and closings in conversation (especially the latter two) by focusing on English. There is limited investigation of how these three speech events are performed in non English languages (e.g. Chinese), a language with speakers that make up about one quarter of the world's population. The need to analyse speech behaviours in other language communities has been expressed by a number of scholars (Duranti 1985; Hopper 1992; Hopper, Doany, Johnson & Drummond 1990/1991; Moerman 1988, 1991). As Moerman (1991: 175-176) argues, it is crucial to "examine talk-in interaction as it is conducted throughout the entire vast and varied range of living peoples."

64 Second, the majority of studies discussing addressing forms, openmg and closing performance have mostly rested on face-to-face or telephone interaction in English and a number of other languages. But there are few studies of how addressing, opening and closing work in written interaction. Whereas these speech behaviours are examined among the few examinations in the written area, the focus has mainly been on professional settings (Cherry 1988; Kalman 1996) or in the business domain (Bargiela Chiappini & Harris 1996; Pilegaard 1997; Yeung 1997), the research on these topics in the personal circle like personal letters is rare. Kalman (1996: 193) believes that personal letters are one of the communicative tools which are used widely among different cultures. Dittmer (1991) similarly observes that personal letters are potentially an important resource for understanding how social role relationships, emotions and societal conditions figure in written language. To build an entire base for investigating social behaviours like addressing, opening and closing performance in culturally varied situations, empirical studies on these social phenomena in personal written communication are clearly in order.

Finally, previous studies have been content to look at the relationship between social and situational factors (e.g. status, age, gender, settings) and linguistic realizations of these three routines in human's interaction, but they have paid little attention to considering the sociocultural construction in its larger contexts. Haberland and Mey (1977: 9) mention "no single aspect of linguistics can be divorced from the study of human societal practice." Thus, if the examination of addressing, opening, closing behaviours is treated as a study of pragmatics that looks at context-dependent meanings, the larger socio-cultural context should be taken into account (Lee-Wong 2000: 32). Putting it differently, these three social behaviours in interaction should be viewed in relation to what is considered to be culturally appropriate within specific contexts.

Due to limitations identified from previous studies, there arises the need to carry out further study of how addressing, opening and closing behaviours are accomplished in various types of interaction (e.g. personal letters, across different languages and cultures). The need for further investigations, which provides the opportunities for the present research, exist~ in the following areas:

65 The extent to which address forms, openmgs, closings in written interaction, particularly in personal letters, are accomplished in a similar way to those in the oral conversation;

Whether address forms, openings and closings function linguistically and pragmatically in written interaction, particularly in personal letters, in similar ways to those in oral conversation;

Whether the production of addressing, opemng and closing behaviours in written interaction, particularly in personal letters, is not only related to the social but also to the larger cultural context as part of contextual meanings.

2.4 Concluding remarks This survey of the literature and identification of some unresolved problems and new issues in this chapter demonstrates that the study of the performance of addressing, opening and closing behaviours in various types of interaction across different cultural contexts has yet to be developed. In order to build a strong case for examining the performance of these speech events in personal written communication in the Chinese context, a theoretically framework needs to be formulated and an appropriate method should be chosen to carry out the present research. The consideration of the theoretical framework and the methodology will be discussed in detail in the next chapter.

66 CHAPTER THREE METHODOLOGY

3.0 Introduction

While some routine speech behaviours and speech acts in oral communication have been discussed extensively, investigations into these aspects in written interaction have been limited. In this study an attempt is made to examine address forms, openings and closings in Chinese personal letters. The purpose of this study is to look into how Chinese people use language in these three speech behaviours. In particular, it is to explore their basic linguistic and pragmatic features as well as roles of social variables and the cultural constraints on language. In order to answer the research questions, the first step is to decide on the methods of data collection and data analysis. The former is to explicate a scientific approach for the data collection. The latter conceptually incorporates issues in relation to addressing, opening and closing behaviours, thus formulating an analytical framework and analytical methods for this study.

3:1 Data collection This section discusses collection in relation to the choices of a particular method in conducting the research. It concerns the consideration of methodological issues, including sampling (i.e. place, materials, selection of writers and recipients), and the procedure of data collection and data translation.

3.1.1 Sampling One of the most significant issues researchers have to consider when designing a project concerns the sampling principles, such as the type of data and the number of the respondents who will be included in the study. In this regard, Sankoff lists three types of decisions which influence the choice of sampling procedure

(l)The first is to delineate, at least roughly, the geographic, social, etc. boundaries of the speech community or subcommunity of interest, i.e. define the sampling universe.

(2)The second is to assess the possibly relevant geographic, social, and sociolinguistic dimensions of variation within the community, i.e. to construct a stratification for the sample. As the goal of sampling is to tap the existing linguistic variation in a community, it is important for the

67 investigator to consider carefully all types of variation which exist... Might age, sex, class make a difference? ... Do individuals speak differently in different social circumstances? Sampling should take account of at least these dimensions, as well as any others which might be relevant to a particular situation.

(3)The third type of decision is to settle on the number of informants and the amount of material to be collected from each, i.e. to fix the sample size. (1974: 21)

As suggested above, it is important to note that the sampling procedure IS not a straightforward one, but it depends on a number of considerations and factors. It is these considerations and factors that enable researchers to study a relatively small number of units in place of the target population, and to obtain data that can represent the whole target population. In short, sampling can be seen as a standard procedure of social research (Sarantakos 1993: 126). For this reason, the sample units for this study are selected by means of the methodological principles described by Sankoff (1974).

3.1.1.1 Place Data were collected in the Xinjiang Autonomous Region of the People's Republic of China between 1998-1999. This place was chosen because I worked in Urumqi, the political and industrial capital of Xinjiang, and had support networks there. This made it easy to access people and minimal time was needed for establishing local contact with prospective subjects. The second reason was methodological. Xinjiang, situated in Northwest China, and in the centre of the Eurasian continent, is the largest of China's provinces and autonomous regions. This region has many nationalities with the Uygurs and Han Chinese being the two major ethnic groups (47% and 38% respectively according to 1995 census). This is an important historical point of view, as Han Chinese, the largest ethnic group in China, moved to Xinjiang from other parts of the country both before and after the founding of the People's Republic of China. Mandarin is their native language in both spoken and written form. Thus, the distinctive immigration feature of Han Chinese in Xinjiang is in stark contrast to that of other provinces or regions with the minimization of influence of local culture on people's speech behaviours. Such a choice of the sampling area, therefore, could guarantee a high representation by participants as Chinese in a general sense.

68 3.1.1.2 Data 3.1.1.2.1 Chinese personal letters As stated by Chen (1999: 1), writing personal letters is a relatively common practice in China. An analysis of the content and functions ofpersonalletter writing indicates that personal letters are sent to family members, relatives, friends and familiars. They serve a variety of functions, such as exchanging information with each other, maintaining contact with those whom the writer is missing, expressing personal feelings and affections, admonishing younger people, and paying respect to older people (Nin & Liu 1993: 240).

A brief survey of the history of letter writing in China will shed light on some cultural characteristics of modem personal letter writing in China. As far as we can see from written records (Chen 1999), the presence ofletters in China dates back to the period of the Oracle Inscription, Shang Dynasty (11th to 16th century B.C.). The earliest letters on record are those that were exchanged between the emperor and supernatural beings by means of 'puci, ' the main content of the oracle inscription that was carved on bones or tortoise shells. The function of 'puci' in that period was mainly that of exchanging information between the emperor and the supernatural beings about the portents of agriculture, weather, and farming and hunting. This kind of letter was richly coloured by religion in relation to humans and Gods. In the Zhou Period (1027 B.C.- 777 B.C.) there was the introduction of another kind of special letter, 'jinwen, ' an inscription that was carved on to bronze objects. The contents of these inscriptions were mostly related to the rituals of certain social events, such as sacrificial rites, citations of punitive expedition and praise for ancestors. Within this inscription the structure of the content began to be fixedly formulated and the language was fairly formulaic.

As time passed, 'puci' and 'jinwen ' lost their original status as letters, and they were replaced by the old-style letters. Old-style letters referred to the classic letters used by people after the period of Spring and Autumn (770-475 B.C.). They were seen as an important means of communication between people at that time. They served to inform recipients of events, express ideas, affection and emotions, and pursue relationships (Chen 1999: 1). As the classical letters were written in the classical Chinese language, they were characterized by a very formulaic text format. For example, people used fixed formulae for the structure of letter writing (e.g. opening and closing parts) and

69 also used fixed expressions in certain sections, such as those of addressing and signing, which varied depending on the status of the writer and recipient. This style of letter writing remained essentially the same until early this century (Chen 1999).

Following the Vernacular Movement (1917-1919), classical letters were replaced by letters in the vernacular language, the Chinese language spoken by the people. There were many changes in the language and style of the vernacular letters during their development, but even so, some of the formulae exhibited in classical letters can still be found in modem letters, such as certain ritual expressions which depend on rank, age and other variables.

In more recent times, the tendency in China has been the increasing use of multiple means of communication, especially by the younger generation. However, as a useful communicative tool, personal letter writing is still the most salient activity in which writing skills are put into practice in people's daily life.

The data analyzed in this study consists of 259 recent (1997-1999) personal letters written in Mandarin by native speakers of Chinese. The 259 letters were part of the writers' normal personal correspondence and were addressed to 201 named Chinese recipients (some of them were written by different writers to the same recipient). They are divided into two categories: family letters (N=108) and non-family letters (N=151). Each type of letter is categorised by the nature of the relationship between the writer and the recipient. For example, the family letter is defined as a type of letter which is sent to immediate family members and relatives. It serves the purpose of informing recipients of family events, educating the young and showing respect to the old. An essential aspect of this type of letter is a heavy emphasis on the emotional and affectionate bonds between the writer and the recipient. The non-family letter is defined as a type of letter which is addressed to friends or familiars who have personal relationships with the writer, including teacher-student relationships. The primary aim of this type of letter is to exchange ideas or thoughts in order to make or maintain contact between two parties, although emotional and affectionate features can be found.

70 Letters within each type are also divided into the initial letter (N=l45) and the reply letter (N=ll4). Initial letters are defined in this research as letters in which the writer first contacts the recipient through correspondence, or resumes the writing after a long period of little contact. Reply letters are defined as a mode of writing with which the writer responds to a previous letter from the recipient in a situation of more frequent contact.

T a bl e 3 ..1 C ompos1t1on o fthe corpus o fltte ers Type Family letter Non-family letter Total Initial type 58 87 145 Reply type 50 64 114 Total 108 151 259

These 259 letters were selected from a total database of 318 letters. They were selected based on a closer examination of the letters, including checking the writing date (e.g. whether they are up-to-date letters), the writer's name (e.g. whether the letters are written by the same writer), the specific background information about the writer and recipient (e.g. whether the information identifying the role-relationship, age or gender is clear), the letter types (e.g. whether the letters are personal) and the readability of the letters (e.g. whether the sections are easily identified). This initially allows control over the independent variables of the letter type, the functions of the letter, and the writer and recipient. These independent variables are all difficult to account for in a large corpus of naturally occurring texts from various writers.

3.1.1.2.2 Reasons for choosing naturally occurring personal letters There are two reasons for choosing naturally occurring personal letters as the source of data for this research. The first one lies in the strengths of the ethnographic approach, which enables one to describe speaking rules and norms of interaction in people's real lives (Fong 2000: 221). Elicited data drawn from questionnaires or interviews can be umeliable (Dale 1978; Wolfson 1976), because such data cannot show the real use of particular linguistic forms or structures in naturally occurring contexts, nor show when or whether people other than those selected by the researchers (e.g. college students) perform certain speech behaviours or acts in real contexts (Holmes 1990: 164). However, the ethnographic approach seeks to minimize the umeliability of the behaviours of the people being studied. Through looking at particularly speech

71 behaviours in a natural setting, one can see authentic patterns and rules of interaction, and explain the relationships between speech behaviours or acts and social variables (Hammersley 1998: 8).

Another advantage of the ethnographic approach is that ethnographers can gain an actual set of information from the naturally occurring data, which is helpful for providing a much clearer picture of structural, constitutive and socio-cultural features of human interactional behaviours. As argued by Olshtain and Cohen (1983), one of the disadvantages of elicited data is the minimization of responses invoked by pre-designed contexts. This probably results from the respondents' seeking to reduce meanings to what is "observable" (Hammersley 1998: 9) due to their considerations of "being polite" (Olshtain & Cohen 1983: 32). On the other hand, respondents involved in the elicited data procedure (e.g. filling in space on a tape or a form; being videotaped in a role-play situation) provide responses which are possibly not shown in real interaction. Thus, elicited data in this sense are "likely to be misleading" (Holmes 1990: 165). In contrast, natural occurring data will offer an actual flow of authentic information, hence contributing more to our understanding of the sociolinguistic reality of human verbal communication.

Limiting the choice of letters to those of a personal kind is a socio-methodological consideration. It is widely accepted that personal letters are the most common written practice (Dierks 1999: 32). It is a writing mode which is suitable for all occasions in life and for all people in society (Richardson 1741, cited in Hornbeak 1934, cited in tum in Dierks 1999: 32). As a significant practice, personal letters cast light on the social boundaries of class, identities and gender. In addition, personal letters provide a glimpse into the cultural norms and beliefs of society underpinning the writer's practice. Thus, personal letters are indeed a reflection of the social richness of letter writing in general.

Personal letters are also most concerned with affection, which people do not express orally (Barton & Hall 1999: 5). Lakoff (1982: 245) notes: "Once we attempt to translate oral communication to the written page, we find ourselves having to translate meaning, as much as form." In this regard, Besnier's studies (1989, 1991) in

72 Nukulaelae, an African language, are quite suggestive. There, personal letter writing is considered to be a cathartic communicative event. The written mode is considered to be intrinsically more affective than the spoken mode. Specifically, the opening and closing sequences of letters accurately reflect the affective state of the writers. In an analysis of the relationship between affect and letter-writing in a Japanese context, Kataoka (1997) demonstrates that young Japanese women tend to exploit affective shape, form and function in order to establish relationships of intimacy and solidarity. He argues that in casual letters these features serve to achieve reciprocity, but also create a new mode of literacy influenced by social change. Taken together, this evidence shows that this kind of written contact provides an opportunity for research into immediate affective interaction.

In addition to its social and affective function, personal letter writing is another activity in which politeness.is expressed in actual usage (Cherry 1988). There are two reasons for this. First, certain parts of personal letters, such as the addressing, opening and closing sections, are particularly relevant to the level, form and distribution of politeness strategies (Nevalainen & Raumolin-Brunberg 1995; Pilegaard 1997; Zhu 1997). Second, the aim of personal letters is primarily to establish or maintain relationships between writer and recipient, thus, a great deal of attention is paid to politeness, conveying a high degree of respect or subordination. In this sense, personal letters are characterised by the conveyance of politeness.

The most important features of personal letters are their fulfilment of a set of symbolic conventions which demonstrate contexts of hierarchy and equality, and their fulfilment of a set of formal rules which indicate deference or intimacy. It is necessary and important to systematically investigate personal letters with respect to their particular forms and functions of specific sections or contents. The aim is to better understand how the writers use these language forms to accomplish their objectives and confirm their relationships with the recipients (Pong 2000: 221).

3.1.1.2.3 Writers and recipients This section provides a description of the general background of the writers and recipients who produced and received the letters used in this study. As mentioned

73 previously, these letters were written by 259 individual native speakers of Chinese. They included 26 parents, 18 children, 48 siblings, 16 relatives, 115 friends, 10 teachers and 26 students (see Table 3.2). All writers were urban residents of different parts of Xinjiang. The recipients (N=201) were drawn from university classes where the researcher worked, or were friends or acquaintances of the researcher.

Table 3.2. Information about social roles of writers Social role N Family letters Parent 26 Child 18 Sibling 48 Relative 16 Non-family letters Friend 115 Teacher 10 Student 26 Total 259

Writers and recipients in this study include both males and females, and they vary in age and educational level. Among the writers there are 124 females and 135 males. Forty five were aged 40-60; eighteen were aged 30-39; one hundred and ninety six were aged 18-29. One hundred and twenty two writers had a secondary education and one hundred and thirty seven had a university degree. All of the age, sex, and education background information about the writers have been set out in Tables 3.3 and 3.4.

T abl e 3 ..3 B'10grap1 hi ca1 intiormation ab out wnters Age (years) 18-29 30-39 40-60 Total Gender F M F M F M F M Parent 0 0 0 0 8 18 8 18 Child 9 9 0 0 0 0 9 9 Siblings 23 14 5 4 1 1 29 19 Relatives 4 3 1 1 3 4 8 8 Friends 46 62 4 3 0 0 50 65 Teachers 0 0 0 0 3 7 3 7 Students 17 9 0 0 0 0 17 9 Total 99 97 10 8 15 30 124 135

74 Table 3.4. Education background of writers Qualification Secondary school University Gender F M F M Parent 8 14 0 4 Child 0 9 9 0 Sib linE_ 17 10 12 9 Relative 4 6 4 2 Friend 17 35 39 24 Teacher 0 0 3 7 Student 2 0 15 9 Total 48 74 82 55

This selection was motivated by the need for a sample of writers from different social backgrounds in terms of role, age and gender. This will help to achieve both psychological and sociological validity and will reflect the perspectives of both individuals and social groups or communities (Sherzer 1974: 17). On the other hand, the diversified social and situational backgrounds will reveal that the different writers perform certain speech events and express certain feelings in different ways. Thus, the sample will be a partial reflection of the social differences within China as a whole. However, this is not to say that this sample includes all social groups. Rather, it was limited to writers of a certain range of ages (18-60) and educational background. Hence, the results can be generalized to populations similar to those from which the sample was drawn.

3.1.2 Methods for data collection Following the ethnographic approach of data collection, in the spring of 1998 I first asked my students at Xinjiang Normal University, friends and acquaintances to gather letters from family and non-family settings. During the first contact, I explained the general objective of the study and the criteria of the role (listed in Tables 3.3 and 3.4) and age (18-60). Considering the sensitivity of privacy, I further explained that the letters were only for research, and no personal information would be open to the public (All names presented in this study are pseudonyms). I also explained that this study was only to investigate how address forms, openings and closings appear in personal letters. The referential information in the main body would not be read or revealed publicly or privately. After the potential subjects agreed to lend their letters for this study, I gave each ofthem pieces of face sheet to fill in, which assessed the information about writers in regard to age, gender, role-relationship to the subjects (recipients) and educational background. In the Autumn of 1999 I returned to Xinjiang, to collect

75 additional data following the same procedure. Once the letters were collected, I listed them according to the information about the writers and recipients cited in the face sheets. Finally I selected those which met the criteria set up for this study.

The resulting sample is probably less heterogeneous than one could collect by randomly sampling the writers in the Xinjiang area. This sampling procedure ensures a reasonable spread according to the criteria, limiting the degree to which the result can be generalized to the same broad social groups. In real life, it is likely that a writer would have a similar behaviour to a person in the same broad social group. Thus, this desire maximizes the potential influence of roles, age and gender. Thus, examining the writers' relevant performance in letters within a fixed social group allows this research to automatically control the social variables.

3.1.3 Transliteration As the written form for the Chinese language is in characters and there are technical difficulties in printing Chinese characters in the text paper, a Romanization system is thus adopted to present the data. 'Hanyu pinyin' (Chinese alphabetic system), the most commonly used system for Chinese Romanization, which was officially adopted in 1958, was applied to transcribe the data cited as examples in this study.

Since the Chinese language is a tonal language, a syllable in Chinese 'pinyin ' without a tone marker can have various referents if not contextualized. Even if there is a tone marker, it is still difficult to figure out the meaning of a non-syllabic word in 'pinyin ' due to a great number of homophones, that is, words with the same pronunciation, but with different and unrelated meanings. For this reason, two ways for translating the data are provided in this study: glossing and a combination of literal translation and free translation.

Glossing refers to word-for-word translation. In this study it covers all the words and particles like 'rna' or 'ba.' However, the markers of tense, aspect and passive voice are not shown in the translation. The key point for this part of the translation is to comprehend the original Chinese version thoroughly, and then put it into idiomatic Chinese (Xu 2000: 7). In order to achieve this result, the researcher strives to retain the sentiments and styles of the original by taking sentences as basic units and reproducing

76 both their ideological content and style. When it is difficult for the researcher to do literal translation, however, free translation is adopted. According to Xu, free translation is an alternative way adopted to reveal the meanings and spirit ofthe original without trying to reproduce its sentence pattern or figures of speech (2000: 8). For example:

Example (1) (daughter to father) (1) Jian xin ru mian. (2) Jinlai shentijiankang, shenhuo yukuai, see letter as face recent body healthy life happy

)'lqte dou hao ba? everything all good (isn't it) It is as if we were face to face when you are reading this letter. How has your Health, your life and everything else been recently?

In this example, 'Jian xin rumian ', glossed as 'see letter as face,' can be translated literally as 'As if you saw my face when you are reading this letter,' but the writer's affective meaning, an important feature of this Chinese conventional greeting, is lost. Actually, in the Chinese, 'rumian' means 'as if we were face to face,' conveying the writer's affective feelings towards the recipient and also emphasizing mutual understanding and mutual love between them. Therefore, to capture the implication of the phrase 'rumian, ' the free translation in English is provided.

3.2 Data analysis 3.2.1 Framework for analysis This section explains the analytical framework created to investigate the phenomena with regard to the research questions. The purpose of this research is to examine Chinese people's performance in addressing, opening and closing behaviours in personal letters, arriving at understanding of the constructions of form, function, and meaning, of how they are realized linguistically and socially, of how they interact with the social variables and of how they are derived from that in the context of the prevailing socio-cultural constraints. Therefore, the analytical perspectives rest on three aspects: 1) identification of sections of letters; 2) identification of constitutive components occurring in the three sections; 3) identification of sequential organization of components in certain behaviours; 4) identification of social values elicited in performed speech behaviours and acts (e.g. formality/informality, intimacy/distance). Below is a summarised definition of each category and the coding rules giving a general

77 background for the analytical framework. Detailed explanation about the rules associated with each phenomenon will be provided in later chapters.

3.2.1.1 Identification of sections of personal letters These identifications include the coding of the sections of personal letters at the text- sequential level. According to Chen's division (1999), the personal letter is divided into five sections, as illustrated in Figure 3.1 : the address section which contains the address form; the opening section which contains salutation and other strategies, such as inquiries or certain statements making reference to previous correspondence; the main body section, where the central communicative information and illocutionary acts are expressed; the closing section which indicates that there is no further topic and the end of exchange is reached, and often contains some routines, such as good wishes or the sign-off phrase; and the signing s.ection which contains the writer's own name. As the first, second and fourth section are relevant to the present study, they are described in details.

Addressing section

Opening• section

Main• body

Closing• section

Signing• section (based on Chen 1999)

Figure 3.1. Text-sequential realization of a personal letter

In reality, the opening and closing sections vary extensively or very occasionally may be missing altogether. For the sake of illustration, the opening and closing sections are identified based on these criteria: the opening is made up of single or combined ritual statements used to identify the relationship between the writer and recipient (e.g.

78 greetings, inquiries); the closing contains single or combined ritual expressions, which directly or indirectly indicate the end of written interaction and imply the continuation ofthe relationship between writer and recipient at the same time (e.g. pre-closings, good wishes). One distinctive feature in both sections is that the constitutive components are characterized as being relationship-oriented expressions (phatic expressions), rather than referential messages (with a few exceptions) no matter whether it is a single strategy or combined one.

3.2.1.2 Identification of linguistic forms/components/sequences 3.2.1.2.1 Identification of linguistic forms in the addressing section Zhu (1997: 17), in her study of politeness phenomena in Chinese personal letters, identifies twelve themes of address forms. Her categories are as follows:

In family: 1) Intimate term alone 2) Intimate term + name 3) Name+ kinship term 4) Kinship term alone 5) Full name 6) Intimate term + kinship term 7) Kinship term + name 8) Intimate term + name + pronoun + kinship term 9) Intimate term + kinship term + intimate term + name

In non-family: 1) Simple structure 2) Dual structure 3) Triple structure

Generally speaking, Zhu's framework has provided a detailed classification of address forms in Chinese letters, which lays a basis for future comparative studies. The main disadvantage of this classification system is that it is hard to fit it into a general analytical framework because of a lack of consistency in structuring. For example, Zhu categorizes constituents from (1) to (9) in family letters by looking at what terms are involved or combined, while (1) to (3) in non-family letters focus on the grammatical structure in each category. This non-consistent classification makes it difficult to compare the addressing structures between family and non-family. Also there is confusion in the division of the addressing structure in Zhu's classification and as a

79 result, it is difficult to leave open the possibilities as address forms in real life. In order to achieve clarity and consistency, therefore, the present study classifies each address form into three super-categories:

Table 3.5. Descriptive framework for address forms in the addressing section ( 1) The name-alone strategy (2) The compound name strategy (3) The no-name strategy

These super-categories are formulated in the examination of the Chinese written address forms by focusing on three viewpoints: 1) addressing practice with name; 2) combination of name with other term (e.g. title, kinship term); 3) addressing practice without name. The advantage of this identification system is that it can systematically investigate how each address form reflects linguistic and social meanings, and how the form reflects the characteristics of the writer and recipient, the social relationship between them, and the socio-cultural norms within which it is embedded. What can be derived from these categories is the influence of hierarchy position, social distance, and the attempt to cope with cultural norms in the practice of addressing.

3.2.1.2.2 Identification of components in the opening and closing section In order to proceed with the analysis, it is necessary to identify and code the components or strategies occurring in the opening and closing sections. For the present purpose, the identification and coding are centred on two interrelated aspects: major types of speech acts of each component and specific speech acts of each component.

Major types of speech acts of each component As is discussed in chapter two, based on the content and intent of speech acts in human interaction, Searle (1976) proposes a framework that has been regarded as the basis for an empirical classification system. This framework contains five categories of speech acts:

a. Directives: e.g. requesting, asking, advising, inviting.

b. Assertives: e.g. claiming, affirming, suggesting, swearing, boasting, concluding.

80 c. Commissives: e.g. planning, promising, threatening, vowing, opposing, betting.

d. Declaratives: e.g. declaring, res1gnmg, adjourning, appointing, nominating.

e. Expressives: e.g. greeting, apologizing, thanking, condoling, congratulating. (based on Searle 1976)

This set of classifications is built based on performative verbs. This theory can be applied to "natural language, to examine text and discover how the syntax places a Speech Act (SA) in one category" (Cicognani & Maher 1997: 4). However, there are problems with this and similar classifications of speech acts. Specifically, there are some acts that we cannot simply classify into the categories of speech acts only according to their speech act verbs. For example, a problem encountered in this study is that there is difficulty in determining an inquiry after a recipient's state of health and well-being as an 'expressive' or a 'directive,' an acknowledgment of the receipt of a previous letter as an 'expressive' or an 'assertive,' and 'offering help' as an 'expressive' or 'commissive' act, because these acts appear to contain some other function (e.g. requesting, informing and promising) apart from expressing the writer's feelings and attitudes.

In order to establish a criterion for determining the nature of speech acts which emerged from this study, this research adopts the pragmatic classification approach partially based on Stekele-W eithofer and Kannetzky' s framework regarding the "pragmatic analysis of the form(s) of communicative and joint actions" (2000: 1). Normally, Searle's classification theory presupposes propositional content and conventional linguistic meaning. The pragmatic classification approach of speech acts here, in contrast, does not focus on the pre-established meaning and the content of expressions, rather, it highlights the writers' implicit intention and particular social meanings (functions) expressed through a particular act in a particular context. Borrowing Austin's (1962) treatment ofperlocutionary effect, it emphasizes the particular feelings expressed in certain sentences. If intending and particularizing are presupposed as basic for identifying the nature of speech acts, it can be assumed that most speech acts

81 involved in the opening and closing sections m this study can be described and explained under the category of' expressives.'

It is commonly accepted that 'expressives' are essentially subjective and have few things to do with the world (Mey 2001). They are generally used "simply to express the speaker's inner psychological state or to express a particular attitude that is represented by the propositional content of the utterance" (Holtgraves 2002: 15). Also, sometimes they are used to signal the speaker's hope that something positive will happen to the addressee. As pointed out by Mey, because of this subjective characteristic, the expressive speech act is subject to limitations and changes according to different conceptualizations of a certain social behavior (2001: 121). This view offers a relevant backdrop for this study. In their examination of openings in oral conversation, sociolinguistic studies of English and other language communities (e.g. Kasper 1989; Omar 1992; Shih 1986;·Yahya-Othman 1995) treat the 'inquiry' act as a 'phatic inquiry' or 'phatic question.' According to them, the main concern of the inquiry is not to exchange information about the participants, but simply to make the language reflect concern for the addressee and for those who are intimate to the addresser (Schmied 1991: 92). Also an inquiry, especially between intimates, serves to indicate recognition that the addressee is a member of a group and not an isolated individual (Yahya Othrnan 1995: 218).

This holds true for Chinese society. In Chinese culture, asking about the recipient's current situation in health or other personal affairs in personal letters is regarded as an important means of conveying the writer's consideration and affection to the recipient based on the notion of 'renqing' (human warmth) since ancient times. The purpose of doing so is not to get a truthful or detailed answer, but is to strengthen 'expressive ties' (Bond & Hwang 1989) and the affective relationship between participants, especially between members of a family. Thus, in considering its psychological characteristics and the writer's phatic intentions, the act of 'inquiry' in this study is therefore classified as an expressive act.

This is the same in the case of the acknowledgment of receipt of a prevwus correspondence. In China it is a convention that writers acknowledge the receipt of the

82 previous correspondence by writing a reply letter. Although such an announcement of the arrival of the previous letter bares some factual information to some extents, it mainly serves to convey the writers' attitude or feeling towards the recipients (e.g. catering to their expectations). It does this mostly through the expression of the writers' positive feelings towards the receipt of the pre':'ious letter (i.e. confirming the writer and recipient's shared membership of a social group) and through the implication of the writer's intention of making the recipient relaxed. Based on this, the acknowledgment is psychologically real and salient, and hence, an 'expressive.'

The speech act of 'offering help' in the closing section of personal letters is another example of having expressive functions with the features of other type of acts. From the surface level, this speech act, such as 'Ruyou shenme kunnan, jinguan gaosu wo, wo yiding hui bangmang de.' (Ifthere is any difficulty, please let me know, and I will give a hand.), seems to commit the writer to some future course of action, which appears to be related to a real-life situation. But under the sentence meaning, it is found that the writer's purpose by using this act is not to assume a particular action he or she would take to help the recipient, rather, to convey his or her positive feelings or attitudes towards the recipient (e.g. in-groupness, friendship) and to emphasize the intention of making the recipient happy. Here, too, the emphasis of this act is its emotional and attitudinal aspects, rather than the referential content, hence an expressive act.

Based on such a pragmatic rule of judgments about the nature of a speech act (Stekeler Weithofer & Kannetzky 2000), this study finds that in the letters analyzed, the speech acts used were overwhelmingly expressives, among which very few were sentences at 'declarative': pre-closing statements, such as 'Zhanci, yuyan houxu' (I have to stop here for now. The rest will be talked about next time.). Thus, it proposes one category of classification for speech acts occurring in both opening and closing sections: the expressive act. This framework is developed to provide flexible ways of deciding the characteristics of some speech acts by focusing on intentional semantics in relation to phatic aspects.

83 Specific speech act of each component Within this category, each component is classified according to the speech function the writer tends to achieve in a particular context. As Wierzbicka points out, "every culture has its own repertoire of characteristic speech acts and speech genres" (1991: 149). Wierzbicka (1991: 150) also states that one particularly appropriate way of approaching the repertoire of speech acts and speech genres characteristic of a given culture is through their folk names, the special lexical units which have come to encode a culture's view of its most relevant forms of talk. In Gumperz's words, "members of all societies recognise certain communicative routines which they view as distinct wholes, separate from other types of discourse, characterized by special rules of speech and . nonverbal behaviours" (1972: 17). Thus, as Hymes states, "one good ethnographic technique for getting at speech events, as at other categories, is through words which name them" (1962: 11 0).

Wierzbicka also insists, however, that some acts such as 'promising,' 'ordering' or 'warning' cannot be viewed as culture-specific (1991: 152). As argued by Walters, "every language makes available to the user the same basic set of speech acts, such as requesting, apologising, declaring and promising, with the exception of certain culture specific ritualised acts such as baptising, doubling at bridge, and excommunicating" (1980: 78). The notion underlying this assumption is that if "one language permits an act such as requesting, every other language will" (Walters 1980: 79).

Following Wierzbicka (1991), the components of openings and closings are identified from two approaches. In the first, some speech act characteristics are described and encoded as culture-independent natural kinds through their illocutionary and perlocutionary forces, such as greeting (e.g. Malinowski 1972; Orner 1992; Yahya Othman 1995), apologies: (Holmes 1990, 1995; Olshtain & Cohen 1983), thanking (e.g. Aijmer 1996; Coulmas 1981), or good wishes (e.g. Pong 2000; Verschueren 1981). In the second, some speech acts are described in relation to Chinese cultural norms. This approach is free of any ethnocentric bias and is culture-specific.

In the first approach, each component associated with expressive acts is due to the corresponding verb used in performing it. On the illocutionary level, these components

84 are semantic in origin and to be described at the semantic level in terms of relationship between sentences (Norrick 1978: 277). On the perlocutionary level, in identifying components in this approach, both the function of the act and the writer's attitudes and intention are taken into account. In other words, the components cannot be coded without considering the speech function and the writer's intention.

In the second approach, some components are identified on the basis of the characteristics of certain communicative routines in Chinese culture. For instance, according to DeCapua and Dunham (1993), the sentences of 'The final examinations are approaching, I hope you are studying hard and learning a lot, and making good friends' can be heard as the speech act of advice due to "the recommendation of a course of action which is recognizable as a potential solution" (Bayraktaroglu 2001: 178). The present study, however, identifies these statements as a speech act of hope. The reason is that these statements do not seem to suggest that the recipient is having trouble, in which case the recipient would interpret them as negative criticisms and as insults to their social competence. Here, these statements are similar to good wishes, being sympathetic and conveying the writer's concern for the recipient. Say that whilst this can be interpreted as advice, it still has strong functions of an expression of hope, an expressive act.

Another example is related to the closing section. Following Okamoto's conceptual framework (1990), the sentence 'On the whole, I have been all right here, so don't worry about me' can be regarded as a 'summary statement' of the content of the conversation of the writer, that is, an assertive. In this study, it is not treated as a purely summary statement. Rather, it serves to comfort the recipient with positive messages (which may or may not be true), thus aiming to relieve the recipient of worry or anxiety. In this sense, it is appropriate to take the phrase as an act of 'reassurance about oneself,' an expressive act.

Therefore, this study categorizes a range of speech acts or speech genres that are either universal or specific to the Chinese context, as shown in Figure 3. This framework is used to analyse components of the opening and closing sections ofletters:

85 Table 3.6. Descriptive framework for components in the opening and closing section Components of opening* Components of closing* (1) Greetings (1) Expressions of good wishes (2) Inquiries (2) Pre-closings (3) Acknowledgments ofreceipt of a previous letter (3) Longing for future contact (4) Expressions of apologies (4) Sign-off phrases (5) Expressions of good wishes (5) Expressions of concern (6) Expressions of missing (6) Reassurance about oneself (7) Expressions of concern (7) Farewell messages (8) Expressions of thanks (8) Sending regards to others (9) Reassurances about oneself (9) Offering help (1 0) Expressions of congratulations (1 0) Expressions of thanks (11) Expressions ofhope (11) Expressions of apologies (12) Sending regards to others * Ordering the components is based on the frequency of occurrence of each component in this study

3.2.1.2.3 Identification of sequences in the opening and closing section As discussed in chapter two, most studies have developed their sequential schemata of openings (e.g. Hopper 1991; Hopper & Chen 1996; Houtkoop-Steenstra 1991; Lindstrom 1994; Schegloff 1968, 1986) and closings (e.g. Clark & French 1981; Okamoto 1990, 2002; Placencia 1995; Schegloff & Sacks 1973) by focusing on turn taking in talk-in-interaction. These studies are based on conversations in which the interactants are visually or audibly accessible to each other, and the opening and closing exchanges systematically establish an interactional space in which possible relevant topics can be worked out by the participants (Hutchby 1999: 44). The personal letter, however, is a particular kind of written discourse, in which communication takes place in a non-interactive and one-way mode divorced from immediate response and exchange of information and affection. The lack of a visual or audible channel thus results in written openings and closings taking distinctive sequential forms.

Laver (1975, 1981) is among the first sociolinguists to treat linguistic routines of openings and closings (i.e. greeting and parting) as extremely significant communicative strategies for the negotiation and control of social identities and social relationships between interactants in conversation. Underlying Laver's functional approach is that the course of conversational interaction develops according to two needs: the need to communicate as efficiently as possible and the need to be polite (1981: 289). Based on this principle, Laver comes to the conclusion that the functions of opening and closing "in the crucial marginal phrases of encounters when their participants' psychological comfort is most at risk" are "to establish and consolidate the

86 interpersonal relationship between the two participants [and to ease] the transition to and from interaction" (1975: 236).

Laver's proposal provides a helpful tool for interpreting the written data for this study. The advantage of Laver's framework is that he puts the opening and closing encounters into a social process in which the social relationships and social identities of participants are negotiated and controlled. Thus, this perspective defines and shapes how the ritual and phatic expressions in these two events are exploited to fulfill social functions, and how they interact with social variables and personal circumstances. This theoretical perspective can be used to ask and answer how the initial component generates the second, why the intense or elaborated structures occurs, how the components are ritually and structurally related to each other, and how purposes or meanings of the writer are deployed in a certain sequence. Therefore, this study adopts Laver's function-oriented framework of phatic communion in the examination of the opening and closing sequence m personal letters. This identification framework contains five key ingredients:

Table 3.7. Descriptive framework for sequences in the opening and closing section (I) To observe the basic components of sequences (2) To identify social and sequential functions of the basic components (3) To observe how the preceding component makes the next move available socially and ritually ( 4) To identify some kind of structural and social relationship between components (5) To seek the causes of the relationship between the components

Developing this framework brings to light the structural properties of opening and closing events in personal letters. It aims to disclose the system or ritual properties of these two behaviours in the one-way interaction, that is, how the preceding component constrains the next move, or in tum, how the next action is related to the first component. In addition, this framework also emphasizes that rules of the sequences cannot be understood without considering social meanings, conventions and norms embedded within the sequence. That is to say that management and deployment of components may tell us something about the structure of social relationships and the general relationship between ritual structures and social norms. In brief, formulating

87 such a framework provides a helpful tool for analyzing writers' performance of openings and closings in written interaction.

3.2.1.3 Identification of social values To identify key social values reflected m the addressing, opening and closing performance in personal letters, a framework is built on previous research concerning certain relevant values, such as formality/informality, distance/intimacy, respect/consideration and the Chinese cultural norms and beliefs underpinning these values. This framework is used to reveal different attitudes and different ways in which Chinese writers produce their language in certain events, and to uncover the various kinds of social phenomena associated with Chinese cultural norms and beliefs.

3.2.1.3.1 Seniority/juniority As defined in the Oxford English Dictionary (1989: 455, 631), seniority is related to the position of a senior or superior who has or who exercises dominion. In this study, seniority is restricted to the conventional assumption that Chinese are submissive to the authorities of age and status (e.g. parents, old relatives, teachers) (Yang 1989: 127). From a historical point of view, Chinese people's attitudes towards authority are based on father-son relationships (Ebrey 1991: 46). Fathers' linguistic behaviours towards their sons are characterized by the strict, personalized and direct imposition of authority. This family-based assumption of authority is found throughout China, where people share many assumptions about human relations. Juniority, in contrast, refers to the state or condition of being junior in age or rank. It means youthfulness, lowly position and lowly standing (e.g. children, young relatives, students). For Chinese people, juniority is usually characterised in linguistic behaviour by the showing of respect, obedience and dependence towards seniors or superiors.

3.2.1.3.2 Formality/informality Formality and informality are concepts frequently used by sociolinguists to describe social occasions and associated behaviours (Irvine 1979). For this study, formality is narrowed down to the concept which refers to linguistic behaviour emphasizing respect (for an established order of social position and identities) and the degree of social distance (as opposed to intimacy). The main property of this concept is asymmetrical:

88 Writers use a deferential and formal strategy for recipients who are of high status and are distant. For example, the address form 'Zunjingde Wang laos hi' (Respected Teacher Wang) used by a student to a teacher in a college conveys both formality and respect, emphasizing the high rank of the teacher and the hierarchical relationship between them.

The concept of informality is culturally specific. According to Wierzbicka, informality refers to "the purposeful rejection of any overt show of respect, with the implication of familiarity, friendliness, and equality" (1991: 111). Lee-Wong (2000: 170) argues that informality for the Chinese is concerned with familiarity and friendliness, like the relations between social equals (e.g. siblings, friends). For example, equals often use given names to indicate informality, emphasizing in-groupness. However, for Lee Wong, familiarity is not always associated with equal relationships and the notion of familiarity does not mean rejecting any indication of hierarchy (2000: 171). This holds true for this study. For example, a teacher addresses his student as 'Hailin tongxue' (Bailin fellow student), which is an informal expression denoting great familiarity but also hierarchical difference. Such a phenomenon can be explained by the fact that the Chinese hierarchy in address forms is not based only on the achieved status but also on the ascribed status encoded in such forms (Lee-Wong 2000: 172). In this sense, Wierzbicka's perception that identifiable values cannot be understood without considering the cultural norms and values of a given society is very important for social research (1991: 113).

3.2.1.3.3 Distance/intimacy Social distance is defined on a spectrum ranging from intimate to distant (Lee-Wong 2000: 65). According to Hudson, distance or intimacy concerns how many social characteristics people share (e.g. sex, age, identity of occupation, interests) and how far they are prepared to share intimacies and other factors (1996: 122). In this study, intimacy is defined by the term 'qin' (close): Two people know one another well and have 'good feelings' for one another (Wierzbicka 1991: 105). It incorporates not only kinship ties but also friendliness. As Lee-Wong states, it is related to 'qinren ' (close people) which refers to one's parents, children and one's family members and 'qinyou' which is reserved for more distant relatives and friends (2000: 65). In contrast, distance

89 is defined as a relationship where "there is a complete or marked absence of structured relationship" (Lee-Wong 2000: 65). One would, for example, treat as distinct the social relationship between two strangers or interactants for transactional purpose in talk-in interaction. On the other hand, even with an established or structured relationship as in the workplace, two persons can also be described as distant. In this study, teacher and student are considered people distant in relationship, although one can form a close relationship with the other (see chapter four).

There are many factors which can contribute to determining the degree of social distance or intimacy between people, such as relative age, sex, social roles or membership of the same family and so on (Holmes1992: 247). As an example in this study shows, on a certain level, members of one's immediate family have a closer relationship with each other than with relatives of the family.

3.2.1.3.4 Respect/consideration It is generally accepted that respect is a way of conveying that the other is of higher status than the self (O'Driscoll 1996: 25). Since ancient times, respect has been regarded as a central part of the Chinese life (Lee-Wong 2000: 243). According to Confucian ethics, the basis of human goodness 'ren ' is filial piety and deference: Good people should exhibit loyalty to their rulers, filial obedience to their parents, and brotherly deference to their older brothers (Ebrey 1991: 55). By conducting themselves according to this principle, people most commonly show deference verbally by using an honorific and polite form when speaking upward towards seniors (e.g. parent, teacher). For example, when sending a greeting to seniors in letter writing, juniors (e.g. child, student) usually use the honorific pattern of 'Ninhao' (How are you?), with which the social status of age or rank is explicitly marked. In Chinese society respect for seniority is always regarded as polite behavior. This complex politeness strategy simply reflects the rigid hierarchy of society (Zhan 1992: 8).

Consideration can be seen as a demonstration of concern for the other in any interaction (Lee-Wong 2000: 244). In Chinese society, showing consideration for another person is an example of human goodness (Bitchener 1999: 32). According to McNaughton (1974), consideration is both a natural feeling for others according to one's relation to

90 them and a mutual belief on the basis of knowing how it would feel to be the other person. As argued by Zhan (1992: 9), conveying concern for the other can be seen as the core of the positive politeness strategies of Chinese: Chinese speakers try to be as close to the addressee as possible. For example, family members and close associates usually like to inquire after the recipient's health status or other personal state in letter writing as is shown in this study. This phenomenon indicates that Chinese people highly value kindness and compassion highly. To Chinese people, the warmer the feelings, the smaller the distance between the speaker and the hearer.

This section illustrates the analytical framework which will be used to investigate the linguistic and pragmatic features of addressing, opening and closing behaviours in Chinese personal letters. The illustrations are concerned with how to identify the sections of a letter, how to identify linguistic forms of address, how to identify both components and sequences of openings and closings, and how to identify social values elicited in these three behaviours.

3.2.2 Data analysis After developing the analytical framework of the investigation, the next task for the researcher is to study the data and search for trends, patterns and relationships relevant to the research questions. To do so, it is important to choose an effective way to analyze the data. It includes deciding on either a quantitative or qualitative orientation. Thus, before making any decision, it is necessary to compare the nature and the characteristics of the alternative methods.

3.2.2.1 Quantitative vs. qualitative analysis Quantitative data analysis focuses on giving numerical results which can be reported in tables, graphs and charts presenting numbers, proportions, or trends. Quantitative approaches are used to answer questions such as 'how many' and 'how often,' which allow researchers to discover which phenomena are likely to be genuine reflections of the behaviour of a language or variety (Bouma 2001: 171). In quantitative research people classify features, count them and construct more complex statistical models in an attempt to explain what is observed. Findings can be generalized to a larger population and direct comparisons can be made between two corpora, so long as valid sampling

91 and statistical measures have been used. Thus, the main concern for quantitative researchers is to seek explanations and predications that will be generalizable to other persons and places (Glesne & Peshkin 1992: 6). However, the picture of the data derived from quantitative analysis is less rich than that obtained from qualitative analysis.

In a different manner, qualitative data analysis is used to describe an impression, to tell what kinds of things there are and to reveal what they look like (Bouma (2001). The aim of qualitative analysis is to find out what is going on and what is happening in a group or in a community, such as who speaks to whom "with what message, with what feelings and what effect" (Bouma 2001: 172-173). Instead of focusing on empirical and analytical aims, this kind of research focuses on "how social life is organized within institutional discourse, and how knowledge and power are implicated in them" (Miller 1998: 36). Malinowski once argued that anthropologists should "grasp the native's point of view, his relation to life, to realize his vision of his world"(1922: 25). Also, Bryman notes that "the most fundamental characteristic of qualitative research is its express commitment to viewing events, action, norms, values etc. from the perspective of the people who are being studied"(1988: 61). These kinds of reasoning challenge the usual ethnomethodological preference for examining social interaction by analyzing how textual realities are constituted. Qualitative analyses advance social research practice by focusing researchers' attention on to the social organization of everyday life.

Putting together, it is clear that the main characteristics of qualitative approaches are open and evolutionary although it leads to a lack of standardization. It is the openness of qualitative inquiry that allows the researcher "to approach the inherent complexity of social interaction and to do justice to that complexity, to respect it in its own right" (Glesne & Peshkin 1992: 7). Qualitative researchers deal with multiple socially constructed realities, which are complex and indivisible into discrete variables. Their major tasks are "to understand and interpret how the various participants in a social setting construct the world round them" (Glesne & Peshkin 1992: 6). The main disadvantage of qualitative approaches is that their findings cannot be extended to wider populations with the same degree of certainty with which quantitative analyses can.

92 From this brief discussion it can be deduced that both quantitative and qualitative analyses have something to contribute to social science, but there are differences between them. Bouma believes that the distinction between quantitative and qualitative analysis is not that one is regarded as better or worse, rather, whether it is appropriate to the question asked (2001: 174-175). Recently, there has been a rise in the social sciences towards the use of the multi-method, which tends to use both quantitative and qualitative analyses. Bouma argues that well-designed qualitative research often serves as essential preparation for significant quantitative research and vice versa (2001: 174). Johnstone points out that analyzing sociolinguistic data often involves both some explicit or implicit counting to answer questions regarding how often things happen and some description to answer questions regarding how and why things happen (2000: 37).

3.2.2.2 Combining statistical analysis and in-depth interpretation This present study adopts both simple statistical measures using quantitative techniques and the in-depth interpretation ofthe qualitative approach. There are several reasons for combining these two analytical methods. First, this research involves the objective of answering questions such as 'how many?', 'how often?', 'what proportion of writers?', hence, it is necessary to give the numerical results by using statistical measures. Counting how many there are of each form, pattern and sequence and finding out the occurrence of each among social roles, ages or genders allows differentiations, and classifications and orderings. Comparing these numerical results through counting and ranking enables the researcher to unearth a trend, such as the most commonly used form or pattern in the addressing, opening and closing events and the perspective of a certain group of people in the use of language in particular contexts and so forth. In general, the brief statistic analysis employed here is to help readers get a general picture of the frequency of particular phenomena, thus discovering a range of issues confronting people(Bouma2001: 174).

Second, this research aims to investigate how Chinese people behave in the addressing, opening and closing sections of personal letters, how linguistic variations are related to social variables (e.g. social roles, age) and the impact of the Chinese culture on the speech acts or events. That is why it is important to employ the in-depth interpretation of qualitative analysis, which is helpful for answering the above questions.

93 The key strength of the in-depth interpretation is that it reminds the researcher to look at the results qualitatively. There are many research focuses that this approach stimulates in this study, such as 'What is the form and function that a certain pattern/sequence reflects?', 'What are differentiations in functions between different types of strategy (e.g. single vs. combined)?', 'Why does a group of writers use a particular form/pattern/sequence more (or less) than an other?', 'How does the range of address forms, opening and closing patterns convey the writer's feelings and attitudes towards the recipient?', and 'How do the traditions and conventions affect the Chinese use of language in particular events and situations?' Describing and interpreting these allows the researcher to discover ritually and socially constructed realities of social interaction in relation to the addressing, opening and closing performance in the Chinese society, and also to reveal the culture-specific association of language and social factors in those speech acts and behaviours, and lastly to do justice to the complexity of such social phenomena embedded in the cultural context. Thus, using this kind of analytical method avoids simplifying social phenomena, and instead explores the range of linguistic behaviour and gives deeper understanding and clearer reflection of concepts, notions, and issues related to these social phenomena.

In summary, an efficient way to examine Chinese people's performance in certain speech acts and behaviours should be both descriptive and exploratory. The combination of the statistic measures and in-depth interpretation thus permits an attempt to focus on how the Chinese manipulate language. This is more than simply capturing overall tendencies in the incidence of linguistic features. These multiple methods provide useful baseline information, by which the pragmatic language use and the roles of various social variables are explicitly mapped out.

3.3 Concluding remarks This chapter has described the samplings, procedure of data collection and how the data were translated and analyzed based on a framework of categorization and identification developed from both previous studies and the researcher's introduction. This chapter has also sketched very briefly the analytical methods adopted for this study, namely, statistical measures and in-depth interpretation. Detail analysis and discussion of findings are given in the next three chapters.

94 CHAPTER FOUR ADDRESS FORMS IN CHINESE PERSONAL LETTERS

4.0 Introduction This chapter examines the address forms used in Chinese personal letters. In particular, it looks at the lexical and pragmatic features of each form and explores the relationship between the use of language and social factors, such as social roles, ages/generations and gender in the family and non-family settings. It also investigates the influence of the Chinese culture on this practice. The aim is to better understand how the social and pragmatic aspects of address forms in personal written interaction are practiced within the context ofbroad social norms and cultural values.

4.1. Address forms and functions in Chinese personal letters Three macro strategies of address forms have been identified from the data: (1) the name-alone strategy; (2) the compound name strategy; and (3) the no-name strategy. The name-alone strategy refers to the strategy in which there is only a name (e.g. full name or given name) without any modifier before or after. This is the most commonly used strategy found in the letters studied here (occurring in 46.7% of the letters). The compound name strategy is the strategy of modifying the name by adding a term of address either before or after the name. The modifying term might be an intimate term as in 'Qin 'ai de Meiling' (Dear Meiling), a deferential term as in 'Zhou jun' (Gentleman Zhou), a title as in 'Ma laos hi' (Teacher Ma), or a kin term as in 'Dongjing ge' (Dongjing elder brother). This is the second most commonly used strategy (occurring in 32.4% of the letters). The no-naming strategy refers to the addressing forms in which there is only the person's title, kinship term, or a combination of these together with a modifier, such as with 'Jingai de baba mama' (Respected dad and mum). This strategy is used less frequently than the other two (occurring in 20.8% ofthe letters).

As indicated in Table 4.1, each macro strategy can be divided into several substrategies. The name-alone strategy has six sub-strategies, the compound name strategy includes eight sub-strategies, and the no-name strategy contains seven sub-strategies. The following section will examine that which constitutes each individual strategy, that is, its linguistic forms and pragmatic functions.

95 Table 4.1. Categories and distribution of address forms in Chinese personal letters (N=259) Address forms N % a. The name-alone strategy 1. Full name 54 20.8 2. Given name 45 17.4 3. Milk name 18 6.9 4. Nickname 2 0.8 5. Family name 0.4 6. Foreign name I 0.4 Subtotal 121 46.7 b. The compound name strategy 7. Name+ kinship term 36 13.9 8. Name+ title 20 7.7 9. Intimate term+ name 13 5.0 I 0. Name + relation term 6 2.3 11. Diminutive term + name 5 1.9 I2. Intimate term+ kinship term+ 2 0.8 name 13. Name+ deferential term 0.4 14. Deferential term+ name +title I 0.4 Subtotal 84 32.4 c. The no-name strategy IS. Kinship term 32 12.4 16. Intimate term+ kinship term 11 4.2 I7. Deferential term + title 3 1.2 18. Deferential term+ kinship term 3 1.2 19. Title 2 0.8 20. Intimate term+ relation term 2 0.8 2I. kinship term +deferential term 1 0.4 Subtotal 54 20.8 Total 259 100

4.1.1 The name-alone strategy Among the six sub-categories of the name-alone strategy, the full name (FN) is the most commonly used, making up 20.8% of the total number cases. This strategy comprises a complete name, consisting of a family and a given name, such as 'Liu Xiangling. ' 'Liu' is the family name and 'Xiangling' is the given name. As with the Chinese oral address terms, in written Chinese the given name always follows the family name. A full name can be viewed as a familiar or casual address form. It is often used in non-family letters,

96 particularly by teachers writing to students or between friends. The use of a full name by teachers conveys a kind of familiarity and also age or rank seniority. Used by friends, it suggests familiarity and a level of in-group membership. Occasionally it can be adopted by the members of a family, particularly in letters from the older to the younger (e.g. father-son, elder sister-younger brother, aunt-nephew). However, this usage is often limited to those cases where a full name is made up by only two Chinese characters or syllables, such as 'Gao Yan' and 'Li Ji'. According to Chao, monosyllabic given names, such as 'Yan' or 'Ji' are not free, and thus must be bound and can be only linked together with a family name or with a following kinship term (1976: 315). Therefore, it is not common for Chinese people to address others with the monosyllabic given name, even within the family, unless a special intimacy is involved, as between husband and wife or amongst very close friends (Lee-Wong 2000: 171). For the older family members, a two-syllable full name seems to function like a given name.

The second most frequently used form in the name-alone strategy is the given name (GN), which is used in 17.4% of the total number of cases. In the data, most of the given names consist of two syllables like 'Ruifen, ' 'Yuanfei, ' and 'Hailin, ' and only rarely of one syllable like 'Man,' 'Yan,' and 'Ji.' The given name can be seen as an intimate address term. It is used often by parents to children, elder siblings to younger siblings, or among friends. The use of a given name by the older members of a family suggests in-groupness and a level of intimacy, while its use between friends denotes a high level of familiarity. Occasionally a given name also can be used by teachers addressing students. For example, when a teacher feels a close relation with a student he or she may address the student named 'Chen Hailin' as 'Hailin, ' signaling greater familiarity intermingled with seniority. It can be also suggested that such naming is an attempt to diminish the distance while retaining the authority.

The third sub-category is the milk name strategy. The milk name is a very intimate term of address, which is usually given by parents or an older member in the family for use during a child's early childhood, and sometimes after they become adults (Shih 1986: 104). A milk name is usually used among family members, particularly from parents to children, older to younger siblings or relatives, and also between close friends, indicating affection and intimacy. Five sub-types of the milk name are identified in this

97 study. The first is formed by duplicating the last character of the given name, such as 'Tingting' and 'Jingjing. ' The second is the strategy of addressing a child by referring to their physical characteristics or personality traits. For example, a boy may be called 'Xiao shitou' (a little piece of stone), probably because he is strong like a stone. A girl is named 'Ximao' (meaning thin hair), because she has little hair, or because she is the youngest child in the family, as 'xi' in some southwest regions of China (e.g. Hunan Province) represents 'young'. The third type is based on the name of the animal of the Chinese zodiac which was dominant in the year of the person's birth. Thus, a boy is named 'Xiao long' (little dragon), because the year he was born, such as 1952 or 1964, belongs to the year of 'long' (dragon). The fourth type is formed with numerals in the order of seniority among siblings or in separate series for the brothers and the sisters (Chao 1976: 315). This is often prefixed by the diminutive term 'xiao' (little) together with junior kinship terms (sometimes with the noun 'zi' following the kinship term), such as 'Xiao mei' or 'Xiao meizi' (little sister). Sometimes the prefix is simply a numeral such as 'san' (three). The fifth type is made up by the last given name prefixed or suffixed by a diminutive term. Examples of prefixed terms are the 'xiao' in 'Xiao Ling' (little Ling) and the 'a' in 'A-xiong.' Examples of the suffix are the 'zi' in 'Yingzi, ' 'Lingzi' and 'Meizi' and the 'er' in 'Xuer. ' The milk name is less frequently used than the given name, being used in only 6.9% of the total number ofletters.

Nicknames (NN) are the names which are given by others based on one's personality and physical appearance, or as "capsule histories" (De Klerk & Bosch 1999: 1). Examples of these are as 'Zidan' (Bullet), 'Pangzi' (fatty), and 'Suizuizi' (lippy) (Tian 1998: 300). Generally speaking, a nickname contains the playful element and often yields a sense of humor, irony or banter (Tian 1998: 300). It shows the strong involvement of physical, personal, historical and cultural factors in addressing forms. For example, in one of the letters studied a male is called 'Mao Qi' by his friend, probably because his behavioral attributes could be much like those of 'Mao Qi, ' an important figure in the Second World War. A young boy is addressed as 'Shamo shen deng' (A magic lamp in the desert), who was a fictional swordsman. These nicknames show some indications of "the ongoing enjoyment that human beings find in playing with language" (De Klerk & Bosch 1999: 1). The use of nicknames can be regarded as a means of creating some amusement or a light atmosphere (Shih 1986: 105). In the

98 letters studied here, nicknames are only used between friends, implying in-group membership and a higher degree of familiarity. This is a very infrequent address form, accounting for only 0.8% of the total.

Another alternative among names is the family name alone. This is a relatively new naming practice which often occurs among university students and friends (Shih 1986: 106). In Chinese, most family names are monosyllabic such as 'Wang,' 'Li' and 'Yang,' and there are only a few polysyllabic family names like 'Ouyang' and 'Situ.' Chao notes that, grammatically, a monosyllabic family name is not a free form, and to form an independent term it must be bound to something else, such as a given name, a deferential term or a title (1976: 314). Thus, the family name alone is not widely used to address people, except when the addresser and addressee are in a friendship context. Generally, when a family name is used alone, it implies a high level of intimacy. The use of family name alone is infrequent, accounting for only 0.4% of the total.

The final name-alone strategy involves a foreign name (FoN). Usually it is an English first name, such as 'Susan' or 'John.' It is another intimate address term which is often used among classmates or friends who have knowledge of English, such as English language majors. It may serve as an effective means "to signal one's aspiration to quicken the transition towards a relationship of solidarity and intimacy" (Li 1997: 505). In recent years the use of the western style first name as address form has grown with the popularity of learning English in China. However, the use of a foreign name as an address term, even in a situation where there is a great deal of contact with the English language, still seems to be uncommon. For example, as with the use of the family name, it occurs least commonly, and only makes up to 0.4% of the total.

4.1.2 The compound name strategy The compound name strategy is composed of eight sub-types. The most commonly used form is that of the name plus a kinship term, which accounts for 13.9% ofthe total. This pattern is often formed with a given name (monosyllabic or polysyllabic) or a milk name and a kinship term, usually one that refers either to juniors of a different generation, such as 'nuer' (daughter), or to both seniors and juniors of the same generation, such as 'ge' (elder brother) or 'mei' (younger sister). This pattern is often

99 used in family letters to address children, such as 'Chunli nuer' (Chunli daughter); to address both elder and younger siblings, such as 'Hong jie' (Hong elder sister), and 'Dongjing xong di' (Dongjing younger brother); or to address younger relatives, such as 'Tingting mei' (Tingting younger sister). The use ofthis pattern can also be extended to a circle of friends. For example, one can address his friend as 'Wenhui xong' (Wenhui elder brother). For family members, this pattern serves to indicate in-group membership, and also reflects a hierarchical kin relationship mixed with intimacy. For friends it stresses a kin-like friendship and intimacy.

The second most frequently used pattern in the strategy is a name plus a title, which accounts for 7.7% of the total. It is made up by a family or a full name accompanied by a title, such as 'Ma Wenjing laoshi' (Teacher Ma Wenjing) or 'Wang laoshi' (Teacher Wang). The occupational title is used here as a deferential term, and it signals the recipient's seniority in social rank. Generally, the family or the full name plus the occupational title is a more formal address form, and conveys the writer's high degree of respect for the recipient. This pattern is usually used by students writing to teachers.

An intimate term plus a name is the third most frequently used pattern within the compound name strategy, and accounts for 5% of the total. It includes an intimate term and a given name or a milk name. Here, intimate terms refer to some endearing adjectives or nouns like 'Qingai de' (dear), and 'Zui keai de' (most lovely). The combination of an intimate term and a given or a milk name can be seen as a very intimate address pattern. It is often used among siblings and friends, serving to indicate a great degree of intimacy and affection. It is worth noting that the choice of the intimate term depends on the relationship and the level of intimacy between the writer and recipient. For example, 'dear,' the intimate term most commonly encountered in the letters, carries the general sense of intimacy, and is often used among siblings and friends. The term 'Zui keai de, 'which is valued along with 'the most lovely,' is usually used between relatives of the same generation (e.g. cousin to cousin).

The next pattern is that of a name plus a relation term. The relation terms refer to those items reflecting a kind of social relationship between writer and recipient, and also the status of the recipient, such as 'wo you' (my friend), 'zhanyou' (comradeship-in-arms),

100 'laoxiang' (village fellow), or 'tongxue' (classmate). Basically, the choice of these terms is related to the relationship between writer and recipient. For instance, the first three are often used between friends, and they signal a particular level of group membership and friendship. However, the term 'tongxue' has two dimensions. When it is used by friends, it means 'classmate' of the same class, which means "a relationship institutionalized as egalitarian" (Blum 1997: 367). Used in this context, 'tongxue' emphasizes a relationship of mutual acquaintance. When it is used by a teacher writing to a student, however, it means 'student fellow', which suggests a teacher-student relationship. Used in this context, 'tongxue' emphasizes a higher level of formality in the relationship between the writer and the recipient. This form occurs infrequently in our data, making up only 2.3% of the total number of addressings.

The other pattern under this category is a diminutive term plus a name. It occurs infrequently in this study (in 1.9% of addressings). The diminutive term here refers to the lexical items 'lao' (old) and 'xiao' (little). In Chinese, the 'xiao' plus a family name is often used to address very familiar persons who are junior in age or rank. This usage indicates a greater familiarity and some degree of intimacy. As is evident from the data, when a student is addressed by the teacher by the term of 'Xiao Xing' (Little Xing), it not only shows that the teacher is quite familiar with the student, but it also implies that the teacher feels close to this person. The use of this form in the context of a teacher student relationship can be seen as an attempt to diminish the distance between correspondents while maintaining the age or rank hierarchy. Apart from being used by teachers, this pattern is also used between friends. In contrast, the pattern of 'lao' (old) plus a family name is often used to address elder familiars, indicating a certain degree of deference and as well as intimacy (Tian 1998: 297). However, no such case was found within the limited sample used in this study.

The most complex affectionate form found in the data is that of an intimate term plus a kinship term followed by a given or a milk name. Examples of this form are 'Baobei nuer tingling' (Precious daughter Tingting) and 'Keai de meimei Yingzi' (Lovely younger sister Yingzi). It appears that the use of this pattern is limited to letters to family members, as it expresses a special kind of closeness. For example, when a mother calls her daughter 'baobei nuer' (precious daughter), or an elder cousin

101 addresses her younger cousin 'keai de meimei' (lovely younger sister), they tend to convey a strong affection to the younger recipients, emphasizing the close kin relationship between the two. However, this pattern is very uncommon, making up only 0.8% ofthe total number ofaddressings.

Another sub-strategy is that of a name plus a deferential term. It is an archaic form of address, which consists of a family name and a deferential term, such as with 'Zhou jun ' (Gentleman Zhou). In the Chinese tradition, 'jun' was a typical deferential term with which to address a man. By using this term preceded by a family name, the writer tends to convey his or her respect for the recipient. The data shows that this pattern is used among friends. However, it occurs very rarely, and there is only a single case in the data (0.4% of the total number of addressings).

The last pattern within the category of the compound name strategy is that of a deferential term plus a name followed by a title, such as 'Zunjing de Wang laos hi' (Respected teacher Wang). This pattern is used by students writing to teachers, indicating a high respect to the recipient and reflecting the recipient's high status in both age and rank. Again, it occurs only once in the data (0.4% of the total number of addressings).

4.1.3 The no-name strategy The no-name strategy includes seven sub-types: a kinship term; an intimate term plus a kinship term; a deferential term plus a title; a deferential term plus a kinship term; an intimate term plus a relation term; a title; and a kinship term plus a deferential term. The most commonly used pattern in this category is that of the plain kinship term, and this is used in 12.4% of the total of number of addressings. This form is often used in letters written to one's parents, to elder relatives, and to elder siblings, but only rarely to younger siblings. It serves to emphasize the recipient's seniority and in-group membership, conveying the writer's intention of maintaining a solid, hierarchical and affectionate relationship. Interestingly, it appears that the plain kinship term can be also used between friends, although this occurrence is very rare (only one case). Such a usage indicates a high level of group membership and intimacy, reflecting a change in the character of the relationship from one of friendship to one of kinship. Sometimes,

102 the plain kinship term is preceded by numbers indicating the order of seniority among siblings, such as 'Da ge' (Eldest brother) or 'Si jie' (Fourth elder sister).

The second most commonly used sub-strategy under the no-name strategy is that of an intimate term plus a kinship term, which makes up 4.2% of total number of addressings. The intimate terms used in this pattern include 'guai ' (sweet) and 'qin 'ai de ' (dear). The former is adopted by parents (usually mothers), such as 'Guai erzi' (sweet son), while the latter is often used by younger siblings, such as 'Qin 'ai de jiejie' (Dear elder sister). It is clear that this pattern is family-oriented, and serves to express intimacy intermingled with relations of seniority.

Three other patterns found in this category are that of a deferential term plus a title, that of a deferential term plus a kin term, and that of a kin term plus a deferential term. They occur very infrequently in the data: The former two patterns make up 1.2 % while the latter only 0.4% of the total number of addressings. As can be seen from the data, these three patterns include three different words of a deferential kind, which differ in meaning and usage. For instance, the phrase 'zunjing de' (respected) suggests a sense of respect, and is often accompanied by a title, such as 'Zunjing de laos hi' (Respected teacher). It is used by students writing to teachers, serving to indicate a greater deference to the recipient, and emphasizing the recipient's seniority in rank.

The phrase 'jingai de' (beloved) suggests a sense of respect mixed with affection. It is accompanied by a kinship term which addresses seniors in the family, such as 'Jingai de baba mama' (Beloved dad and mum), or 'Jingai de yifu yimu' (Beloved uncle and aunt). The pattern serves to convey a greater degree of respect and affection towards recipients, emphasizing both hierarchical kin relationships and intimacy.

The third deferential term is 'daren' (great man). In traditional Chinese, 'daren' was a very common deferential term used to address senior family members (Chen 1991: 369). It was often preceded by a kinship term, such as in the phrase 'Fuqin daren' (Father great man). Combined with a kinship term, this address form serves to convey the writer's high respect for the recipient by emphasizing a hierarchical relationship. On the other hand, it implies a certain level of formality in comparison with other

103 deferential terms like 'jingai de' (beloved). It should be noted that the 'daren ', like 'jun' (gentleman), is used very rarely in modem personal letters. For example, there is only one case occurring in this sample (0.4% of addressings). It thus appears that the archaic deferential items are gradually disappearing in modem China.

The next no-name strategy is the plain title. For the present study, a title is narrowed down to the occupational title, such as 'laoshi' (teacher). Normally, the occupational title in this study is often prefixed by a full name or a family name, such as with 'Wang laos hi' (Teacher Wang) and 'Ma Shuwan laos hi' (Teacher Ma Shuwan). This phrase indicates a great respect for the recipient, and it emphasizes the recipient's seniority and distance from the writer. It can be seen as a more impersonal address form. However, a plain occupational title as an address form appears to be rather personal, serving to denote both respect and a great formality. This is because the title with a name is used not only by students but also by everybody who knows the addressee, whereas the plain title appears to be used only by the students who feel very familiar with their teachers. Hence, while the plain title may indicate hierarchy, it does not indicate distance. However, this usage is not common in the data, only two students (0.8% of addressings) use the plain title to address teachers.

The last sub-type of the no-name strategy is that of an intimate term plus a relation term. The intimate term in this pattern is the common term 'qinai de' (dear) and the relation term is 'tongzhuo' (classmate sitting at the same desk). 'Tongzhuo' is a relation term which has only recently emerged and which is prevalent among university students. It might be more personal than the term 'tongxue' (classmate), as the very word itself carries a sense of closeness. This pattern combines an emphasis on in-group membership with an emphasis on intimacy. However, it is also a very infrequent pattern in this sample, only two (0.8%) of the writers use it.

4.1.4 Discussion Based on the description of address forms that occur in the data, several suggestions can be made regarding the linguistic and pragmatic features of this behaviour. First, the address forms in Chinese personal letters are characterized by their diversity and elaborateness. The complexity in the addressing practices in written communication, as

104 it is in oral communication, reveals the importance of correct naming conduct in the interaction between Chinese people. This reflects "a view of the Chinese person as thoroughly embedded in a world of speakers and hearers," or of writers and recipients (Blum 1997: 358). Thus, addressing practices in written discourse, as in spoken language, are regarded as complex in China.

Second, as with oral communication, address forms in written communication are found to be a strong social index. They clearly announce the identities of the writer and the recipient in regard to age, status, gender, occupation, and the degree of intimacy or distance between them. Address forms also indicate the formality of the context. Furthermore, they constitute an important part of the Chinese language's politeness strategies, through which the writers indicate their respect and solidarity towards the recipients.

How have Chinese written address forms become such an important source of social information about the identities of writers and recipients? The reasons for this are related to the traditional Chinese mode of ethics. As is well known, China has had several thousand years of feudal history, and the ethics of Chinese traditional society were expressed in the phrase 'Zhongjun xiaoqin, sangang wuchang ', which translated as 'Loyalty to king; filial piety to kin; the three cardinal guides: Ruler guides subject; father guides son; husband guides wife; and the five constant virtues: benevolence, righteousness, propriety, wisdom; and fidelity' (Zhao 1997: 18). Influenced as they are by these traditional principles, people addressing each other obey the principle of 'Guijian youdeng, zhangyou youxu, nannu youbie' (There is a ranking between the noble and the humble; there is an order between senior and junior; there are differences between man and woman) (Wu 1990: 64). Therefore, the use of address forms in Chinese personal letters reflects a distinct characteristic of this hierarchical society, and displays many different kinds of identities ofthe writer and recipient.

The third feature of Chinese address forms is the use of kinship terms to address non kin. This usage cannot be taken literally, as in the sense of defining the recipients as actual family members. What is expressed through these forms is the writer's intention of establishing a relationship which is similar to that of 'family solidarity' (Brown &

105 Gilman 1960: 276). According to Tian (1998: 33-34), an important point about the use of kinship terms is that they are an indication of one's status, at least when they are used within the family. Kin status predetermines one's social life by stipulating the duties and rights that are part of one's social relationships. For example, when the kinship term ge (elder brother) is used to address a person within family, it indicates the recipient's seniority over the addresser (a younger sibling), while at the same time implying that the senior has a responsibility towards the junior (Wu 1990: 86). However, when kinship terms are used for non-kin they suggest a shift of the relationship, that is, a friend relationship, for example, is replaced by a kin relationship. In this case, according to Wu, the writers tend to establish polite and close interpersonal relationships between the recipients and themselves (1990: 86). This phenomenon reflects the influence within the Chinese subconscious of the principle of 'Tianxia yijia ' (One family under the sky) or 'Sihai zhinei jie xiongdi' (People of the world are all brothers) (Ma & Chang 1998: 47). On the other hand, using kinship terms for non-kin can also be seen as an effective communication measure. As Wu notes, by doing so, the writers tend to achieve a special communicative goal, as a Chinese saying says 'Tao jinhu' (to try to win someone's friendship) (1990: 87).

Another noticeable feature of address forms in personal letters is their complex structure. Findings show that some writers like to add before or after a name certain affectionate or deferential adjectives or nouns, such as 'Qingai de Meizi' (Dear Meizi). They might also like to add a kinship term, such as 'Fuqing daren ' (Father great man), or a title, such as 'Zunjing de Wang laoshi' (Respected Teacher Wang). In general, the addition of a modifier has a social meaning. For example, when a writer uses the term 'Zunjing de Wang laoshi' (Respected Teacher Wang) to address a teacher, he not only expresses his deference to the recipient, but also highlights the recipient's positive quality.

In Brown and Levinson's view of politeness, the use of additives to modify the address forms can also be seen as a way of expressing positive or negative politeness. For example, a combination of a given name with an intimate term like 'Qingai de Meiling' (Dear Meiling) strongly claims an in-group membership by emphasizing intimacy. The selection of a term with a positive orientation such as 'dear' may serve the purpose of indicating that the writer wants to "come closer" to the recipient (Brown & Levinson

106 1987: 103). Hence, such a strategically mixed address form should be considered more positive than the plain given name.

It is worth noting that some additives carry both positive and negative orientations of politeness, and the reflection of the orientations of politeness is related to the particular contexts in which the address forms occur. For example, when a writer combines a deferential term with a kinship term to address members of the family, such as 'Jingai de shushu shenshen' (Beloved uncle and aunt), the writer tends to express respect towards the recipients. This phrase emphasizes the recipients' seniority, and to some extent also expresses affection, stressing in-group membership and intimacy.

In summary, there are a variety of address forms that Chinese people use in personal letters. These forms are related to the writer and recipient's status in both age and social rank, and to the circumstances of their interpersonal relationship. Chinese people use kinship terms to address friends in letters in order to establish and maintain a polite and close relationship. Moreover, it is common for Chinese writers to modify a name, a kinship term or a title with an intimate or a deferential term in order to create or increase intimacy, deference or a mixture of respect and affection. The particular orientation of politeness expressed through the mixed address forms depends on the selection of a particular term and on the particular context.

4.2 Address forms and social variables This section examines how social variables influence Chinese in their choice of address forms in personal letters. First, we will look at the connection between the choice of address patterns and the social roles of writers and recipients.

4.2.1 Address forms and social roles As shown Table 4.2, when referring to children, parents overwhelmingly use the name alone strategy (65.4%), especially using a given name (42.3%), or a given name plus a kinship term like 'Li er' (Li son) (26.9%). The use of a given or milk name to address children indicates a kind of informality, emphasizing an in-group identity. It also implies the parents' seniority over children in both age and generation. The combination of a given or milk name with a kinship term appears to encode the parent's

107 indices of both seniority and affection, since these attached kinship terms not only indicate the children's junior status in the family, but they also stress the affectionate blood bond with parents. In this sense, the attachment of the junior kinship term can be seen as a way of increasing the degree of in-group identity and affection.

However, children do not use names to address their parents, but use plain kinship terms (44.4%) like 'Baba mama' (Dad and mum), or a kinship term preceded by an intimate term (38.9%), such as 'Qingai de baba mama' (Dear dad and mum). These two major address forms used by children serve respectively to convey respect or unite affection and respect towards their parents. Occasionally, children may use a deferential term plus a kinship term (11.1 %) and a kinship term followed by a deferential term (5.6%) to address their parents, denoting a great respect and a level of formality.

Siblings use a range of address forms in letters, including plain kinship terms (35.7%), such as 'Si jie' (Fourth elder sister) or 'Didi' (Younger brother), a name plus a kinship term (31.3%), such as 'Pen di' (Pen younger brother), or a given name (12.5%). Looking at the data closely, it is interesting to find that the use of the address forms among siblings is two-dimensional. The elder siblings prefer to use a name alone strategy (e.g. given name, milk name) (20.8%), while the younger siblings prefer to use a kinship term (31.3%). Sometimes both elder siblings and younger siblings use the combination of a name with a kinship term to address each other (26.7% and 29.1% respectively). The finding that the younger do not use a name to address the older in the sibling context again reflects the fact that addressings in the Chinese context are structured hierarchically even among siblings.

Relatives prefer to use a name strategy (43.8%), especially a given name (18.8%) and a milk name (18.8%), or a kinship term alone (37.5%). It is worth noting that within the group of relatives the choice of address forms varies with the differences in age and generation. As is shown in the data, it seems that the use of kinship terms, such as 'Gugu' (Aunt) or 'San sao' (The third elder sister-in-law), is preferred by juniors, whether their generation be the same as or different from that of the recipient. It should be noted that the samples of this phenomenon are limited to two nieces, one nephew, and two younger brothers-in-law. The use of a given or a milk name like 'Dongjing',

108 however, is prevalent among the seniors of different generations (four uncles, three aunts). Thus, it can be suggested that the addressing practices in the indirect blood relationships (e.g. uncle to niece, aunt to nephew) are also hierarchical. Moreover, it is interesting to note that indirect blood relatives can be addressed with reference to their junior status. For example, one elder cousin addresses his younger cousin by the kinship term 'Biaomei' (Younger cousin). The use of such an address form can be regarded as a way of showing a close relationship between the relatives of the same generation. However, this kind of usage is rare and only one case is found in the data.

In the case of non-family letters, the findings show that the name alone strategy, especially a full name (40%), is appropriate for friends. This is probably because confirmation of in-group identity and familiarity appears to be crucial for friends in the Chinese culture. Apart from the common use of a name, friends also tend to address their friends with a name plus a kinship term (11.3%), such as with 'Yuchen xong' (Yuchen elder brother). From a friend's point of view, using a name followed by a kinship term can bring them closer to their friends, since to some extent the equal friendship seems to be replaced by the filial piety expressed through the kinship term. Thus, the relationship between the friends is strengthened.

109 Table 4.2 Frequency and distribution of address forms across social roles (N=259) Address Forms Social roles Total Family letters: I 08 Non-family letters: !51 Parent Child Sibling Relative Friend Teacher Student (26) (18) (48) (16) (115) (10) (26) (259) a. The name-alone strategy 2 0 I I 46 4 0 54 I. Full name (7.7) (0) (2.1) (6.3) (40) (40) (0) (20.8) 2. Given name II 0 6 3 22 3 0 45 (42.3) (0) (12.5) (18.8) (19.1) (30) (0) (17.4) 3. Milk name 4 0 3 3 8 0 0 18 (15.4) (0) (6.3) (18.8) (6.9) (0) (0) (6.9) 4. Nickname 0 0 0 0 2 0 0 2 (0) (0) (0) (0) (1.7) (0) (_0) (0.8) 5. Family name 0 0 0 0 I 0 0 I (0) (0) (0) (0) (0.9) (00 (0) (0.4) 6. Foreign name 0 0 0 0 I 0 0 I (0) (0) (0) (0) (0.9) (0) (0) (0.4)

Subtotal 17 0 10 7 80 7 0 121 (65.4) (0) (20.8) (43.8) (69.6) (70) (0) (46.7) b. The compound name strategy 7. Name + kinship 7 0 16 0 13 0 0 36 term (26.9) (0) (33.3) (0) (11.3) (0) (0) (13.9) 8. Name+ title 0 0 0 0 0 0 20 20 (0) (0) (0) (0) (0) (0) (76.9) (2.2) 9. Intimate term+ 0 0 3 0 10 0 0 13 name (0) (0) (6.3) (0) (7.8) (0) (0) (5.0) I 0. Name +relation 0 0 0 0 4 2 0 6 term (0) (0) (0) (0) (3.5) (20) (0) (2.3) II. Diminutive term 0 0 0 0 4 I 0 5 +name (0) (0) (0) (0) (3.5) (10) (0) (1.9) 12. Intimate term+ I 0 0 I 0 0 0 2 kinship term + (3.8) (0) (0) (6.3) (0) (0) (0) (0.8) name 13.Name+ 0 0 0 0 I 0 0 I deferential term (0) (0) (0) (0) (0.9) (0) (0) (0.4) 14. Deferential term 0 0 0 0 0 0 I I + name + title (0) (0) (0) (0) (0) (0) (3.8) (0.4)

Subtotal 8 0 19 1 32 3 21 84 (30.7) (0) (39.6) (6.3) (27.8) (3()} (80.8} (32.4) c. The no-name strategy 0 8 17 6 I 0 0 32 IS.Kinship term (0) (44.4) (35.4) (37 .5) (0.9) (0) (0) (12.4) 16.1ntimate term+ I 7 2 I 0 0 0 II kinship term (3.8) (38.9) (4.2) (6.3) (0) (0) (0) (4.2) 17. Deferential term 0 0 0 0 0 0 3 3 +title (0) (0) (0) (0) (0) (0) (11.5) (1.2) 18. Deferential term 0 2 0 I 0 0 0 3 + kinship term (0) (11.1) (0) (6.3)) (0) (0) (0) (1.2) 19. Title 0 0 0 0 0 0 2 2 (0) (0) (0) (0) (0) (0) (7.7) (0.8) 20. Intimate term+ 0 0 0 0 2 0 0 2 relation term (0) (0) (0) (0) (1.7) (0) (0) (0.8) 21. Kinship term+ 0 I 0 0 0 0 0 I deferential term (0) (5.6) (0) (0) (0) (0) (0) (0.4)

Subtotal 1 18 19 8 3 0 5 54 (3.8) (100) (39.6) (50) (2.6) (0) (19.2) (20.8) Total 26 18 48 16 115 10 26 259 (100) (100) (100)_ (100) (100) (100) (100) (100)

110 Teachers overwhelmingly prefer to use a name alone strategy (70%) to address their students. Addressing students by a name indicates a level of familiarity, while at the same time implying the teachers' higher status and a degree of social distance. It is interesting to notice that occasionally teachers may use the pattern of a name plus a relation term (20% ), such as with 'Hailin tongxue' (Hailin student fellow), or a diminutive term plus a family name (10%), such as with 'Xiao Xing' (Little Xing). Using these elaborate terms, the teachers tend to minimize their distance from the students, signaling a great familiarity and a certain degree of closeness.

As with children writing to parents, students writing to teachers address their recipients with a name (family name, full name) plus a title (76.9%), such as 'Ma laoshi' (Teacher Ma) or 'Zhang Min laoshi' (Teacher Zhang Min). By referring to the recipient's higher social status, such forms indicate both the students' respect for the teachers and a certain level of distance. Apart from using these forms, students may occasionally choose a very deferential term to address their teachers, such as a deferential term plus a title (11.5%), as with 'Zunjing de laoshi' (Respected teacher), or a deferential term plus a name then followed by a title (3.8%), as with 'Zunjing de Wang laoshi' (Respect Teacher Wang). This finding shows that there is a tendency for students to use a more neutral deferential address pattern towards teachers, serving as a way to avoid creating a great distance between them.

4.2.2 Discussion Based on the findings, two models can be constructed to summarize the structural rules that underlie the use of address forms by writers of various social roles. Figures 1 and 2 show the networks of the writer-recipient relationships that give rise to different address forms in the family and non-family letters, respectively. It is necessary to say this suggested network is classificatory, not a model covering all possible choices occurring in the data, since some less commonly used forms are omitted. For example, in the context of child-parent relationship, a kinship term plus a deferential term, such as 'Fuqing daren' (Father great man), is left out of the figure.

111 Given name I I I Milk name Parent Child Name+ I I I kinship term

I I I I Kinship term Child Parent Intimate term I I I +kinship term

Given name Milk name Elder I Younger Name+ ~ kinship term

I Siblings I r- Kinship term Younger Elder Name+ I I kinship term

Given name Niece, Milk name Aunt, uncle nephew Different rl r r- generation Kinship term Niece, Intimate term nephew Aunt, uncle '-- ~ ~ +kinship term

Relatives Kinship term r Elder cousin Younger Intimate term cousin r- 1- 1- +kinship term Same .__ generation

Younger Elder cousin Intimate term '-- cousin 1- 1- +kinship term

Figure 4.1. Summary diagram of social roles that constrain the choice of address forms in family letters.

112 Full name I Friend I I Friend I Given name I I I I Milk: name Name+ kinship term Intimate term + name

Full name Name +relationship Teacher Student 11 I I I term

Name+ title Deferential term + I Student I I Teacher I title I I I I Title

Figure 4. 2. Summary diagram of social roles that constrain the choice of address forms in non-family letter

As is seen in Figures 4.1 and 4.2, one feature of the address system in Chinese personal letters is its sensitivity to the social role relationship between writer and recipient. Whenever roles are unequally distributed, the use of address forms is asymmetrical, even within the family letters. Parents receive terms of address stressing respect and a mixture of respect with affection from their children and give back names denoting informality and in-groupness. Elder siblings are often treated with terms showing respect or uniting respect and affection, and give back terms expressing informality and solidarity. Older relatives of a different generation receive respect and respect mixed with in-groupness from younger relatives, and give back terms of address indicating informality. Teachers receive address forms emphasizing respect or mixing respect with formality from their students, and give back terms signalling familiarity and distance. On the other hand, when the roles are socially equal, the use of address forms is symmetrical in general. Friends receive reciprocal address forms from each other, emphasizing familiarity and solidarity.

113 Differentiations in the use of address forms across social roles in both family and non family letters reflect the influence of the Confucian principle that people should treat each other according to the distinctions set up in the hierarchy of the society (Shih 1986: 30). According to Confucius, there are five human relationships in the traditional Chinese society, which are known as 'wun tun. ' These are the relationships between the king and the minister, between father and son, between husband and wife, between the elder and the younger brothers, and between friends (Bond & Hwang 1986: 215). Based on these five relationships, good people should exhibit loyalty to the ruler; filial obedience to their parents; deference to elder brothers; kindness to younger brothers; and cordiality to friends (Ebrey 1991: 70). These requirements are guided by rules of correct behaviours, or li, which means "to follow what is appropriate" (Shih 1986: 56). According to Confucius, once people behave by following the requirements of their roles in the hierarchy, harmonious human relations will be achieved. These features of the Confucian ethics have continued to be the basis of the Chinese address system in modem times. Therefore, for example, it is not a surprise that there is a very strong avoidance of the use of names by children, by younger relatives of a different generation, and by students. The choice of address pattern differs even between elder and younger siblings.

The second feature is the influence of asymmetrical roles integrating affection and closeness. The findings indicate that in the asymmetrical role relationships, especially within the family, the writers who hold a higher status (e.g. parents, elder siblings) sometimes address the recipients of lower status (e.g. children, younger brother) with terms uniting seniority and affection, such as 'Meilin nuer' (Meilin daughter) or 'Pen di' (Pen young brother). By combining seniority and affection in their addressing practices, these writers tend to convey both informality and intimacy. This phenomenon can be seen as a typical characteristic of written Chinese family discourse, which is similar in some respects to some features of the family speech style in other cultures (Blurn-Kulka 1990).

The role relationships within the family hierarchy are of particular importance in the Chinese context. The asymmetry of these relationships results in the asymmetry of the use of different address forms. However, as in some other cultures, Chinese families

114 are also the place of affection, every individual in a family plays both the roles of an intimate and of unequal member. Such values and practices within the family may affect people's speech and actions. Thus, parents or elder siblings address their children or younger siblings using terms that indicate both seniority and affection, emphasizing both authority and intimacy. From the point of view of politeness, the involvement of affection in a status-oriented term may increase politeness (Blum-Kulka 1990). This finding is interesting, as it confirms Blum-Kulka's claim that "relationship affection" should be viewed as another social parameter which influences people's expressions of politeness (1990: 284).

It is worth noting that the use of forms denoting both asymmetry and closeness can be extended to the non-family circle (e.g. from teacher to student), although the incidence of this is very rare. For example, the data shows that teachers sometimes (in 20% of cases) address their students with terms indicating both familiarity and a level of closeness, such as 'Hailin tongxue' (Hailin student fellow), or 'Xiao Xing' (Little Xing). This finding reflects the fact that even in the non-family setting the asymmetry of unequal roles does not seem to prevent the formation of close attachments. This is because, to use Blum's words, in Chinese society "considerations of closeness sometimes override those of distance: closeness can be part of distance if it is generational or conventional" (1997: 371). On the other hand, the use of these less distant and more status-oriented terms also reflects the writer's aim of achieving politeness, as the terms used to express closeness such as 'tongxue' (student fellow) or 'xiao' (little) can be seen as in-group identity markers. They serve to claim common ground and in-group membership (Brown & Levinson 1987: 103).

In sum, in the Chinese context social roles strictly constrain the use of address forms in both family and non-family letters. In general, asymmetrical social relationships result in address structures being constructed asymmetrically, while the equal social relationships (i.e. friendship) result in address structures being constructed reciprocally. However, despite the asymmetries of address forms, indices of both seniority and intimacy can be embodied in them. This is especially so in family letters, and even in non-family letters in which considerations of closeness may override concerns of distance.

115 4.2.3 Address forms and age/generation This section examines the influence of age/generation on the use of address forms in Chinese personal letters. It aims both to investigate how age/ generation differences influence Chinese written addressing practices and to investigate the extent to which Chinese cultural values and norms act as a constraint on the pragma-linguistic repertories ofwriters.

Based on findings summarized in Table 4.3, it appears that in family letters members of the older generation (e.g. parent, uncle) often use a name alone strategy (72.7%), especially a given name (42.4%), while members of the younger generation (e.g. child, nephew) often use a kinship term (56.5%), or an intimate term plus a kinship term (30.4%). Older members of the same generation (e.g. elder brother) often use a name (40%), or a name plus a kinship term (36%), while younger members of the same generation (e.g. younger sister) often use a plain kinship term (62.9%) or a name plus a kinship term (25.9%).

In the case of non-family letters, persons of an older generation to persons of a younger generation (e.g. from a teacher to a student) often use a name (in 70% of cases). A full name is used in 40% and a given name in 30% of cases. Persons of a younger generation to persons of an older generation often use a name plus an occupational title (76.9%). Older persons of the same generation often use a full name (27.9%), or a name plus a kinship term (21.3%), while younger persons of the same generation often use a given name (33.3%) or a full name (25%). A majority ofletters written to persons ofthe same age use a full name (61.9%).

116 Table 4.3. Frequency and distribution of address forms across age/generation

(N=259) Address forms Age/generation Total Family letters Non-family letters DG:5.6 SG:52 DG:36 SG:ll5 0-Y Y-0 0-Y Y-0 0-Y Y-0 0-Y Y-0 SA (33) (23) (25) (27) (10) (26) (61) (12) (42) (259) a. The name-alone strategy I. Full name 3 0 I 0 4 0 17 3 26 54 (9.1) (0) (4.0) (0) (40) (0) (27.9) (25) (61.9) (20.8) 2. Given name 14 0 5 I 3 0 12 4 6 45 (42.4) (0) (20) (3.7) (30) (0) (19.7) (33.3) (14.3) (17.4) 3. Milk name 7 0 4 0 0 0 4 0 3 18 (21.2) (0) (I 6) (0) (0) (0) (6.6) (0) (7. I) (6.9) 4. Nickname 0 0 0 0 0 0 I 1 0 2 (0) (0) (0) (0) (0) (0) (1.6) (8.3) (0) (0.8) 5. Family name 0 0 0 0 0 0 0 I 0 I (0) (0) (0) (0) (0) (0) (0) (8.3) (0) (0.4) 6. Foreign name 0 0 0 0 0 0 I 0 0 I (0) (0) (0) (0) (0) (0) (I .6) (0) (0) (0.4) Subtota 24 0 10 1 7 0 35 9 35 121 (72.7) (0) (40) (3.7) (70) (0) (57.4) (75) (83.3) (46.7) b. The compound name strategy 7. Name+ kinship 7 0 9 7 0 0 13 0 0 36 term (21.2) (0) (36) (25.9) (0) (0) (21.3) (0) (0) (13.9) 8. Name+ title 0 0 0 0 0 20 0 0 0 20 (0) (0) (0) (0) (0) (76.9) (0) (0) (0) (7.7) 9. Intimate term+ 0 0 3 0 0 0 6 I 3 13 name (0) (0) (12) (0) (0) (0) (9.8) (8.3) (7. I) (5.0) I 0. Name +relation 0 0 0 0 2 0 2 0 2 6 term (0) (0) (0) (0) (20) (0) (3.3) (0) (4.8) (2.3) I I. Diminutive term 0 0 0 0 I 0 3 0 I 5 +name (0) (0) (0) (0) (10) (0) (4.9) (0) (2.4) (1.9) 12. Intimate term+ I 0 I 0 0 0 0 0 0 2 kinship term + (3.0) (0) (4.0) (0) (0) (0) (0) (0) (0) (0.8) name 13. Name+ 0 0 0 0 0 0 0 I 0 I deferential term (0) (0) (0) (0) (0) (0) (0) (8.3) (0) (0.4) 14. Deferential term+ 0 0 0 0 0 I 0 0 0 I name+ title (0) (0) (0) (0) (0) (3.8) (0) (0) (0) 0.4 Subtotal 8 0 13 7 3 21 24 2 6 84 (24.2) (0) (52) (25.9) (30) (80.8) (39.3) (16.7) (14.3) (32.4) c. The no-name strategy 0 13 I 17 0 0 I 0 0 32 15. Kinship term (0) (56.5) (4.0) (62.9) (0) (0) (I .6) (0) (0) (12.4) 16. Intimate term+ I 7 I 2 0 0 0 0 0 II kinship term (3.0) (30.4) (4.0) (7.4) (0) (0) (0) (0) (0) (4.2) I 7. Deferential term+ 0 0 0 0 0 3 0 0 0 3 title (0) _(0) (0) (0) (0) (11.5) (0) {Q2_ (0) (1.2) 18. Deferential term+ 0 3 0 0 0 0 0 0 0 3 kinship term (0) (13.0) (0) (0) (0) (0) (0) (0) (0) (1.2) 19. Title 0 0 0 0 0 2 0 0 0 2 (0) (0) (0) (0) (0) (7.7) (0) (0) (0) (0.8) 20. Intimate term+ 0 0 0 0 0 0 0 I I 2 relation term (0) (0) (0) (0) (0) (0) (0) (8.3) (2.4) (0.8) 21. Kinship term + 0 0 0 0 0 0 0 0 0 0 deferential term (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) Subtotal 1 23 2 19 0 5 1 1 1 1 (3.0) (100) (8.0) (70.4) (0) (19.2) (1.6) (8.3) (2.4) (0.4) Total 33 23 25 27 10 26 61 12 42 259 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: different generatwn; SG: same generatiOn; 0-Y: older to younger; Y-0: younger to older; SA: same age

117 4.2.4 Discussion Based on the above findings, a number of points can be summarized and discussed. First, findings show that age and generation are determining variables in the choice of address forms in Chinese personal letters. Within family letters, members of older generations receive respect and respect mixed with affection, and give back informality or informality with intimacy. Older members of the same generation receive respect or respect mixed with closeness, and give back informality or informality mixed with closeness. In non-family letters, persons of older generations receive deference or deference mixed with distance, and give back familiarity and distance. In the case of letters between persons of the same generation, although there are no fundamental differences in the use of address forms, elder and same-aged writers generally prefer to use terms indicating familiarity, while younger writers prefer to use terms emphasizing informality and a level of intimacy.

The influence of age and generation on the choice of address forms in letters is analogous to the way in which age and generation influences verbal formulations in the Chinese culture. As stated by Shih, age in the Chinese context is a symbol of knowledge, experience, power, and status (1986: 36). There is often a correlation between seniority and social rank. Age is thus an important determinant in the choice of address forms in Chinese. The older the recipient, the greater the deference received. Even for people of the same generation there is still a distinction of seniority based on age differences. As is apparent from the data, elder brothers are often addressed with deferential kinship terms, while younger brothers are often addressed by their given names, which suggest informality and in-group identity.

Similarly, distinctions of seniority are also found in non-family letters, particularly in those written to friends, although the difference is not as significant as it is within the family circle. For example, the data shows that elder and same-age friends prefer to use a full name (27.9%) to indicate familiarity, while younger friends prefer to use a given name (33.3%) to emphasize intimacy. These phenomena can be explained by the fact that in the Confucian tradition a man would be subject to his sovereign, as a son is to his father, a younger brother to his elder brother, and a younger friend to an elder friend (Bond & Hwang 1986: 216). Then why are younger friends inclined to use a given

118 name to address elder friends instead of some deferential pattern, as younger brothers do in the family letters? Answering this question, one might hypothesize that the subordination of a younger brother is deference-oriented, while subordination from a younger friend is intimacy-oriented. From the younger friend's view, overuse of deferential address patterns would produce a distance, and a crucial concern is with maintaining friendship rather than age status.

In determining the orientation of the use of address forms the impact of age is more significant between persons of different generations than between persons of the same generation. Findings indicate that in the family setting the address system for members of different generations is completely asymmetrical, while the addressing practices for members of the same generation is characterized by both asymmetry and symmetry. In the non-family setting, the use of address forms is totally asymmetrical for persons of different generations. For persons of the same generation, however, address forms are usually symmetrical, although there are some differences between writers of different ages in their preferences for particular terms.

The different orientations of address forms used by writers of different generations and of the same generation (asymmetry vs. both asymmetry and symmetry) could be interpreted as a reflection of the differentiations of authority in Confucian social philosophy. According to Confucius, authority is created through the hierarchical structuring of relationships between father-son, elder brother-younger brother; teacher student; and elder friend-younger friend (Ebrey 1991: 46). Among them, the father-son relationship is a strict and unambiguous source of authority. This relationship strongly stipulates the behavioural inclinations of both sides. Thus, a father does not address his son by a kinship term, but by a given name, while a son never refers to his father by name. However, the elder-younger brother relationship does not appear as strictly hierarchical as the father-son relationship, since there can exist a feeling of true friendship from the younger to the elder, and a warm-hearted attitude from the elder to the younger (Ebrey 1991: 70). As a result, when addressing each other, it is appropriate for elder and younger brothers to use hierarchical structures, and it is also suitable for both of them to use terms emphasizing informality and affection.

119 In the non-family circle, the teacher-student relationship contains a stronger authority than the relationship between friends, as it is more hierarchical. There is a Chinese saying: 'Yiri weishi, zhongsheng weifu' (So long as someone has been your teacher even for one day, you should respect him as your father for a lifetime). Therefore, it is natural for teachers to receive terms indicating deference and to give back terms indicating familiarity. In contrast, the authority contained in friend-friend relationships is much weaker, although something of a hierarchical nature still exists because of age differences. For friends, solidarity and friendship are more important than authority. Therefore, a writer can use hierarchical structures to address an older friend, but he or she is more likely to use the reciprocal forms, indicating familiarity and solidarity.

In summary, age is an important social factor influencing the choice of address forms in both family and non-family Chinese letters. In addition, age has a different impact on address forms in letters between persons of different generations and the same generation. The use of address forms is asymmetrical between persons of different generations, while it may be both symmetrical and asymmetrical between persons of the same generation. For writers of the same generation, asymmetries outweigh symmetries in family letters, while symmetries are stressed more than asymmetries in non-family letters.

4.2.5 Address forms and gender This section analyzes the effect of gender, another social factor influencing the use of address forms in Chinese personal letters. In particular, it looks at how women use address forms differently from men, and how the Chinese cultural norms affect the sex role differentiations of this behaviour. As is shown in Table 4.4, in general the impact of gender on the addressing practice is not as striking as that of the other two social factors (role, age/generation). The differences between the address forms chosen by males and females are not very significant, although they still exist.

120 Table 4.4. Frequency and distribution of address forms across gender (N=259) Address forms Gender Total

Family letters Non-family letters F:53 M:55 F:71 M:80 259 N % N % N % N % N % a. The name-alone strategy 2 3.8 2 3.6 26 36.6 24 30 54 20.8 I. Full name 2. Given name 5 9.4 15 27.3 9 12.7 16 20 45 17.4 3. Milk name 6 11.3 4 7.3 2 2.8 6 7.5 18 6.9 4. Nickname 0 0 0 0 I 1.4 I 1.3 2 0.8 5. Family name 0 0 0 0 I 1.4 0 0 I 0.4 6. Foreign name 0 0 0 0 0 0 I 1.3 I 0.4 Subtotal 13 24.5 21 38.2 39 54.9 48 60 121 46.7 b. The compound name strategy 7. Name + kinship 12 22.6 II 20 0 0 13 16.3 36 13.9 term 8. Name+ title 0 0 0 0 13 18.3 7 8.8 20 7.7 9. Intimate term+ 3 5.7 0 0 10 14.1 0 0 13 5.0 name I 0. Name +relation 0 0 0 0 0 0 6 7.5 6 2.3 term II. Diminutive term +name 0 0 0 0 3 4.2 2 2.5 5 1.9 12. Intimate term+ kinship term + 2 3.8 0 0 0 0 0 0 2 0.8 name 13.Name+ deferential term 0 0 0 0 0 0 I 1.3 I 0.4 14. Deferential term + 0 0 0 0 I 1.4 0 0 1 0.4 name+ title Subtotal 17 32.1 11 20 27 38 29 36.1 84 32.4 c. The no-name strategy 16 30.2 15 27.3 0 0 I 1.3 32 12.4 15. Kinship term 16. Intimate term+ kinship term 6 11.3 5 9.1 0 0 0 0 II 4.2 17. Deferential term+ title 0 0 0 0 2 2.8 I 1.3 3 1.2 18. Deferential term+ kinship term 0 0 3 5.5 0 0 0 0 3 1.2 19. Title 0 0 0 0 I 1.4 I 1.4 2 0.8 20. Intimate term + relation term 0 0 0 0 2 2.8 0 0 2 0.8 21. Kinship term + deferential term I 1.9 0 0 0 0 0 0 I 0.4 Subtotal 23 43.4 23 41.8 5 7.0 3 3.7 54 20.8 Total 53 100 55 100 71 100 80 100 259 100

Findings indicate that in family letters males use a name alone strategy more than females do (38.2% vs. 24.5%), especially in the use of a given name (27.3% vs. 9.4%). Females use the compound name strategy more than males do (32.1% vs. 20%). Specifically, females prefer to use the intimate term plus a given or milk name (5.7%), such as 'Qingai de meizi' (Dear Meizi), or the intimate term plus a kinship term then followed by a milk name (3.8%), such as 'Keai de xiaomei Tingting' (Lovely little sister Tingting). There was no such case found among the letters written by males. Gender

121 differentiation in the use of the no-name strategy is very insignificant (43.2% vs. 41.8%).

Further examination suggests that sex-role differentiation in the choice of address forms is related to the age variable. For example, the preference for a given name by males is found among older writers (father 10; uncle 2; and elder brother 3), while the choice of the intimate address pattern by females is also found among older writers (mother 2; elder sister 2; and elder cousin 1). It is also noted that the intimate address pattern is only used from females writing to females (mother-daughter; elder-younger sister; elder-younger cousin). The exception is the given name, which can be used by males writing to both females and males (father-son; uncle-niece).

For non-family letters, the overall difference between females and males in the choice of the name-alone, the compound name and the no-name strategies are even less significant (54.9% vs. 60%; 38% vs. 36.1%; 7.0% vs. 3.7%). Further inspection indicates that there are some differences in the choice of sub-strategies within the compound name strategy. For example, there are some cases where females use an intimate term plus a given name (14.1%), but no such case is found among the letters written by males. There are cases where males use a name (i.e. given name) plus a kinship term, such as 'Yuchen di' (Yuchen younger brother) (16.3%). No such case is found among the letters written by females. It is worth noting these gender differences only occur in friend-friend letters. Another difference is in the choice of the pattern used for connecting the name and the title. It is evident that more females than males use this pattern (18.3% vs. 8.8%). However, this result may be misleading, because the percentages are made against the total numbers of females and males in the non-family letters. If the percentages were calculated against the total numbers of each gender in the user's group, the result would be quite different: No significant difference appears (female students: 13 out of 17 addressings = 76.5%; male students: 7 out of9 = 77.8%).

It is also worth noting that in non-family letters, particularly in friend-friend letters, the intimate pattern of address is often used by older females (6 out of 10 addressings)

122 writing to other females, while the pattern of a name plus a kinship term is used by older male friends (13 out of 13) writing to other male friends.

4.2.6 Discussion Based on the findings, three interesting features can be observed. First, females and males have different preferences in their choice of some address forms, although there are also some similarities in their addressing practices. In general, females tend to use intimacy-oriented terms, while males tend to use status-oriented terms. These gender differences can be explained by the different focus and the different purposes of male and female letter-writers.

Females' preference for the use of intimate-oriented terms reflects their tendency of giving great attention to verbal intimacy in their social interactions (Schaef 1985: 121). Through the use of the intimate address form, females tend to convey their affection, and their desire to maintain a close kin relationship or a good friendship with the recipient. For example, when a mother addresses her daughter with the terms 'Baobei nuer Tingling' (Precious daughter Tingting), it is clear that she tends to emphasize an affectionate bond between her and her daughter, conveying a very firm kin relationship through the affectionate term 'baobei' (precious). When a girl addresses her female friend with the terms 'Qin 'aide Jingrong' (Dear Jingrong), it is evident that by using the endearment term 'qin 'ai de' (dear) she wants to show her like of and affection for the recipient, indicating a close friendship tie.

Males' preference for a given name or a name plus a kinship term reflects their focus on indicating familiarity, in-groupness and hierarchical relationships in the addressing practices. For example, by addressing his nephew with a given name 'Mingfeng,' a male tends to indicate in-group membership and informality, and also to specify a kind of hierarchical relationship (uncle-nephew), thus implying his seniority over the recipient. When addressing his male friend with the terms of 'Dongjing xiongdi' (Dongjing younger brother), this man intends by the use of the given name 'Dongjing' to indicate in-group membership and informality. By using the kinship term 'xiongdi' (younger brother) to refer to his junior friend, the writer also implies a level of seniority

123 over the recipient. This finding supports the view that men often choose terms encoding a role relationship which maintains their own status (Lee-Wong 2000: 230).

These differences between female and male preferences for specific address forms provide a good example of the gender orientation of linguistic choices in general. Males tend to stress the social role-relationship and their own social status by using informal and status-oriented address forms, while females tend to emphasize an emotional bond with their recipients by using an intimate and affectionate pattern of address. These results concur with the view suggested by Zimin that "the sexes have different speech styles or perhaps different ways of dealing with the world" (1981: 38). For males it might be the world of reality, and for females, that of feelings and emotions.

It is also worth noting that the males' focus on social role relationships and status in their addressing practices differs in degree between letters within the family and the non-family circles. Within the family, males often use the given name to address the recipient, a practice which encodes their own authority. Within the non-family circle, males often use a name plus a junior kinship term to address the recipient, a practice which encodes an element of age seniority and also a level of intimacy. It can be suggested that the address forms of males' letters to their families appear to be more hierarchically-oriented than those of their letters to non-family. The possible reason for this may be that when writing to the family a man tends to pay more attention to the hierarchy of age, while in the friend-friend context a man is more concerned about in groupness.

The second noticeable feature is that age plays a role in the formation of respective preferences in the choice of address form for both females and males. The data shows that those address forms which reveal gender differences often occur in the letters written by the older to the younger. Why do older females like to use the intimate address pattern? Why do older males like to use the terms indicating informality and authority? In the case of the former, a possible explanation is that older females may feel more obliged to show intimacy to younger persons. Such a tendency results from the nature of females and their roles in a family or a social group. In Chinese culture, an older female, especially one from a nuclear family, tends to view the display of

124 affection and intimacy as part of her responsibility for caring for and educating the younger. For example, a mother, who is assumed to be a symbol of tenderheartedness, devotes her whole life to the cause of caring for and educating her children. The affectionate ties between this mother and her children are very firm, and thus she tends to convey her affective feelings to her children through using the intimate address patterns. This practice enables the mother to effectively communicate her love and care for young. In the case of an elder sister, the preference for the intimate address patterns can be seen as a reflection of her adopting maternal attitudes and behaviours towards her younger siblings. In a similar way to her mother, she tends to view the showing of affection to the young as her duty as a female, and as an important way of maintaining close ties with her younger siblings.

As for older female friends, the motivation for choosing the intimate pattern seems quite similar to that for elder sisters. A closer examination of the data indicates that when an older female addresses her younger friend by a certain intimate term, she usually has a very close friendship with the recipient. In the Chinese culture close friends are often treated as family members. Given this situation, it is no surprise that the older females tend to show affection to young females by using an intimate address pattern, in the same way that elder sisters often address their siblings.

Simpler reasons would seem to account for the preference by older males for terms of informality and authority (a given name for family and a name plus a kinship term for non-family). The older males' preference for addressing family by the given name reflects a perceived need to emphasize both in-group membership and their own authority. This result suggests that older males writing to the family could well use the address as a way of presenting their age, and hence power. In comparison, the older males' preference for a name plus a kinship term in their letters to non-family, particularly in the friendship context, is related to their perception that it is necessary to indicate to younger recipients both intimacy and hierarchical age differences.

The third observed feature is the influence of the gender of the recipient upon the choice of address forms. As discussed earlier, the gender differences are mainly demonstrated by the differences between the three address strategies: the intimate address pattern; the

125 name plus a kinship term pattern; and the given name pattern. Of the three, the given name pattern can be used by males writing both to persons of the same sex and to persons of the opposite sex. However, the intimate address pattern is mainly used by j females writing to females. This phenomenon can be interpreted by reference to the J fact that intimate terms, such as the term 'precious' or 'lovely,' carry a rich colour of feminization, and are thus suitable for addressing females. On the other hand, the ! pattern of a name plus a kinship term often occurs in letters written by males to their male friends. This phenomenon may reflect the tendency for males, particularly male friends, to value their mutual friendship as a kind of brotherhood. By using kinship terms, a male writer is able to express his close kin-like relationship with his friend and to reveal his awareness of the recipient's need to be regarded as a kin member. Thus, it can be suggested that in the Chinese context letters written between male friends are other-oriented, a style of language which is normally assumed to be a female trait (Holmes 1995: 38).

In summary, regardless of the similarities in the choice of some address forms by both females and males, females generally prefer to use terms emphasizing intimacy and in groupness, while males generally prefer to use terms indicating role-relationships and informality. In addition, the age factor plays a role in the gender differentiation and these differences occur when writers are addressing younger recipients. Moreover, the differences in the choice of address forms across genders also depend on the gender of the recipient.

4.3 Concluding remarks This chapter has examined the linguistic and pragmatic features of address forms used in Chinese personal letters. Based on the analysis of the findings generated from the empirical data, some concluding remarks can be made. First, the written addressing practices in the Chinese context are characterized by their diversity and elaborateness. Second, the use of written address forms is governed by social factors such as social roles, age and generation and gender, and the influence of each factor varies with the context of the communication. Third, the use of address forms in personal letters reflects the fact that the Chinese culture, especially its Confucian tradition, imposes strong conventions on the address practices used in written communication. It appears

126 that socio-cultural rules and norms, rather than personal habits or preferences, have the greatest bearing on the address practices used in Chinese personal written interaction.

127 CHAPTER FIVE OPENINGS IN CHINESE PERSONAL LETTERS

5.0 Introduction This chapter investigates openings in Chinese personal letters across family and non family contexts. The first part of the chapter describes the grammatical and pragmatic features of opening strategies and the ways that writers organize the start of written interaction. Thus, it attempts to explore the types and properties of openings, the main sequential structures, and the operation of the rules of sequencing based on the empirical data for the present study. The second part examines the relation of the opening behaviour and social factors (e.g. roles, age, gender) and cultural constraints in the Chinese context. It aims at examining to what extent these factors affect Chinese people's performance in the opening section of letters, reflecting socio-cultural ways of interactive written communication in this particular society.

5.1 Opening strategies This section looks at the linguistic realizations of opening strategies m order to understand the language patterns and pragmatic functions of these strategies. Twelve main opening strategies are identified from the data. As shown in Table 5.1, the overall preference for each strategy is in the following order: greetings (42.6%), inquiries (19.4%), acknowledgements of receipt of a previous letter (19.2%), apologies (5.5%), expressions of wishes (4.1 %), expressions of missing (3.5%), expressions of concern (1.2%), expressions of thanks (1.2%), reassurance about oneself (1.1 %), congratulations (0.8%), expressions of hope (0.7%), and sending regards to the other (0.4%). It appears that the greeting is the most commonly used strategy, with the inquiry and the acknowledgement of receipt of a previous letter also used extensively. The use of other strategies is limited, no more than 6 percent for each of the strategies. It should be noted that the strategies are not mutually exclusive, as they may co-occur in an opening. Some strategies are often combined to formulate a sequential opening, which explains why 558 instances of strategies are included, far more than the total number of 259 letter samples in this study.

128 Table 5.l.Categories and distribution of opening strategies in Chinese personal letters (558 instances)

Strategies Functions Patterns N 0/o 1. Greetings Reveal the 1.1 How-are-you phrase. 210 37.6 writer and the e.g. 'Nihao!' (how are you?) recipient's 1.2 Conventional formulae. 28 5.0 status and e.g. 'Jianxin yukuai!' (May you be happy to establish or read my letter.) maintain relationship Subtotal 238 42.6 between them 2. Inquiries Express concern 2.1 How has your 'x' been ?(health/happiness) 73 13.1 about well- e.g. 'Jinlai shenti jiankang shenhuo yukuai being of the rna? '(Have you been healthy and happy recipient recently?) 2.2 Are you well? 12 2.2 e.g. 'Ni hai hao rna? '(Are you well?) 2.3 Precondition+ how has your 'x' been? 12 2.2 e.g. 'Haojiu rnei gei ni xiexin le, jinlai shenti jiankang rna?' (I haven't written to you for a long time, have your health been ok?) 2.4 Precondition+ I'm just wondering 11 2.0 whether ... e.g. 'Henjiu rnei tongxin le, buzhi ni shifou yiqie dou ruyi?' (We haven't been in contact for a long time, and I am just wondering whether everything is going well with you.) Subtotal 108 19.4 3 .Acknowledg- Attend to 3.1 Bald acknowledgement. ments of receipt recipient's a. Passive structure. 15 2.7 of a previous expectation and e.g. 'laixin shoudao. ' (The letter indicate an in- letter has been received.) group b. Active structure. 6 1.1 membership e.g. 'Nide laixin wo yi shoudao. '(I have received your letter.) 9 1.6 3.2 Expressing happiness. e.g. 'Hen gaoxing shoudao nide laixin.' (I'm happy to receive your letter.) 1 0.2 3.3 Expressing gratitude. e.g. 'Xiexie ni zai bairnang zhizhong laixin. ' (Thank you for writing even though you're busy.) 76 13.6 3.4 Acknowledgement + expressing feelings. e.g. 'Zuotian shoudao nide laixin, wo xinli shifen gaoxing. ' (I received your letter yesterday, I am very happy.) Subtotal 107 19.2

129 4. Apologies Provide an 4.1 An explicit expression of apology. 5 0.8 apology for an e.g. 'Hen baoqian meiyou geinijishi quxin.' offence to ( Many apologies for not having written you restore a good earlier.) relationship 4.2 Admitting offence +requesting forgiveness. 12 2.2 e.g. 'Meiyoujishi geini huixin, haiwang yuanliang. '(I haven't written back, please forgive me.) 4.3 An explanation+ admitting offence+ 14 2.5 requesting forgiveness. e.g. 'Youyu wo zuijin henmang, meiyoujishi geini huixin, jingqing yuanliang. ' (Because I have been very busy recently, I haven't written back on time. Please forgive me.) Subtotal 31 5.5 5. Expressions Create happy 5.1 Wish you .... 23 4.1 of good wishes ambience by e.g. 'Zhuni shentijiankang, wanshi yuyi.' expressing good (Wish you health and happiness with wishes everything.) Subtotal 23 4.1 6. Expressions Express 6.1 Expressing missing. 9 1.6 of missing affection and e.g. 'Shennian!' (I miss you so much!) emotion 6.2 Preground + expressing missing. 11 1.9 e.g. 'Likai jia yi wuge yu le, wo feichang xiangnian nimen. ' (It has been five months since I left home, I am missing you very much.) Subtotal 20 3.5 7. Expressions Express 7.1 Expressing concern. 2 0.4 of concern consideration e.g. 'Xiwang ni duo zhuyi ziji shenti ,zhuyi referring to anquan. '(I hope you look after yourself well recipient's and keep safe.) future action 7.2 Precondition+ expressing concern. 5 0.8 e.g. 'Chunhan douqiao, wang ni duo zhuyi shenti. ' (In the chilly air of the spring period, please take care of your health.) Subtotal 7 1.2 8.Expressions of Express 8.1 Express gratitude. 5 0.8 thanks gratitude for e.g. 'Xiexie ni duiwo de wenhou he guanxin. ' benefit obtained (Thank you for your greeting and concern.) 8.2 Stating the benefit+ express gratitude. 2 0.4 e.g. 'Xiri de bangzhu, zaici biaoshi ganxie. ' (For the past help, I am grateful.) Subtotal 7 1.2 9. Reassurances Show 9.1 Reassurance about oneself+ ask the other 6 1.1 about oneself involvement and not to worry. consideration. e.g. 'Wo zaizhe li yiqie douhao, wunian. ' (I'm all right here, don't worry.) Subtotal 6 1.1 lO.Expressions of Express 10.1 Congratulations on .... 5 0.8 congratulations appreciation to e.g. 'Zhuhe ni chuwei renmu.' what the (Congratulation on becoming a recipient has mother.) achieved. Subtotal 5 0.8

130 11. Expressions of Indirectly 11.1 Express hope. 3 0.5 hope express concern e.g. 'Wang ni duo geijiali xiexin. '(Hope about the you write to the family more.) recipient 11.2 Preground + express hope. 1 0.2 e.g. 'Kuaiyao qimo kaoshi le, xiwang ni xinku yidian, duo xue yixie dongxi, duo jiao yixie hao pengyou. ' (The fmal examinations are approaching, I hope you are studying hard, learning lots and making good friends.) 4 0.7 Subtotal 12. Sending Send good 12.1 Sending regards to other with 'please.' 1 0.2 regard to other wishes to the e.g. 'Qing daiwen Liuming hao. ' (Please third party give regards to Liuming.) 12.2 Sending regards to other without 'please' 1 0.2 e.g. 'Daiwen Li shushu hao. '(Remember me to Uncle Li.) Subtotal 2 0.4 Total 558 100

5.1.1 Greetings Greetings are the most frequently used strategy among all the strategies (42.6%). This strategy serves to reveal the writer and recipient's social status in age or rank, to express the writer's attitude towards the recipient (e.g. respect, familiarity) (Chen 1991: 55) and to convey s desire for establishing or maintaining a relationship between them (Coupland et al. 1992, Goody 1972, Omar 1992,). Following Malinowski (1972: 151), the greeting constitutes a central example of phatic communion, since the main concern of the writer is not to exchange cognitive information, but to reflect that the writer wishes the recipient well, indicating they still share membership in a social group (Yahya-Othman 1995: 210).

As evident from the data, greetings in Chinese personal letters consist of two patterns: the 'Nihao' phrase and the conventional formulae. The former is far more preferred to the latter (37.6% vs. 5%). The phrase of 'Nihao' (literally 'you good.') here can be translated to 'How are you?' in English. However, unlike 'how are you?' in English, it is a statement, which serves to give regards to the recipient rather to make an inquiry. In the data, there are four variations for the 'How are you?' phrase, distinguished by the use of different terms of the second person pronoun: 'ni' (familiar, singular), 'nin' (honorific, singular), 'nimen' (familiar/honorific, plural), and 'ninmen' (honorific, plural).

131 'Nihao' (familiar, singular) is the most commonly used pattern among the four (162 out of 210). It is often used by parents to children, older relatives to younger relatives, teachers to students, or among siblings and friends. It can be seen as the basic unmarked greeting form. When it is used by Chinese people, the interpretations of its meanings may depend on the context. For example, when used downwards to juniors, this term appears to indicate an in-group membership, age hierarchy and a level of intimacy. When used among siblings and friends, it indicates in-group membership and intimacy as well. When used by teachers to students, it may emphasize familiarity.

'Ninhao' (honorific, singular), the second most widely used term (23 out of 210) is a respectful form of 'Nihao '. In Chinese, the 'nin' is used to reflect respect when people are speaking upwards to seniors (e.g. parents, older relatives, teachers). Similar to the 'ni ', this honorific singular pronoun is also relationship-oriented. For example, when used to seniors in the family, the 'nin' conveys not only the writer's respect, but also a level of intimacy. However, when used to seniors of the non-family (i.e. teacher), it indicates deference and distance as well.

'Nimen' (familiar/honorific, plural), an infrequently used term (12 out of 210), is the plural form of the 'ni ', where 'men' is the plural-meaning marker added to 'ni. ' This form is normally used to refer to double addressees (e.g. father and mother, aunt and uncle), and it often occurs among members of families or friends. Like other patterns of the same class, the 'nimen' is also clearly associated with the status and relationship. For example, when used upwards to the seniors, it emphasizes not only respect, but also intimacy as well. When used within the same generation or downwards to the younger generation, it signals in-group membership and intimacy.

The phrase of 'ninmen ' (honorific, plural) is the least common pattern among the four (11 out of211). This is the plural form of the honorific singular term 'nin.' In Chinese, this form is often found in written language, particularly in letters, and usually used upwards to double seniors in the family, indicating deference and closeness.

The other type is the conventional formulas of greetings. This type could be understood as a formal-traditional style of greeting. Mostly it is composed of an idiom, which may

132 tell of a writer's hopes and desires to the recipient by referring to some aspects (e.g. hope the recipient is happy when this letter arrives). In comparison to the greeting of 'How are you?', this type appears to be more personal, as it is connected to a complex display of affection and emotion of the writer. On the other hand, this type of greeting does not focus more on signaling the status difference between the writer and recipient, whereas the phrase of 'Nihao' always does. The main concern in using this kind of greeting is to strengthen the emotional bond with the recipient, and thus to enhance the relationship. As shown in the data, this pattern usually occurs among family members or among friends. The following patterns for the conventional greetings are extracted from the data:

1.2 Conventional formulae a. 'ADV+ADJ' Jin hao! recent good Hope you are well! b. 'V+ NP+V' Jian xin hao! see letter good Hello to you when you are reading this letter! c. 'V+NP+ADY' Jian xin yukuai! see letter happy May you be happy to read my letter. d. 'V+NP+V+NP' Jian xin zhiqing. see letter know information Here is some information when you are reading this letter. e. 'V+NP+PRO+V' Jian xin ru m1an. see letter as face It is as if we were face to face when you are reading this letter.

From these examples, it can been seen that a writer sends regards to the recipients by denoting hopes and desires through some highly restricted sets of adjectives, such as 'hao' (good) or 'yukuai' (happy), or verb phrases, such as 'hao' (greeting), 'zhi qing' (know something) or 'rumian' (as if we were face-to-face). In Pattern (1.2a) the writer

133 wishes the recipient is well by a very short phrase 'Jin hao' (Hope you have been well). Here the 'hao' is an adjective, which is used to indicate the writer's hope that the recipient's recent condition has been good. The expression of the writer's psychological state 'Hope you ... ' is important to the recipient since it shows the writer's positive attitude towards the recipient. This pattern is often used in friend-to friend contexts, indicating familiarity and in-group membership.

In Pattern (1.2b ), the writer gives regards by using the phrase 'jianxin hao' (Greeting to you when you are reading this letter). It is worth noting that the adjective 'hao' in this pattern acts as a verb, which means 'to greet.' It appears that the writer in this pattern emphasizes the action of greeting. The whole pattern can be interpreted as 'I am greeting you when you are reading this letter.' The emphasis on the greeting action seems to be important to the recipient in the same way as the hopes or desires expressed in 'Jinhao' (hope you have been well), as the expression of 'to greet' comes to reveal the writer's attention to the recipient, showing deference and affection. This pattern is often used by children towards their parents or among friends.

In Pattern (1.2c), the writer wishes the recipient well by focusing on hopes relating to the recipient's positive state of mind 'Jianxin yukuai' (May you be happy to read my letter). Through this pattern the writer reveals the emotion towards the recipient since it involves the anticipation of pleasure, which refers to the recipient's good feelings by referring to the reading of the letter, conveying a familiarity and in-group membership. Friends are often found to use this pattern in the data.

In Pattern (1.2d), the writer implicitly sends regards by using the phrase 'jianxin zhiqing' (Here is some information when you are reading this letter). From the literal meaning, it seems little emotion or attitude is expressed. But reading this pattern more carefully, it can be seen that the writer's emotion is conveyed through the verb phrase 'zhiqing' (know something). 'Zhiqing' here may not refer to literal meaning of 'knowing some factual things,' rather, to the writer's affection and attitude towards the recipient. The whole pattern can be interpreted like this: 'You may know my good feelings to you when you are reading this letter.' Based on this interpretation, it appears that this is a. writer-involvement pattern since a greeting is given to the recipient by

134 referring to the writer's own personal feelings, which may produce an impression on the recipient, thus enhancing the relationship. However, this pattern occurs infrequently, and is only found in parents' letters.

Pattern (1.2e) appears to be a more affectionate formula. In this pattern the writer gives regards by using the very intimate phrase 'Jianxin rum ian' (It is as if we were face-to face when you are reading this letter). 'Rumian' (as if we were face-to-face) is a metaphor which demonstrates the writer's strong affection. This pattern can be characterized as a writer-recipient-oriented greeting. The writer's psychological state expressed is of importance to the recipient, because it emphasizes that the writer and recipient share a 'common ground' (Brown & Levinson 1987), which would result in enhancing the relationship between them. This pattern is often used by friends, and sometimes by children to parents.

5.1.2 Inquiries Inquiries refer to expressions which ask about the recipient's current state in terms of health, well-being, or feelings. Inquiries here can be regarded as 'phatic inquiry' (Omar 1992: 16), since they are expressed not to ask the recipient to give a truthful or detained answer (Shih 1986: 138), but to show the writer's concern about the recipient in terms of some personal aspects, serving to mark "involvement and consideration" (Tannen 1984: 27). As the second most common opening strategy (19.4%), inquiries are prevalent among friends and siblings. The typical patterns occurring in the data are as follows:

2.1 'How has your 'x' been?' Jinlai shenti xuexi hao ba? Shenghuo deng yiqie dou ruyi recently health study good (isn't it) life and everything all satisfied

ba? (isn't it) Have things been going well in health, studying, life and other things recently?

2.2 'Are you well?' Ni hai hao rna? you still good (is it true) Are you still well?

135 2.3 'Precondition+ how has your 'x' been?' Jiu wei le, yiqie dou hao ba?' long time not see (particle) everything all good (isn't it) It's been a long time since we met last. Have things been going well?

2.4 'Precondition+ I don't know whether ... ' 'Hen chang shijian mei you nide xiaoxi, buzhi ni shifu )'lqte ruyi? very long time not have your news not know you whether everything happy There has been no information from you for a long time, and I wonder are you happy with everything?

As seen from the examples, patterns (2.1) and (2.2) are inquiries into the recipient's current situation. However, there are differences between these two. First, Pattern (2.1) is characterized as a syntactical formula, and it often involves two or three parallel sentences emphasising specific aspects of the recipient like 'Jinlai shengti hao ma? xuexi mang ba?' (How have you been in health? Have you been busy at study?). In Chinese culture, it is common for the writer to ask about the recipient's concrete state, such as 'health', 'life' or 'happiness.' Normally such inquiries are given not for getting a real answer to the question, rather for showing the writer's concern about the recipient, in order to "establish bonds ofpersonal union" (Malinowski 1972: 151). This pattern is often used by family members (e.g. siblings, children, parents) or by friends. Pattern (2.2), however, is constructed as a minimal version with an emphasis on the general state of the recipient. Interestingly, it appears to be a more personal pattern since the writer asks about the recipient's current state by referring directly to the recipient as in 'Ni hai hao ma?' (Are you still well?). The 'ni' (you) here can be seen as 'in-group identity marker' (Brown & Levinson 1987: 107), by which the writer tends to emphasize a strong consideration for the recipient, to assume a certain level of intimacy, and thus to maintain an emotional bond between each other. This pattern is often found among siblings and friends.

Patterns (2.3) and (2.4) are both realized through a compound structure where the core inquiry is preceded by a precondition, such as 'Haojiu mei gei ni xiexin le, xianzai shenti jiankang ba?' (I haven't written to you for a long time. Have you been in good health?). The preceded preconditions lay a "scene-setting relation" (Chu 1998: 231) for the following inquiries. Such preceded setting conditions appear to imply a close connection between the writer and recipient. With such preconditions, the inquiries

136 may produce a "dramatized" (Brown & Levinson 1987: 105) emotional atmosphere, from which the relationship with the recipient can be established or maintained.

It is worth noting that the core inquiries in these two complex patterns often take the form of 'How has your 'x' been?', as in 'Kaixue zheme chang shijian le, nide xuexi mang ma? Shenhuo fangmian ke hao?' (It has been a long time since the school started. Has your study been good? And has your life been good?), or 'I'm just wondering whether ... ', as in 'Zanliang haojiu mei tongxin le, buzhi ni shifou shiying le nali de shenghuo?' (We haven't been in contact for a long time. I am just wondering whether you have been adapted to the life over there?). It is clear that the core inquiry in the former is the 'A-not-A question' (Li & Thompson 1981: 550), which directly expresses the writer's concern about the recipient in relation to the aspects of study and life. However, the core inquiry in the latter is a 'disjunctive question' (Li & Thompson 1981: 550), which indirectly conveys the writer's concern about the recipient's adaptation to a new environment. With the disjunctive question of 'I'm just wondering whether you ... ', the writer signals a certain degree of worry and anxiety about the recipient's current situation, implying stronger concern about the recipient. Moreover, this pattern is a writer-then-recipient-oriented pattern where the writer makes the phatic inquiry through involving a psychological state of mind. It can be seen as a typical way of showing the writer's "involvement and consideration" (Tannen 1984: 27). The data show that siblings and friends are inclined to use these two patterns.

5.1.3 Acknowledgements of receipt of a previous letter As an opening strategy, acknowledgements of receipt of a previous letter occur third most commonly (19.2%). On the linguistic surface, this strategy is to give information about the previous letter from the recipient, (received or not). But pragmatically, it belongs to the phatic expressions, which have minimal commitment to 'factuality' (Pavlidou 2000: 126). It serves to meet with the recipient's expectation by providing a message that the recipient wishes to know, and to reflect the writer's positive attitudes towards getting the letter, indicating that they still share a social co-membership and confirming a personal connection. Linguistically, this strategy is achieved through four patterns as follows:

137 3.1 'Bald acknowledgement' a. 'Passive structure' Wuyue fen lai xm yi shoudao. May period come letter already receive The letter dated in May has been received.

b. 'Active structure' Nide lai xin wo yi shoudao. your come letter I already receive I have received your letter.

3.2 'Expressing happiness' Hen gaoxing shoudao nide lai xm. very happy receive your come letter I'm very happy to receive your letter.

3.3 'Expressing gratitude' Xiexie ni zai bai mang zhizhong gei wo x1e xm. thank you in hundred busy among give I write letter Thank you for writing me even though you're busy.

3.4 'Acknowledgements+ expansion' Lai xm shoudao. Zhidao ni qingkuang dou hen hao, come letter receive know you situation all very good

jiali dou hen gaoxing. family all very happy Your letter has been received. Knowing you are ok, all the members in the family are very happy.

Pattern (3.1a) and Pattern (3.1b) can be seen as a bald acknowledgement of the arrival of the letter. The difference between them is that the former is a passive structure without the 'bei' phrase (Li & Thompson 1981: 492), while the latter is an active structure, such as 'Nide laixin wo yi shoudao' (I have received your letter.). For the former, the focus is on the object (letter) ofthe verb 'shoudao' (receive). With such an emphasis, the writer tends to satisfy the recipient's expectation by confirming the arrival of the previous letter. In contrast, the focus in the latter is on the subject of the object (letter)-'wo' (I), 'women' (we) or 'jiali' (family). The first personal pronouns 'I' and 'we', or the noun 'family' here can be seen as 'in-group markers,' which are used to express the writer's attention to the recipient, emphasizing in-group membership. Thus, comparatively, the active structure of acknowledgement may be regarded as a more interpersonal oriented type. It is often used in family letters (e.g. parent to children,

138 children to parent, sibling to sibling), but the passive structure can be used in both family and non-family letters.

Patterns (3.2) and (3.3) are fairly similar to the expressions in English 'Very happy to receive your letter' and 'Thank you for writing.' They can be seen as typical expressive acknowledgements of receiving the letter since the underlying goal is explicitly associated with establishing or enhancing the relationship with the recipients. For example, the focal message contained in Pattern (3.2), exemplified by 'Hen gaoxin shoudao nide laixin. '(I'm very happy to receive your letter.), is on the writer's positive feelings towards receiving the previous letter. By framing the acknowledgement of receiving the letter within an expression of happiness, the writer shows interest in the letter, claiming a common ground (Brown & Levinson 1987: 103). The adjective 'gaoxin ' (happy) often occurs along with some intensifiers, such as 'hen ' or 'feichang' (very), which are used to express a strong degree of happiness. The focus of Pattern (3.3) is on expressing gratitude for being sent the letter, as exemplified in 'Feichang ganxie ni zai baimang zhizhong laixin. ' (Thank you very much for writing me even though you're busy.). Involving thanks with the acknowledgement of receiving letter shows the writer's attention and respect to the recipient. However, these two patterns occur infrequently in the data, nine cases (1.6%) for Pattern (11), and only one (0.2%) for Pattern (12). Pattern (11) is often used among friends, while Pattern (12) is only for a younger to elder sibling.

Pattern (3.4) is the most frequently used acknowledging pattern, which accounts for 13.6% of the total instances. It is an elaborate structure, where a bald acknowledgement of receiving letter is followed by an expression of the writer's feelings or attitudes towards the receipt (e.g. happy, excited) such as 'Laixin shoudao. zhidao ni qingkuang dou heng hao, jiali dou hen gaoxing. ' (The letter has been received. Knowing you are ok, all the members in the family are very happy.). Such a rhetorical pattern reflects the writer's perception of the need to increase the phatic function of the acknowledgement by a separate expression of feelings. The separate expressive act after the acknowledgement could be distinguished from the minimal version 'I was very happy to receive your letter' by separately emphasizing the spiritual effect of the previous letter on the writer. This kind of structure appears to be more helpful for the writers to

139 claim in-group membership, a level of intimacy, thus reinforcing the relationship with the recipients. This pattern is often used between siblings and by parents to children.

There are several adjectives which are used to express the writers' emotion or attitudes such as 'gaoxin' (happy, 61 out 76), 'jidong' or 'xingfen' (excited, 5 out of 76), 'wennuan' or 'wenxin' (warm, 2 out of76), and 'anwei' (comforted, 1 out of76). Apart from these adjectives, the verb 'fangxin' (relieved) can also be used to indicate the mental state of emotion (7 out of 76). Not surprisingly, 'gaoxin' (happy) is by far the most common, and is often used by parents, among siblings or friends. Usually, 'jidong' or 'xingfen ' (excited) is used between friends or by children, 'wennuan ' or 'wenxin' (warm) is used between siblings or friends, 'anwei' (comforted) is used by children, and 'fangxin' (relieved) is used by parents.

5.1.4 Apologies In oral communication, apologies are viewed as social and affective speech acts (Holmes 1990). They are primarily used to "change what might be seen as an offensive act into an acceptable one" (Fraser 1981: 259), and to attend to the hearer's face-need in order to maintain a good relationship between participants (Holmes 1990). As an opening strategy in written communication, apologies are related a writer's admission of delay in writing back and a request for forgiveness for the delay. Pragmatically, they function as 'face-supportive acts' (Holmes 1995: 155) and a 'remedy' (Gof:finan 1971: 140) for the delay, serving to restore social equilibrium or harmony (Edmondson 1981: 280; Leech 1983: 125). As an opening strategy, apologies occur infrequently in the data and account for 5.5% of the total instances. It is appropriate for the juniors in family letters (e.g. younger siblings, younger relatives), and for both seniors and juniors in non family letters (e.g. teachers, students). This strategy is mainly realized through the following patterns:

4.1 'An explicit expression of apology' Hen baoqian mei you ge1 m jishi x1e xm. very sorry not have give you on time write letter Sincere apologies for not having written to you earlier.

140 4.2 'Admitting offence+ requesting forgiveness' Wei jishi gei ni hui xin, qing jianliang. not on time give you back letter please forgive Please forgive me not having written back to you on time.

4.3 'An explanation+ admitting offence+ requesting forgiveness' Y ouyu wo hen mang, shuoyi yizhi mei you gei ni xie xin, because I very busy so all the time not have give you write letter

qing yuanliang. please forgive Because I have been very busy all the time, I haven't written to you on time, please forgive me.

In Pattern (4.1) the writer explicitly expresses apology for a delay in writing back. The inclusion of 'hen' (very) in this case intensifies the sense of apology. Usually, the explicit expression of apologies for some offences is regarded as a formal apology, which serves to indicate formality and distance, as suggested by Holmes (1990: 173). In the data this pattern is often adopted by friends, although the results cannot be generalized (5 out of 31: 16.1%). Choosing this formal pattern, the writers tend to emphasize their sincerity embedded in apologies towards the friend, thus conforming to or consolidating the friendship.

Pattern (4.2) is structured through an explicit acknowledgement of a late response and a request for forgiveness, such as 'Weijishi geini huixin, qingjianliang' (I haven't written back to you on time, please forgive me). This type of apology seems to be more indirect than Pattern (4.1 ), as the writer explicitly acknowledges the offence, but does not explicitly do an 'apology' or demonstrate 'regret.' Although there is no strong connection to the explicit expression of apology, the seeking of forgiveness does linguistically entail regret for having brought some uncomfortable consequences to the recipient (i.e. worry or anxiety resulting from having not received a reply letter). Thus, it emphasizes the writer's attention to the recipient's face-needs, that is, to be noticed or valued (Brown & Levinson 1987: 103). This apology strategy is often used by juniors towards seniors in the family letters (e.g. younger to elder siblings, younger to older relatives), or among friends, indicating a level of deference and in-group membership.

Pattern (4.3) is the most frequent apology strategy selected by the writers in the data (14 out of 31 ). It is the most elaborate pattern consisting of an explanation, an admission of

141 the delay, and then a request for forgiveness, such as 'Youyu wo hengmang, suoyi yizhi meiyou geini xiexin, jinqing yuanliang. ' (Because I have been very busy all the time, I haven't written to you on time. Please forgive me.). Expressing apology by providing a reason in the initial position, the writer seems to claim a compelling reason for the delay. According to Holmes (1990: 162), such explanation serves to remedy a hurt to the recipient's 'positive face.' This could be seen as an effective way of restoring social harmony that existed between the writer and recipient. By combining an explanation with an acknowledgement of the offence and a request of forgiveness, the writer tends to convey information that does not mean to offend the recipient. All that has happened is due to a particular reason, and hopes the recipient will forgive the writer. This strategy is often used by the juniors in the family letters (e.g. children, young siblings) and both seniors and juniors in the non-family letters (e.g. teachers, students).

5.1.5 Expressions of good wishes This strategy refers to expressions wishing good fortune on the recipient. They can be seen as good remarks which are said to increase one's good luck in terms of good health, happiness, improvement, enjoyment and so forth. They are used to create a happy ambience (Pong 2000: 224), thus, consolidating the relationship between writer and recipient. This is an infrequent opening strategy in the data, accounting for only 4.1% of the total instances, and it is often used by juniors in the family and non- family letters (e.g. children, young relatives, students), serving to convey respect and affection to the recipients. Usually this strategy is realized through an exclamatory sentence: 5.1 'Wish you .. .' Shouxian zhu ni shengti jiankang, tiantian youge hao xinqing! at first wish you body heathy day day have good mood At first, I wish you good health and good feelings every day!

In this pattern, the writer expresses the wishes that are related to the recipient's state of health and happiness. As the data show, the expressions of good wishes as an opening strategy, are often related to the Chinese values of 'good health' (8 out of 23), 'smoothness' (5 out of23), 'happiness' (4 out of23), 'improvement' (4 out of23), and 'enjoyment on a special occasion' (2 out of 23). Such a well-wishing event, according to Fang (2000: 233; 235), displays the Chinese people's communicative and pragmatic understanding of 'luck talk' which is strongly linked to the Taoist philosophical principle of 'yin' and 'yang' (dualism, detailed in chapter six) in Chinese culture.

142 Within this Taoist principle, the concept of luck involves humans living in a harmonious relationship with the environment and energy lines through proper environmental arrangements, so as to gain maximum benefit, peace, health, and prosperity (Too 1996). It is such belief that enables Chinese people to engage in expressing wishes to the other by choosing common topics, and to show respect and rapport among people. As appeared in the data, the expressions of wishes often consist of more than one topic, such as 'Zhuni shenti jiankang wanshi ruyi. ' (Wish you health and happiness with everything.). Combining two (occasionally three) topics to express wishes may reflect Chinese people's perception of the need to "increase one's good luck in life" by fulfilling cultural values (Pong 2000: 233).

5.1.6 Expressions of missing Expressing missing can be seen as another affective speech act, which serves to express the writer's thoughts about the recipient, indicating strong affection, thus reinforcing the relationship. It accounts for 3.5 percent of the total instances. This strategy appears to be prevalent among children and siblings. The expression of missing in the data is framed either with a simple sentence or a complex pattern where the core act is preceded by an adverbial clause of reason as shown in the following patterns:

6.1 'Expressing missing' Feichang xiang nimen! very much miss you I miss you very much!

6.2 'Preground +express missing' Jiu wei tong xin, shifen dianji long time not go through letter very think about We haven't been in contact for a long time, and I am thinking about you very much.

Pattern (6.1) is a minimal version where the expression of missing is realized through an elliptical sentence in which the subject (I) is omitted such as 'Feichang xiangnian ni. '(I miss you very much.). Occasionally writers would use a traditional-style phrase, such as 'Shennian!' (Missing very much) or 'Yuannian!' (Missing from a far place). As shown in the data, the common verbs denoting 'thinking about' is 'xiangnian' (18 out of 20). Occasionally 'dianji' (thinking about) or 'guanian' (worrying about) are also used,

143 both of which involve some level of anxiety along with missing. 'Xiangnian ' is appropriate for children, siblings or friends to emphasize a kind of mood of thinking over, while the latter two are restricted to parents or elder siblings to imply care and consideration. Usually, the writer may use some intensifiers such as 'feichang' (extremely), 'hen' (very), 'shifen' (quite) or 'zhen' (truly) to modify the verb 'xiangnian' (missing). The first three appear to be used widely among parents, children and siblings, while the last one seems to be appropriate only for friends.

Pattern (6.2) is a complex sentence where the core act of expressing missing is preceded by a preground offering reasons such as 'Henjiu mei gei nimen xiexin le, feichang xiangnian nimen.' (I haven't written to you for a long time, and I'm missing you very much.). On the surface level, this reasoning displays a factual message, but pragmatically it implies the writer's sadness by referring to a long period of little contact in correspondence with the recipient. Thus, it may help to intensify the degree of missing expressed in the main clause. The data demonstrate that the compound structure occurs slightly more than the simple version (55% vs. 45%).

5.1.7 Expressions of concern Expressions of concern refer to expressions which convey a writer's consideration and concern about personal issues related to the recipient. It serves to make an emotional bond with the recipient, aimed at maintaining a good relationship. This is an infrequent strategy used by Chinese writers in the opening section (1.2%). It is more appropriate for parents and friends, and occasionally for children and siblings, whereas no case is found between teachers and students. There are two representative patterns for this speech act:

7.1 'Expressing concern' Xiwang ni duo zhuyi ziji shenti, zhuyi anquan. hope you more attention self body attention safety I hope you are looking after yourself and keeping safe.

7.2 'Precondition+ expressing concern' Yan dong bi jin, wang nin duo zhuyi shenti. deep winter come near, hope you more care health The cold winter is approaching and I hope you are looking after yourself.

144 In Pattern (7 .1) the writer uses a simple sentence 'I hope you are looking after yourself and keeping safe' to convey positive intent. It implies the writer's affection or emotion towards the recipient. The data show that the aspects for concerning about mainly focus on the issue of health. This pattern is most appropriate for parents.

Pattern (7 .2) is a complex formula where a precondition is employed before the main expression of concern, such as 'The cold winter is approaching, and I hope you are looking after yourself.' Mentioning the weather in the initial position seems to play a role in laying the ground for the writer to convey concern about the recipient, making the whole discourse more convincing. Such a structure may help a writer convey the concern about the recipient in a logic and relevant manner, which is important for the relationship the writers tend to establish or maintain. As shown in the data, this elaborate construction is often used by friends.

5.1.8 Expressions of thanks Thanking is a strategy of expressing indebtedness for what the recipient did for the writer. According to Coulmas (1981: 71), the act ofthanking does not imply the factual, but rather conveys warm and friendly feelings and confirms the relationship with the recipient. This strategy appears most appropriate for friends in the data. Linguistically, thanking is often realized through two patterns.

8.1 'Express gratitude' Xiexie ni dui wode wenhou he guanxm. thank you to my greet and concern Thank you for your greetings and concern.

8.2 'Stating the benefit+ expressing gratitude' Xiri de bangzhu, zai ci biaoshi ganxie past help, in here express gratitude For help in the past, thanks again.

Pattern (8 .1) is used more commonly in the strategy of thanking ( 5 out of 7). Through this pattern, the writer expresses a general feeling of gratefulness to a specific object of gratitude, such as 'Xiexie ni dui wo de wenhou he guanxin. ' (Thank you for your greeting and concern.). In Pattern (8.2) the writer places the object of gratitude 'xiri de bangzhu' (for the help in the past) in the precondition position. Expressing gratitude in this way, the writer tends to emphasize the object benefited from, driving the recipient's

145 attention to the significance of the message: The writer has benefited from the help offered by the recipient. This pattern occurs infrequently (2 out of 7). It is worth noting that the common objects of gratitude used in the data are 'help' (4 out of 7), 'greeting and concern,' 'good wishes and comfort,' and 'respect' (3 out of7 altogether).

5.1.9 Reassurances about oneself This strategy refers to a positive description about the writer's current state and claims willingness to comfort the recipient. Pragmatically, this strategy is used not to denote factual information (the message may or may not be true), but to indicate or reaffirm the relationship in paying homage to the recipient's face need, that is, to be noticed, cared. about (Brown & Levinson 1987: 103), by appealing to in-group membership. Based on the limited occurrences in the data (1.1 %), this strategy appears to be most appropriate for children, serving to emphasize an emotional bond with their parents. Normally, this speech act is framed through the following pattern:

9.1 'Describe the current state+ reassure the recipient' Wo yiqie dou hen hao, wu nian. I everything all very good not worry Everything is all right with me. Don't worry.

As shown in Pattern (9 .1 ), the reassurance is realized through an explicit reference to the writer's general state, for example, by using 'Wo yiqie dou hen hao.' (Everything is all right with me.) followed by an expression asking the recipient to be at ease by using 'Wunian' (Don't worry.). By assuming one's positive situation and reassuring the other, the writer tends to show attention to the recipient's emotional state (e.g. worry, anxiety), thus emphasizing in-group membership and greater closeness.

5.1.10 Expressions of congratulations Congratulations here can be seen as a speech act expressing the writer's positive attitude towards what the recipient has achieved. As with some other routines, this strategy is related to the indication and enhancement of the relationship by showing the writer's positive attitudes or emotion towards the recipient. Using this strategy, the writer tends to claim a common ground (Brown & Levinson 1987: 103) with the recipient by indicating that they share a specific want or goal (e.g. to be successful in some respects). However, this strategy occurs infrequently, accounting for only 0.8% of the total

146 instances. It appears to be appropriate for friends. Sometimes it can be extended to parents, siblings or relatives, but is never used by children or students. The typical pattern is as follows:

10.1 'Congratulation on' Zhuhe ni zouxiang gongzuo gangwe1. congratulate you go forward work positiOn Congratulations on your success in finding a job.

This pattern is very similar to that in English as 'congratulations on .... ' Congratulations in Chinese are often expressed through two verbs: 'zhuhe, 'as in 'Zhuhe ni zoushang gongzuo gangwei. '(Congratulations on your success in finding a job.), and 'gongxi,' as in 'Gongxi ni zhangda le.' (Congratulations on your growing up.), which similarly indicate pleasure in light of the achievement by the recipient in some aspects.

5.1.11 Expressions of hope Expressions of hope can be seen as statements referring to the writer's desire that the recipient do something (e.g. study hard, write a letter). Similar to the cases of other opening strategies, the goal for using this strategy is not to have the recipient do something, but rather to show the writer's attitudes towards the recipient (e.g. care). This strategy is often used by parents, serving to convey affection to the children, and implying the parent's higher status as well. The following are two major representative patterns for this act:

11.1 'Expressing hope' Xiwang you kong duo gei jiali x1e xm. hope have time more give family write letter I hope you write to the family more often in your spare time.

11.2 'Preground + expressing hope' Kuai yao qimo kaoshi le, xiwang ni xinku yidian, duo xue YJXIe Fast get final examination (particle) hope you hard little more study some

dongxi, duo jiao yixie hao penyou. things more make some good friends The final examinations are approaching, and I hope you are studying hard, learning lots and making good friends.

147 Pattern ( 11.1) is a simple structure in which the writer implicitly expresses hope through the display of a hope for the recipient, as in 'Xiwang you kong duo gei jiali xiexin. ' (I hope you write to the family more often in your spare time.). On the surface level, the expression in the example seems to reflect that the writer is advising the recipient to do something, but a closer investigation indicates that this speech act is expressing the writer's psychological state of mind: He is looking forward to receiving more letters from the recipient. Therefore, this expressive act is actually conveying the writer's attention to the recipient, indicating a tight relationship between the two, and also revealing a level of authority over the recipient. Pattern (11.2) is a compound structure where the expression of hope is realized through a pre-ground followed by the core expression. In the example, the sentence 'Kuai yao qimo kaoshi le, ' (The final examinations are approaching) serves to lay ground work for why the writer asks the recipient to study harder and make more good friends as indicated in the following main clause, thus making the whole strategy more convincing and affectionate. By doing so, the writer tends to show care about the recipient, emphasizing a close relationship. In the data, the elaborate structure of Pattern (11.2) is less used than simple one in Pattern (11.1).

5.1.12 Sending regards to others This strategy refers to a request of passing a writer's salutation to the third party. In general, it can be seen as a 'consolidatory' remark, which is used to confirm and reinforce a wide network of common acquaintances (Laver 1981: 303). As an opening strategy, it is used the least commonly in the data (0.4%), and it only occurs in seniors' letters ofthe family (e.g. parents, elder siblings). It serves to highlight the attention that the family seniors typically apply to the care of their relationship with other family members. Basically, this strategy is structured as in the following patterns:

12.1 'Sending regards to other with 'please' Qing daiwen liu min hao. please ask liu min good Please give my regards to Liu Min.

12.2 'Sending regards to other without 'please' Daiwen li shushu hao. ask li uncle good Remember me to Uncle Li.

148 As seen in Patterns (12.1) and (12.2), this strategy is realized through an imperative structure, such as 'Remember me to Uncle Li,' and 'Please give my regards to Liu Min.' The difference between the two patterns is that the former is linguistically marked by a politeness marker 'qing' (please), while no such a word is found in the latter. In the Chinese context, the word 'qing' (please) is seen as a form of politeness and it "appears to attach different degrees of emphasis to its pragmatic value" as a modificatory device (Lee-Wong 2000: 261). This holds true for this strategy. By framing the 'giving regards to other' with 'please,' the writer tends to involve a level of deference in addition to the emphasis of in-group membership. It is found in the data that the deferential regard is used by younger siblings, while the general regard is often adopted by parents.

5.1.13 Discussion After examining the linguistic and pragmatic features of opening strategies used in Chinese personal letters studied here, some points can be summarized and discussed. First, Chinese writers use a series of expressive speech acts to achieve openings in personal letters. As the findings show, openings in the data for this study include various expressive acts, such as greetings, inquiries, apologies, thanking, expressions of hope, expressions of missing, and reassuring oneself. In general, these speech acts used in the opening section are not to exchange cognitive information between writer and recipient (that might emerge in the main body of a letter), but primarily to make the acts reveal the status of the writer and the recipient (e.g. senior, junior), display the state of interpersonal relationship between them (e.g. familiarity, intimacy, distance), and convey positive feelings and attitudes from the writer to the recipients (e.g. care, respect). Following Malinowski's analysis of phatic communion, these acts are "not serving any purpose of communicating ideas" (1972: 150), but establishing "bonds of personal union between people brought together by the mere need of companionship" (1972: 151). Therefore, we can say that openings in Chinese personal letters can be seen as a management of a range of expressive acts reflecting the phatic function.

Why then do Chinese writers use these expressive acts, these phatic expressions, to achieve openings in personal letters? In answering this question, one needs to look at the feature of openings in written interaction, particularly in personal letters. Using

149 Laver's functional approaches to openings in conversation (i.e. propitiatory, exploratory, initiatory), it is assumed that openings in written interaction act to help the writer establish or maintain the relationship with the recipient, thus making the written interaction transfer into the main body smoothly (1975, 1981). Since the underlying goals of some expressive acts or phatic expressions are seen as "establishing relationship and achieving transition" (Coupland et al. 1992: 212), and more readily deployed when participants initiate communication primarily for social purposes, Chinese writers, therefore, achieve openings through the use of a range of expressive acts with the phatic functions.

Another explanation for the use of phatic expressive acts in openings is related to manifestations of politeness. Generally, openings in written interaction can be seen as an event which might be face-threatening and hence "sensitive to strategic manipulation" (Blum-Kulka 1990: 265). Thus, writers need to pay tribute to the mutual face wants, that is, to claim common ground, to show respect (Brown & Levinson 1987: 103), and to convey involvement and consideration (Tannen 1984). It is with such face work (positive and negative politeness) that writers can be "building up their esteem" (Yahya-Othrnan 1995: 211) in the eyes of the recipients, and thus keeping their interaction moving successfully. Thus, politeness becomes one of the major considerations in people's opening behaviour. In order to achieve the purpose of politeness, therefore, Chinese writers employ speech acts which carry the function of politeness in their openings.

Second, the findings denote that these Chinese writers display a preference for the choice of 'greetings,' 'inquiries,' and 'acknowledgements of receipt of a previous letter' within the opening strategies. Their prevalence can be understood in a broader Chinese cultural context where these opening strategies are used. According to the Confucian tradition, Chinese people are obliged to adapt their talk to the addresser or addressee's age, status and intimacy of relationships (Hopper & Chen 1996: 299). Greetings as a central aspect of phatic communion (Yahya-Othrnan 1995: 21 0) can be seen as greatly capable in achieving such a goal, since the greeting token indicates a writer's certainty about the identification of the writer and the recipient's status. In addition, because in greetings there are various linguistic patterns that the writer can use to define status

150 relationships, intimacy or solidarity, greetings thus become an important social index in acknowledging the differentiations in social positions between the writer and recipient. Therefore, it is not surprising that greetings are overwhelmingly used by Chinese people m operungs.

The favouring of the inquiry strategy also reflects the influence of Chinese culture on speech acts. As is well known, showing consideration for others is highly valued and longstanding in Chinese culture (Zhan 1994). In his study on Chinese greeting expressions, Chen (1991: 58) points out that China was loosely made up of many small self-confined communities during its long feudal history, and the members in each community were in close contact with each other. In order to maintain hierarchical order and harmonious relationship in a social group, it was natural for people to convey consideration by asking about the other's state of health, well-being, or other personal issues when people dealt with each other. In doing so, according to Hwang (1987: 952), individuals tended to influence each other by means of 'renqing' (genuine human warmth) and 'mianzi' (face), especially when the addresser was a senior. From this, the individuals showed an element of sharing "the same personal network" with the other and expected they would continue this connection (Hwang 1987: 953). Inquiries are a very important act in maintaining interpersonal relationships. Therefore, inquiries have long been an important opening strategy for attaining desirable social interaction in Chinese society.

The preference for acknowledgements of receipt of a previous letters well illustrates Chinese writers' perception of the need to satisfy the recipients' expectation, which can be seen as influence of sincerity, another cultural trait in the Chinese value system (Oliver 1971: 92). In the Chinese context, acknowledging and making comments on the previous letter has been regarded as a 'right' thing to do in writing communication, since it can release the recipient's worry with respect to the state of the previous letter. More important, this strategy can convey the writer's valuing the recipient by showing feelings or positive attitudes towards the letter (e.g. interested, happy or excited). Enhancing one's influence over the other in Chinese society can be seen as a face-work (Bond & Hwang 1989: 225), as the perception that an individual's value can guarantee further connection. If a writer fails in acknowledging and making comments on the

151 previous letter, the recipient may feel disaffected and unimportant to the writer in the initial stage. The potential for the disruption of their relationship and their social network may then exist. Therefore, announcing the arrival of the letter and expressing feelings also becomes significant in openings of Chinese letters.

Third, Chinese people exhibit a preference for shaping the opening strategy in an elaborate manner, and this feature reflects the influence of conversational logic proposed by Grice (1975). As discussed earlier, many opening strategies, such as inquiries, acknowledgments of arrival of a letter, expressing missing, concern or hope, are linguistically structured in a way where the core opening part is often preceded by a precondition or pre-ground, or followed by an expansion. While formulating an opening strategy either in the way of setting the scene or reasoning, the writers may make the discourse more logical and clear, which will generate the effect of increasing the degree of Relevance and Clarity embedded in the opening strategy (Grice 1975). From the rhetoric point of view, such usage also reflects Confucian influence on people's speech style in the opening behavior. Confucius states that "rhetoric is for the purpose of expressing sincerity" (Shih 1986: 62). Applying a Confucianist view to this study, it appears that when a writer does an act of expressing missing within a compound structure 'It has been five months since I left home, and I am missing you very much', this person tends not only to achieve the logic of the whole discourse, but more importantly to reflect his/her real feeling of missing by using the pre-ground which lays a reason for the core act. The elaboration in structuring the opening strategy can be also seen as a way of achieving politeness, since on the quantitative basis, the stronger the manifestation of the value (sincerity in this case) reflected in linguistic forms (i.e. compound structure), the more likely the strategy will be perceived to be polite.

5.2 Opening types This section examines the opening types and describes how each type is linguistically realized in Chinese personal letters. In the data, two types of openings are identified: the single opening and the combined opening. A single opening involves only one opening strategy, while a combined opening consists of two or more strategies.

152 Table 5 .2. Frequency and distribution of opening types in Chinese personal letters (N=259) Types N % !.Single opening 45 17.4 2.Cornbined opening 214 82.6 Total 259 100

As shown in Table 5.2, of a total of 259 openings, 45 cases are made with a single strategy, accounting for 17.4%. 214 cases involve the elaborate structuring in the opening, accounting for 82.6%. This distribution indicates that combined openings are overwhelmingly preferred in Chinese personal letters. The following section will examine these two types of openings respectively. Before doing this, two questions deserve an attention. First, which strategies are appropriate to form a particular type of opening? Second, how are different strategies combined to form a combination pattern?

5.2.1 Single openings As showed in Table 5.3, six strategies can be used to form a single type of opening. Among the six, the first three strategies (greetings, inquiries, and acknowledgements) are used much more frequently (more than 90% in the total). This finding again confirms that greetings, inquiries and acknowledgements of receipt of a letter comprise three key strategies used in realizing the openings, including the single opening.

Table 5.3. Frequency and distribution ofthe strategy of single openings (N=45) Strategies N % I. Greetings 24 53.3 2. Inguiries 12 26.7 3. Acknowledgments of receipt of a 5 11.1 previous letter 4. Apologies 2 4.4 5. Expressions of good wishes I 2.2 6. Expressions of missing I 2.2 Total 45 100

By examining the data more carefully, it can be seen that the occurrence of the single opening varies with the different type of letters. As suggested in Table 5 .4, the single opening occurs more frequently in initial letters than in reply letters (26.2% vs. 6.1 %), and more topics are selected in initial letters than in reply letters. The topics of the single opening in initial letters include greetings (15.2%), inquiries (8.3%), apologies

153 (1.4%), good wishes (0.7%) and expressions of missing (0.7%), while the topics for reply letters only include two strategies: greetings (1.8%), and acknowledgements of receipt of a previous letter (4.4%). These finding suggest that writers are inclined to use the single opening and have more options in choosing topics in initial letters than in reply letters.

Table 5.4. Frequency and distribution of single openings in initial and reply letters (N=259) Strategy Initial letter: 145 Reply letter: 114 N 0/o N % I. Greetings 22 I5.2 2 1.8 2. Inquiries I2 8.3 0 0 3. Acknowledgements of receipt 0 0 5 4.4 of a previous letter 4. Apologies 2 I.4 0 0 5. Expressions of good wishes I 0.7 0 0 6. Expressions of missing I 0.7 0 0 Total 38 26.2 7 6.1

A possible explanation for these differences lies in the factor of contact frequency between writer and recipient (Chen 1991; Jenkin & Hinds 1987; Lang 1994). In the case of initial letters, the writer has not been in contact with the recipient for a period of time, and thus it is understandable that the writer is likely to perform the opening with a single strategy, since there is less common ground the writer and recipient can share with and refer to. Moreover, also because of the lack of contact, the choices of the topics in the initial letters mainly rely on the writer's own experience and are likely to be limited in a narrow range. In contrast, in the case of reply letters, the writer has had recent contact with the recipient, and thus is more likely to have a common ground to share with, and thus more topics to refer to.

5.2.2 Combined openings The combined opening type is structured in a much more complex way than the single opening. What needs to be examined here is which strategies are used and how the structures are constructed. Since the initial and reply letters are two different domains where the constitution of opening components are quite different, they need to be investigated separately.

154 5.2.2.1 In initial letters Four patterns of combined openings are identified from initial letters in the data. As shown in Table 5.5, 'greeting+ inquiry' is the most commonly used pattern (41.1%), followed by 'greeting+ inquiry+ other expressive act' (29.9%) and 'greeting+ other expressive act' (22.4%) and lastly 'inquiry + other expressive act' (6.5%). It is interesting to note that nearly all the combined openings in initial letters (1 01 out of 107) contain the strategy of 'greeting,' which can be seen as almost obligatory in any combined opening. The strategy of 'inquiry' is the second most common component (82 out of 107).

Table 5.5. Combination patterns of openings and their distribution in initial letters (N=l07) Combination patterns N % 1. Greeting + inquiry 41 38.3 2. Greeting +inquiry+ other expressive act (e.g. apology/concern/missing/good 35 32.7 wishes) 3. Greeting+ other expressive act (e.g. apology/concern/missing/good wishes) 24 22.4 4. Inquiry+ other expressive act (e.g. hope/missing) 7 6.5 Total 107 100

Combination 1: Greeting + inquiry Findings in Table 5.5 indicate that 38.3 percent of the combined openings in initial letters are structures with the pattern of 'greeting+ inquiry.' As seen in Examples (30) and (31 ), this is a combination with a greeting in the initial position, and with an inquiry in the second part.

Example (1) (parent to son) (1) Ni hao. (2) Jinlai m )'lqie dou hai hao? you good recent you everything all still good How are you? Has everything been well with you recently?

Example (2) (elder sister to younger sister) (1) Ni hao. you good

155 (2) Zanliang haojiu mei tong xin le, ni ge fangmian hai hao we two very long not go through letter (particle) you every aspect still good

ba? (isn't it) How are you? We haven't been in contactfor a long time. Has everything been ok?

As seen in these examples, both openings are initiated with the familiar form of 'How are you?' (Nihao). The use of 'Nihao' phrase here serves to imply the recipient's social status (junior) and the relationship between writer and recipient (in-group membership) in the family context. After the identification of the status and relationship, the writer asks about the current state of the recipient with a simple pattern 'Jinlai ni yiqie dou hai hao?' (Has every thing been well with you recently?), as in Example (1), or with a compound structure 'Zan liang hao jiu mei tongxin le, ni ge Jangmian hai hao ba? ' (We haven't been in contact for a long time. Has every thing been ok?) as in Example (2).

The presence of such an inquiry after the greeting 'Nihao' (How are you?) can be seen as additional material to upgrade the identification of the relationship between the writer and the recipient, since the inquiry here obtains expressive function: signaling the writer's kindness and consideration to the recipient, thus reinforcing the interpersonal relationship between them. The effectiveness of the inquiry is due not only to its special phatic function in the opening discourse, but also to its consequential effect: It may elicit the recipient's response with some other phatic expressions in the next correspondence, such as reassurance about oneself, or thanks for such concern. Such a recycling function may be very helpful for enlarging the communication channel between the writer and recipient. This may be the major reason why the combination of greeting with inquiry becomes the most common structure of openings in initial letters.

Combination 2: Greeting + inquiry + other expressive act (e.g. apology/concern/missing/good wishes) The pattern 'greeting + inquiry + other expressive act' occurs the second most frequently in the data (32.7%). Linguistically, it can be seen as an expanded pattern of Combination 1, where another expressive act is added after inquiry. Such a combination can be read as the most elaborate pattern. It presents a sufficient identification of social status and the relationship. Using this pattern, the writer tends to

156 show more attention and involvement in maintaining the relationship with the recipient. Some representative examples are presented as follows:

Example (3) (son to parent) (1) Jianxin hao! see letter good

(2) Jinlai shenti jian kang, shenghuo yukuai, yiqie douhao ba? recently body healthy life happy everything all good (isn't it)

(3) Zhuanyan lijia yi wuge yue le, in the blink of an eye leave family already five months (particle)

wo zhai zheli feichang xiangnian nimen. I in here much miss you Hello to you when you're reading this letter! Have you been well and happy? Has everything been ok recently? It has been five months since I left the family. I am missing you very much here.

Example (4) (younger sister to elder brother) (1) Ni hao. (2) Jinlai shenti kehao, gongzuo shunli ba? you good recent body good work smooth (isn't it)

(3) Hao chang shijian mei gei ge qu xin, qing ge yuanliang. good long time not give elder brother send letter please elder brother forgive How are you? How has your health been recently? Has your work gone well? I haven 't written to you for a long time, elder brother, please forgive me.

One apparent feature from these two examples is that the structure of this combined opening interweaves social norms and individual personal feelings according to particular circumstances. Constructing an opening in this way reflects the writer's perception of the need to increase the phatic force of openings by expressing a specific feeling in addition to the identification of the status and the relationship (e.g. by using greetings, inquiries).

As shown in Example (3), when the writer uses the greeting 'Jianxin hao. '(Hello to you when you are reading this letter.) and inquiry 'Jinlai shenti jiankang, shenghuo yikuai, yiqie dou hao ba?' (Have you been well and happy? Has everything been ok recently?), the indication of the recipient's senior status and a close relationship between the two are spelt out. However, from the writer's view, the use ofthese two strategies alone is still not sufficient to share his attitudes or feelings with the recipient based on the

157 personal circumstance (i.e. departing from home for a period of time). He, therefore, performs an extra expressive act: adding an expression of missing. In doing so, the writer further claims common ground with the recipient, emphasizing that the writer and recipient "both belong to some set of persons who share specific wants" (Brown & Levinson 1987: 103). Thus, the extra expressive act added in the current opening helps increases the primary force of affirming and enhancing the relationship between the two. In Example (4), the writer adds an apology for failing to write earlier after the combination of greetings and inquiries. This expressive act (i.e. acknowledging the offensive action, and requesting forgiveness) can be seen as a further claim of the writer's attention to the recipient in accordance with their relationship, displaying a close relationship.

Interestingly, the added expressive act in this combination pattern is often optional, and the choice of topics seems to depend on the relationship between the writer and recipient and a specific circumstance they are engaged in. Usually, the topics available for this combination include expressions of missing (10 out of 36), of good wishes (8 out of 43), of apologies (8 out of 43), of concern (5 out 43), of reassurance about oneself (3 out of 43), and of thanks (2 out 43).

It is worth noting that usually the expressive act is located after the combination of the greeting with the inquiry. But occasionally (3 out of 36), the expressive act, such as expressing missing (2 cases) or expressing good wishes (1 case), can be inserted between the greeting and inquiry, as exemplified in the following example:

Example (5) (friend to friend) (1) Ni hao. (2) hao jiu meiyou nide xiaoxi, wo hen xiang ni. you good good long time not have your news I very miss you

(3) Xianzai ni YJqie hai hao rna? Gongzuo shifou hen mang? now you everything still good (is it true) work whether very busy How are you? There has been no news from you for a long time, and I am missing you very much. Are you well in everything now? Are you busy at work?

158 The violation of the normal sequence may reflect a specific creative style by the writer for the sake of making an emphatic effect. By inserting the expression of missing between the greeting and inquiry, the writer shows her eagerness for sharing personal feelings with the recipient before the identification of the relationship, which is expressed through the inquiry.

Combination 3: Greeting+ other expressive act (e.g. apology/concern/missing/good wishes) The third common combination pattern for openings is the greeting followed by an expressive act, which accounts for 22.4% of the total combinations. In this structure, after using the greeting, there appears to be an opportunity for the writer to upgrade the social identification of the recipient by expressing a specific personal feeling or attitude. The attachment of an expressive act here may have the same function as other 'upgraders,' such as inquiries in Combination 1, serving to share more common ground and membership. The linguistic features of this combination are illustrated in the following examples:

Example (6) (student to teacher) (1) Nin hao! you good

(2) Youyu wo hui beijing laolao jia danwu le, SUOYJ. xianzai cai gei because I back beijing grandma home delay (particle) so now just give

ni xiexin, jin qing yuanliang. you write letter respect please forgive How are you? Because I was held up by going back to my grandma's home in Beijing, I delayed writing you earlier, please forgive me.

Example (7) (niece to aunt and uncle) (1) Nimen hao. you good

(2) Shouxian zhu mmen shenti jiankang, wangshi ruyi. At first wish you body health everything smooth How are you? Firstly, I wish you good health and no trouble in anything.

159 From these two examples, it can be seen that the personal identification and a display of relationship would be insufficient if the greeting phrase 'Ninhao' (honorific, singular) or 'Nimenhao' (familiar, plural) were used alo:q.e. The writers therefore need to denote a kind of emotion or attitude to the recipient by adding some affective speech acts with respect to the relationship. In Example (6), after sending a greeting to the recipient with the honorific 'How are you?' pattern, the writer immediately presents an apology for the delay in writing back to the recipient. This apology reveals the writer's orientation to the renegotiation of the relationship, which might have been affected by the face threatening failure in writing a letter earlier. Such an addition of the apology after the greeting seems aimed at conveying the writer's respect and attention to the recipient, in order to restore harmony.

In Example (7), after the honorific greeting, the writer expresses her wishes to the recipients 'At first, I wish you good health and no trouble in anything.' The good wishes here can be seen as a supplementary material to the greeting 'Nimenhao' (How are you?) since this addition sends concrete wishes relating to the recipient's health and work. By doing so, the writer shows that she is still sharing membership in a kinship with the recipients (aunt and uncle) and still concerned about their situation. In this sense, the expression of good wishes can be understood to help lighten up identification of the recipients' status and relationship.

The commonly adopted expressive act in this combination is apologies (10 out of 25), and others include expressions of missing (6 out of 25), expressions of good wishes (5 out of25) and expressions ofhope (4 out of25). The different choice oftopics seems to denote that a number of options are accessible to the writers and these options may enable them to manipulate the selected expressions to affect the phatic force of the openings, based on the relationship with the recipients and particular personal circumstances.

Combination 4: Inquiry+ other expressive act (e.g. hope/missing) This is the least commonly used combination pattern, accounting for 6.5% of the total cases. It starts the opening with an inquiry, and then completes it with an expressive act. Using an inquiry as the initiator can be seen as a special way of beginning the opening,

160 in which a writer focuses on the identification of the relationship with the recipient, rather on that of the recipient's status. The following expressive act serves to intensify the display of the relationship:

Example (8) (parent to son) (1) Kuai dao qime kaoshi jieduan le, xuexi mang ba? soon arrive final examinations period (particle) study busy (isn't it)

(2) Xiwang ni yonggong yixie, zhenqu kaochu hao chenji. hope you work hard a bit try get good mark. The final examinations are approaching. Have you been busy studying? I hope you are studying hard and trying to get good marks.

Example (9) (friend to friend) (1) Jinlai )'lqie douhao? (2) Zhenshi xiang ni! recently everything all good really miss you Has everything been ok recently? I am really missing you!

In Example (8), the writer uses an inquiry 'Kuai dao qimo kaoshi jieduan le, xuexi mang ba?' (The final examinations are approaching. Have you been busy studying?) to start the opening, through which he firstly claims a close relationship with the recipient by showing concern about his study. Then the writer provides a hope in the next part shaped as 'Xiwang ni yonggong yixie, zhenqu kaochu hao chenji. ' (I hope you are studying hard and trying to get good marks.). The expression of hope here serves to exhibit the writer's greater attention and involvement. The inquiry in this combination also appears to act as a preface. It is used to invite the writer's next expressive act, which is the main concern of the writer. The expression of hope can be seen as a supplement, since it exhibits the writer's thoughts or attitudes elicited from the preceded inquiry.

In Example (9), after expressing the inquiry 'Jinlai yiqie dou hao?' (Has everything been ok recently?), the writer presents a feeling of missing 'Zhenshi xiang ni. ' (I'm really missing you!). The inquiry here functions as a preliminary, providing a background about their relationship (possibly good friends) and their contacting situations (probably they have departed for a period of time) to draw out the writer's next act, an expression of missing feeling. The display of the thinking over can be

161 interpreted as a natural flow of the writer's emotion or attitudes to the recipient resulting from the previous preface, serving to convey more affection to the recipient, which is helpful for enhancing the emotional bond between the two.

From observing these two examples, it can be suggested that this combination is a very personal and intimate opening pattern. By using it, the writer tends to give strong cues on the relationship at the first stage of the opening behaviour. The data indicate that this pattern is restricted to parent-child role relationships (4 out of 7) and few friend-friend contexts (3 out of 7). Another feature is that the topics for the expressive act seem to be limited to a few kinds (e.g. expressing hope and expressing missing.)

5.2.2.2 In reply letters Three combination patterns of openings are identified from reply letters in the data. They are 'greeting + acknowledgement of receipt of a previous letter+ other expressive act', 'greeting + acknowledgement of receipt of a previous letter', and 'acknowledgement of receipt of a previous letter + other expressive act.' As shown in Table 5.6, Combination 1 occurs the most commonly among the three patterns in reply letters (70.1 %). The detailed examination of each type of combination is provided below:

Table 5 .6. Combination patterns of openings and their distribution in reply letters (N=107) Combination Patterns N % I. Greeting + acknowledgement of receipt of a previous letter + other expressive 75 70.1 act (e.g. apology/concern/happiness) 2. Acknowledgement of receipt of a previous letter+ other expressive act (e.g. 18 16.8 apology/ concern/happiness) 3. Greeting+ acknowledgement of receipt of a previous letter 14 13.1

Total 107 100

Combination 1: Greeting + acknowledgement of receipt of a previous letter + other expressive act (e.g. apology/concern/happiness)

This pattern is composed of three components: greeting, acknowledgement of the receipt of a previous letter, and other expressive act. This is the most intense combination pattern of openings in reply letters, and it serves to express the writer's

162 feelings to the recipient to a maximal extent. Some examples are presented to illustrate this pattern.

Example (1 0) (father to son) (1) Ni hao, you good

(2) Lai xin y1 shoudao. come letter already receive

(3) Zhidao ni hen hao, women hen gaoxin. know you very good we very happy How are you? Your letter has been received. Knowing you are all right in everything, we are very happy.

Example (11) (friend to friend) (1) Jin hao. recent good

(2) Nide lai xin shoudao. your come letter receive (3) Hen baoqian wei nengjishi ge1 m hui xm. very apologetic not able on time give you back letter Greeting! Your letter has been received. I am very sorry for not having written back to you on time.

In these two examples, the writers start the openings by a greeting either with the familiar 'Nihao' (How are you?) or a conventional formula 'Jinhao ' (greeting). Then, the greeting is followed by a simple acknowledgement of the receipt of a previous letter, such as 'Laixin yi shoudao. ' (The letter has been received). The addition of the acknowledgement can be understood as extra effort in displaying the relationship, since such an announcement serves to satisfy the recipient's expectation. Thus, a relationship between the writer and recipient is implicated: They are still sharing a common membership in a social group.

After the acknowledgement, an expressive act appears in the sequence to let the writer express feelings or attitudes (e.g. showing happiness, or apologies for not having written earlier) to the recipient. The expressive act can be seen as a sequential consequence in relation to the instance of receiving the letter. Its occurrence gives the writer an opportunity to exhibit some positive feelings (happiness) or affection (apologies) to the

163 recipients, further confirming or consolidating the relationship. In Example (1 0), when a father writes 'Knowing you are all right in everything, we are very happy,' he tends to explicitly express his evaluation of his emotional state which is directed to the fact 'your letter has been received,' thus claiming common ground and emphasizing in group membership and affection. Similarly, when a friend in Example (11) expresses '(I am very sorry for not having written back to you on time,' he tends to denote his attention to his friends, claiming and reinforcing the friendship between them. In both cases, the expression of feelings or attitudes by the writers can be understood as a way of strengthening the initial phatic function displayed by the announcement of arrival of a previous letter. It may lead to a sufficient achievement of identifying the relationship between the writer and recipient.

As showed in the data, there are some expressive acts which can be attached to the acknowledgement. The most common ones are expressions of happiness (30 out of 73) and of apologies (17 out of 73). Other acts which are used infrequently include expressions ofmissing (6 out of63), expressions ofthanking (5 out of73), reassurances about oneself (3 out of 73), and expressions of good wishes (2 out of 73).

As appeared in the data, however, some writers (1 0 out of 73), instead of using an expressive act after the acknowledgement, insert an inquiry between the greeting and the acknowledgement. On the one hand, the insertion of an inquiry serves to focus on the identification of the relationship between writer and recipient. On the other hand, it can be seen as a reflection of the writer's uncertainty about the recipient's current state as shown in the following:

Example (12) (younger brother to elder brother) (1) Ni hao. (2) Jinlai gongzuo shunli shenti ke hao? you good recent work smooth body still good

(3) Lai xin gang shoudao. come letter just received

(4) Nide qingkuang kanqilai bucuo, quanjia yejiu fang xin le. your situation look not bad whole family so put heart (particle)

164 How are you? Has your health and work been ok recently? Your letter has been received. It sounds like everything is ok for you, so everyone in the family is relieved.

As shown in Example (12), after greeting, the writer asks about the recipient's state of health and work before indicating the arrival of the previous letter and the response of the members of the family. In this way, he emphasizes his concern about the recipient. On the other hand, he also conveys some worry about the recipient, although he would be aware of the recipient's state in those aspects from the previous letter.

Combination 2: Acknowledgement of receipt of a previous letter + other expressive act (e.g. apology/concern/happiness) There are 17.8% cases in which the writers initiate the opening by an acknowledgement of the receipt of a letter, and then complete it with another expressive act. By not using the greeting, this combination pattern seems to place a greater emphasis on the identification of the relationship in the beginning of openings. The following examples are given to illustrate this pattern.

Example (13) (father to son) (1) Nide lai xin shoudao le. your come letter receive (particle)

(2) Xin zhong dezhi ni hen mang, wang tebie zhuyi shenti. letter inside get you very busy hope specially attention body Your letter has been received. Knowing from the letter you are very busy, I hope you look after yourself.

Example (14) (friend to friend) (1) Jintian shoudao nide lai xin bei gan qmgq1e. today receive your come letter greatly feel warm

(2) Zaici wo zhuyuan lao penyou huode kuaile haomeng cheng zhen. in here I wish old friend live happy good dream become true I feel so warm from receiving your letter today. I wish my old friend lives a happy life and dreams come true!

In Example (13), when the writer (father) states 'Your letter has been received.' in the initial place, he shows that he identifies the recipient as sharing group membership. Here the writer's attention seems to be on the indication of the relationship, rather than

165 the status of the recipient. After the acknowledgement, the writer exhibits his concern about the writer by saying 'Knowing from the letter you are very busy, I hope you look after yourself.' This expression serves as a consequence resulting from the first strategy. The whole opening can be interpreted in this way: 'I got your letter. After having read it, I know you are very busy studying. I hope you take a good care of yourself.' The writer's addition of the expressive act can be seen as an additional indication the in group membership with the recipient (son).

Similarly, in Example (14), when the writer (friend) starts the opening with 'I feel so warm from receiving your letter today,' he explicitly denotes that he and his friend are enjoying a good friendship. The following expression 'I wish my old friend lives a happy life and dreams come true!' also functions as a consequent feeling resulting from the act of 'knowing something from the letter.' Using such a strategy, the writer brings good fortune to his friend, thus confirming and enhancing the friendship established already.

These findings suggest that initiating the opening with an acknowledgement of the arrival of a letter not only presents the indication of the relationship on its own, but also provides an opportunity for the writer to upgrade the identification with another strategy. There are 19 cases which follow this pattern in the data. Similar to the situation in Combination 1, two major topics serve as the consequential materials in this pattern, including expressions ofhappiness (6 out 19) and apologies (4 out of 19), and the other main topic is expressions of concern (4 out 19).

Also as in the Combination 1, there is a variance emerging from this pattern, that is, in some cases (4 out of 19), an inquiry is used after the acknowledgement of receipt of the letter, as seen in Example (15):

Example (15) (younger sister to elder sister) (1) Hen gaoxin shoudao nide lai xin. very happy receive your come letter.

(2) ni shenti hao rna? gongzuo shunli ba? you body good (is it true) work smooth (isn't it) I'm very happy to receive your letter. Has your health and work been ok?

166 In this example, after acknowledging the receipt of the letter 'I'm very happy to receive your letter,' the writer uses an inquiry 'Has your health and work been ok?' Why does the writer use an inquiry here, rather than an expressive act as in the normal sequence of the pattern? The possible answer is that the inquiry here, similar to other expression acts (e. g. expressing happiness) may serve as a sequential consequence resulting from the previous strategy (i.e. acknowledgement), showing the writer is still thinking about the recipient's state of being, although she was informed in the previous letter. The whole opening may imply that 'I am still concerned about you even though you wrote positively in your previous letter.' The inquiry here functions as a double check to make sure that the recipient is well, thus pragmatically conveying consideration and affection to the recipient.

Another feature also deserves attention for this pattern. The strategy of 'acknowledgment of receipt of a letter' is often shaped by the expressive manner. It usually integrates an offer of factual message with expressing positive feeling, such as 'I'm happy to receive your letter.' As mentioned earlier, expressing happiness or pleasure while acknowledging the arrival of the letter can be seen as an explicit 'expressive acknowledgement.' Using this type of acknowledgement, a writer tends to rapidly spell out his/her feelings about receiving the letter in the short version, producing an effect in greatly identifying common membership that the writer and recipient have had.

Combination 3. Greeting + acknowledgement of receipt of a previous letter This pattern is used the least frequently in reply letters (14%). In this pattern a greeting is used initially and then an acknowledgement of receipt of the letter follows. It is interesting to note that the acknowledgement here is often an 'expressive acknowledgement,' namely shaped by the display of happiness or thanks within one expression. Some examples are presented here to demonstrate this combination:

Example (16) (friend to friend) (1) Ni hao. You good

167 (2) Hen gaoxin shoudao nide lai xin. very happy receive your come letter How are you? I'm very happy to receive your leUer.

Example (17) (younger brother to elder sister) (1) Ni hao. you good

(2) Xiexie ni zai baimang zhizhong lai xm zhidian mtJm. thank you in hundred busy inside come letter point out wrong path How are you? Thank you for writing even when you are busy to point me in the right direction when I am puzzled.

In the two examples, the writers do not just simply acknowledge their receipt of the letters after the greeting. Instead, they express their feelings or attitudes in the same announcement as 'I'm very happy to receive your letter' in Example (16), or 'Thank you for writing even when you are busy to point me in the right direction when I am puzzled" in Example (17). Such an acknowledgement is able to simultaneously perform two functions: conveying some feelings or attitudes (e.g. happiness, gratitude) and announcing the receipt of the letter. Two reasons may be relevant for a writer to choose this opening pattern. First, the writer needs to avoid insufficient support of the relationship by a simply bald acknowledgement if there is no other act following. Second, the writer needs to express feelings relating to a particular situation. A further examination of the data indicates that the users of this pattern have not had contact with the recipients for some period of time. By using this pattern, these writers are able to achieve a quick confirmation of the relationship within a dense opening. Therefore, with 'expressive acknowledgement,' the writer can let the recipient quickly understand a feeling of continuing the connection.

There are 15 cases in the data that adopt the pattern of Combination 3. Among them, most cases use the 'expressive acknowledgement' as an addition after the greeting (11 out of 15). But some cases (4 out of 15) use a 'bald' acknowledgement as in:

Example (18) (father to son) (1) Nimen hao. you good

168 (2) Qian liangfeng xin yiyu liuyue ershi yi ri shoudao. previous two letter already June twenty one day receive How are you? Your two letters were received on the 21st ofJune.

In this example, after the greeting, the writer (father) directly informs the arrival of the previous two letters through the factual expression 'Your two letters were received on the 21 51 of June.' As a phatic expression, the bald acknowledgement here merely indicates that the writer and the recipients are still sharing in-group membership. However, with respect to the display of feelings or attitudes, this kind of acknowledgement seems to be less emotional and affectionate compared with the 'expressive acknowledgement.' The motivation of using bald acknowledgement by the writer (father) here seems to not only come from a search for simplicity but also reflects the writer's intent to show a level of seriousness under this particular role-relationship, that is, authority over the recipient (son).

5.2.3 Discussion Based on the findings regarding the types of opening strategies, there are two points which are of interest. First, these Chinese writers exhibit a greater preference for the use of combined openings compared to the single openings. From a sociological point of view, such a phenomenon can be interpreted as being driven by something similar to the norms of opening sequences in conversation on the level of 'social recognition' (Schiffrin 1977). As discussed in the previous section, written openings are a phatic activity, in which writers aim to establish and strengthen bonds of personal relationship with the recipients. For the sake of being psychologically 'properly involved' (Goffrnan 1963) with the recipient in this behaviour, the writers are required to show identification of the recipient's status and the relationship between the two, even to show an effort of reinforcing the relationship by using some particular opening strategies. Following Schiffrin, a single strategy seems to be neither socially nor ritually sufficient as evidence of personal identification, since the opening activity between acquainted people needs to signify not only the social recognition, but also specific personal feelings to acknowledge "the expectations and obligations" inherent in this activity (1977: 679). Strategies should be, therefore, intensified and efficiently ritualized. This may provide the primary reason why Chinese writers tend to use two or three strategies to form the openings in letters.

169 Chinese writers' favouring of the combined openings also reflects the influence of Grice's conversational maxims on their opening behaviour. The findings indicate that these Chinese writers exploit some combined strategies to achieve the effect of Relevance and Clarity, as proposed by Grice (1975), in their opening section. Take Combination pattern 4 (inquiry + other expressive act) for initial letter as an example. The appearance of a certain expressive act after the inquiry can be seen as the sequential consequences resulting from the previous strategy (i.e. hope is given based on the writer's concern about the recipient's situation indicated in inquiries). In doing so, the writers tend to "increase the importance of the elements focused" (Lee-Wong 2000: 111) for "facilitating the management of interpersonal relationship" (Pavlidou 2000: 138). This result suggests that while performing an opening, the phatic event, Chinese writers attend to the relationship aspect of communication not only by doing something (i.e. choosing necessary strategies), but also by doing something logically (i.e. choosing strategies related to each other). This is an interesting finding, as it assumes that the maxims of Relevance and Clarity (Grice 1975) are not only applicable to the conversation that has a clear referential content, but also to some speech acts bearing phatic functions in a particular speech behaviour.

The predominant employment of combined opening in Chinese personal letters can be also seen as the writer's effort in manifesting politeness. According to Brown and Levinson (1987: 61), politeness refers to attention to face needs. Thus, it assumes that the combined openings are the more polite ones, since they take more account of face needs of the recipient. By selecting more strategies rather than the single one, the writers are able to share more common ground and experience with the recipients, by contributing more feelings or politeness. It would be helpful in making the recipients "feel good" (Lakoff 1973: 298) as to the continuation of the relationship between them. There is a saying in Chinese that 'Li duo ren bu guai, ' which can be translated as 'People do not object to anyone being over-polite' (Lee-Wong 2000: 255).

Combination 1 (e.g. greeting+ acknowledgement+ expression of apologies) in reply letters is another example to illustrate this point. In general, this combination contains both positive and negative face-saving strategies. By using greeting strategy, the writer attends to the recipient's positive face, since a specific form of greeting indicates that

170 the recipient's social identity is accepted by the writer. The announcement of the arrival of the letter is to satisfy the recipient's expectation by socially affirming social worth, which forms another positive politeness strategy. Then apologizing for not having written back earlier can be seen as a negative politeness strategy since it is primarily aimed at face redress associated with the offence of not writing earlier (Brown & Levinson 1987: 189). Adopting these strategies reflects the writer's obedience to the social maxim to be polite (Laver 1981: 296). This finding is consistent with Shih's claim that politeness should not be merely viewed as a conflict avoidance strategy, but also positively as a harmony and friendship enforcement strategy (1986: 22-23).

The second point of interest is that Chinese writers use a systematic way in which they organize the opening structures. These structural patterns can be summarized in Figure 5.1 for initial letters and in Figure 5.2 for reply letters.

Inquiry

Greeting Inquiry •I Expressive act

Expressive act

Inquiry •I Expressive act

Figure 5.1. Summary diagram of the sequential combination of strategies of opening in initial letters

171 Acknowledgement I Expressive act of receipt of a letter I J ,____. rl Greeting Acknowledgement of receipt of a letter

Acknowledgement -----+ J of receipt of a letter I Expressive act I

Figure 5.2. Summary diagram of the sequential combination of strategies of opening in reply letters

As showed in Figures 5.1 and 5.2, combined openings in both initial and reply letters contain two major linguistic components: an initial resource, which indicates the status and or the relationship between the writer and recipient; and an upgrading resource, which is used to increase identification of the status and the relationship. Based on these two figures, a discussion on how the opening sequences are developed is presented below.

The first comes with the combination patterns of openings in initial letters. Two strategies can act as initial resources in initial letters: greetings and inquiries. The greeting is almost an obligatory component for initiating an opening. In using a greeting, the writer puts the focus on answering the question 'to whom the writer is writing?' Compared with the greeting strategy, inquiry is a less preferred initiator. It is used only in the case when the writer is engaged in dealing with an intimate recipient. Using this strategy, the writer's focus is on identifying the relationship with the recipient (e.g. in-group membership) rather than on the status of the recipient. The initial sources in initial letters basically function to establish 'ritual accessibility' (Schiffrin 1977: 679) for the whole opening discourse, and to allow the recipient to accept the "initiation of an engagement" (Goffman 1963: 113).

Three options are available for the upgrading resources: an inquiry alone; a combination of inquiry with another expressive act (e.g. inquiry + expressing missing); and

172 expressive act alone (e.g. apologies, expressions of missing). The first two options are added after the greeting. An inquiry alone is the most commonly used 'upgrader' in the combination, and it serves to further indicate the relationship by conveying concern to the recipient. An inquiry plus another expressive act can be seen as an intense sequential pattern, which reflects a writer's perception of the need of integrating the social norm with individual personal feelings. The last option (expressive act alone) can be placed either after the greeting or the inquiry. The expressive act here can be understood as the flow of a specific personal feeling, which is characterized as an expansion of identification of status and relationships. It is worth noting that when an expressive act is added after the inquiry functioning as the initiator of openings, it has sensitivity to topics since it carries a consequence in the next speech act.

As for the combination patterns of openings in reply letters, the findings show that the greeting and the acknowledgement of receipt of a previous letter are two common resources that are used to initiate an opening sequence. As in initial letters, the greeting is overwhelmingly preferred by writers as a means to identify status and relationship. In reply letters, acknowledging receipt of a letter can also function as an initiator for an opening, although this usage is much less preferred than the greeting. It focuses on conveying a share of membership in a social group, and also acts as a basic sequential unit which provides an opportunity for a writer to express relevant feelings or attitudes in the next act.

There are also three options available for upgrading resources: a combination of acknowledgement with another expressive act; acknowledgement alone; or another expressive act alone. The first two options are used after the greeting. Among them acknowledgement plus an expressive act is dominant, which serves to achieve sufficiency in indication of the relationship by acknowledging in-group membership and showing a specific personal feeling. The acknowledgement alone is preferred when the writer simultaneously indicates the receipt of the letter and expresses some positive feelings within one strategy. This kind of 'upgrader' may exhibit the writer's eagerness and rapidity in confirming the relationship. The last option (expressive act alone) is only used after the acknowledgement. It functions to express the writer's specific

173 feelings or attitudes referring back to the acknowledgement of the letter. It can be seen as a relevant part to the acknowledgement.

To sum up, while performing an opening sequence of personal letters, the first task facing Chinese writers is to identify the recipient's status and the relationship between them with a greeting (unless the writer's focus is on the display of the relationship immediately, rather than the status). After the social identification is completed, the writer comes to the procedure of upgrading the relationship by means of various options depending on particular circumstances.

5.3 Openings and social variables Following the discussion of various patterns of linguistic realization for openings, it is now time to explore the effects of various social factors on the opening behaviour. This section addresses the relationship between social roles and opening types, namely, the single and combined strategy.

5.3.1 Openings and social roles Table 5. 7. Frequency and distribution of opening types across social roles (N=259) Types Social roles Total Family Non-family Parent Child Sibling Relative Friend Teacher Student (26) (18) (48) (16) (115) (10) (26) (259) 1. Single openings 7 2 10 1 19 3 3 45 (26.9) (11.1) (20.8) (6.3) (16.5) (30) (11.5) (17.4) 2. Combined 19 16 38 15 96 7 23 214 openings (73.1) (88.9) (79.2) (93.8) (83.5) (70) (88.5) (82.6) Total 26 18 48 16 115 10 26 259 (100) (100) (100) (100) (100) (100) (100) (100)

As seen in Table 5.7, each social role in the data exhibits a great preference for the use of combined openings, but a close examination indicates that the degree of favour of these two types varies with different roles in both family and non-family letters. The data show that the use of the single opening is highest for parents and teachers (26.9%, 30%), whilst children and students use this type the least commonly (11.1% and 11.5% respectively). This means that the single opening appears to be more appropriate for seniors than for juniors in both family and non-family contexts.

174 5.3.1.1 Single openings and social roles Apart from the different preferences for the single opening across social roles, the choice of a specific strategy also differs among these roles. As shown in Table 5.8, the greeting is most appropriate for children (100%) and students (100%) as a single opening. It is also extensively used by teachers (66.7%), friends (52.6%), and parents (42.9%). It may suggest that while formulating the opening with a single strategy, both seniors and juniors are inclined to choose a strategy which best reflects the recipient's status and relationship with the recipient. Inquiry alone is found to be used more often between siblings (60% ), particularly by elder siblings (4 out of 6). However, it is worth noting that the findings regarding the influence of social roles on the choice of the single opening cannot be viewed as conclusive, but as suggestive, since the sample numbers of this type in the data are rather small.

Table 5.8. Frequency and distribution of the strategy of single openings across social roles (N=45) Strategies Social roles Total Family Non-family Parent Child Sibling Relative Friend Teacher Student (7) (2) (10) (1) (19) (3) (3) (45) 1. Greetings 3 2 4 0 10 2 3 24 (42.9) (100) (40) (0) (52.6) (66.7) (100) (53.3) 2. Inquiries 2 0 6 0 4 0 0 12 (28.6) (0) (60) (0) (21.1) (0) (0) (26.7)

3 .Acknowledge- 1 0 0 1 2 1 0 5 ment of receipt (14.3) (0) (0) (100) (10.5) (33.3) (0) (11.1) of a previous letter 4. Apologies 0 0 0 0 2 0 0 2 _( 0) (0_2 ( 0) ( 0) (10.5) (0) ( 0) (4.4) 5. Expressions 0 0 0 0 1 0 0 1 of good (0) (0) ( 0) (0) (5.3) (0) (0) (2.2) wishes 6. Expressions 1 0 0 0 0 0 0 1 of missing (14.3) (0) ( 0) ( 0) (0) (0) (0) (2.2) Total 7 2 10 1 19 3 3 45 (100) (100) (100) (100) (100) ( 100) (100) (100)

5.3.1.2 Combined openings and social roles As discussed earlier, combined openings present different structure patterns in initial and reply letters. Thus, the influence of social roles on the choice of the combination patterns needs to be addressed separately.

175 5.3.1.2.1 In initial letters The findings shown in Table 5.9 reveal that parents prefer to use Combination 4 (75%), which is initiated by an inquiry about their children's general state, and followed with an expressive act such as expression of hope (2 out of 3) and expression of missing (1 out of 3). Children show a strong preference for Combination 2, the most intense or elaborate combination (63.6%). This is notably realized by the honorific form of 'How are you?' ( 6 out 7: honorific singular: 2; honorific plural: 1; or familiar plural: 3), then a simple inquiry, and then an expression of some specific personal feelings (e.g. expressions of missing: 4 out of 7; expressions of concern: 1 out of 7, reassurances about themselves: 1 out of7; apologies: 1 out of7).

Table 5.9. Frequency and distribution of combination patterns of openings across social roles in initial letters (N=107) Combination Social roles Total patterns Family Non-family Parent Child Sibling Relative Friend Teacher Student (4) (11) (19) (7) (48) (3) (15) (107) 1. Greeting + I 0 11 1 22 1 5 41 inquiry (25) (0) (57.9) (14.3) (45.8) (33.3) (33.3) (38.3) 2. Greeting+ 0 7 8 2 16 0 2 35 inquiry + other (%) (63.6) (42.1) (28.6) (33.3) (0) ( 13 .3) (32.7) expressive act 3. Greeting + 0 4 0 4 6 2 8 24 other (0) (36.4) (0) (57.1) (12.5) (66.7) (53.3) (22.4) expressive act 4. Inquiry+ other 3 0 0 0 4 0 0 7 expressive act (75) (0) (0) (0) (8.3) (0) (0) (6.5) Total 4 11 19 7 48 3 15 107 (100) (100) (100) (100) (100) (100) (100) (100)

Siblings and friends exhibit a preference for Combination 1 'greeting+ inquiry' (57.9%, 45.8%). While realizing this combination, both siblings and friends may choose the familiar form of 'How are you?' to greet the recipient, and use the simple (50%) or complex (50%) inquiries by focusing on the issues of health (for both) and life (for siblings) or work and study (for friends). Thus, the reciprocal use of the familiar 'How are you?' pattern and the concrete inquiries functions to negotiate a great solidarity, emphasizing closeness between siblings or a higher familiarity between friends.

The findings show that relatives, teachers and students prefer to use Combination 3 (57.1 %, 66.7% and 53.3% respectively). It is worth noting that, although their overall

176 preference for this combination type IS similar, the strategies adopted within the structure vary with these three roles. The data demonstrate that relatives show an inclination for selecting the personal-oriented expression (e.g. expressions of missing) (2 out of 4), while teachers and students show equal preferences for the social acts (expressing apologies: 1 out of 2 for teacher; 6 out of 8 for students; expressing wishes: 1 out of 2 for teachers; 2 out of 6 for students). For the relatives, the favouring of the personal-oriented strategy reflects their concern of emphasizing in-groupness, in order to maintain a close kin relationship with their relatives. For the teachers and students, the preference of the social acts shows their attention to social concern and to avoid conveying specific personal feelings (e.g. expressing missing or concern), aimed at maintaining the rank order and a certain social distance with the recipients.

The data also reveals that the linguistic patterns of strategies vary in this combination depending on the status and relationship across role-relationships and even between the social equals. For instance, when relatives construct the opening with the greeting and expressions of missing, the older ones often use the familiar form of 'How are you?' and adds a simple expression of missing like 'Nihao! shu shen dou xiangni!' (How are you? Your uncle and aunt miss you!), while the younger ones often use the honorific form of 'How are you?', then use a complex pattern of expressing missing like 'Ninhao! nin yijing zoule liangnian le, wo hen xiangnian nin!' (How are you? You have been away for two years. I miss you very much!).

When teachers and students realize this combination with the same strategies (the greeting and expressions of apology), the choice of linguistic patterns of the same strategy would be different. For instance, teachers often uses the familiar form of 'How are you?' and express an apology by involving an admission of the offence and an explanation without requesting forgiveness, such as 'Nihao, lai shanghai hou mei gen ni jizao lianxi, zhuyao shi wo yizhi mei you wending xialai. ' (How are you? I haven't contacted you earlier since arriving at Shanghai, because I haven't settled down.) But students often use the honorific forms of 'How are you?' and an apology including a provision of an explanation, an admission of the offence, and then a request for forgiveness, such as 'Youyu wo hui Beijing laolao jia danwu le, suoyi xianzai cai gei ni xiexin, jinqing yuanlang. '(Because I was held up by going back to my grandma's horne

177 in Beijing, I delayed writing you earlier, please forgive me.) By choosing the honorific form of 'How are you?' and the elaborate apologies that take more account of face needs of the recipient (Holmes 1990: 177), the students convey a greater deference and imply a level of distance as well. But using the familiar form of 'How are you?' and the apology without a request for forgiveness, the teachers are likely to simultaneously emphasize a greater familiarity and distance.

5.3.1.2.2 In reply letters As in initial letters, the choice of the combination patterns in reply letters also varies with different social roles. From Table 5.1 0, it can be seen that Combination 1 is appropriate for children, siblings, relatives, friends, teachers, and students. It seems that almost all the social roles are inclined to choose this combination pattern except parents. However, the findings show that, although Combination 1 is chosen widely across the roles, the choice of the specific expressive acts following the acknowledgement of the receipt of a letters in this structure seems to distinguish each role from another. For instance, children often use reassurance about themselves (2 out of 4); siblings and friends prefer to choose happiness (5 out of 13; 20 out of 41), apologies (5 out of 13; 7 out of 41) occasionally congratulations (1 out of 13; 2 out of 41); relatives like to use an expression of happiness (2 out 4), or of good wishes (2 out of 4); teachers choose thanking (2 out of 3) and once apologies (1 out of 3); and students overwhelmingly use apologies (3 out of 5) and thanking (2 out of 5). It appears that intimates in both family and non-family show more personal-orientation in the choice of expressive act (e.g. expressing happiness, congratulating), while distant writers seem to be more social oriented (e.g. expressing apologies, thanking).

The data indicate that Combination 2 is appropriate for parents. This finding suggests that in choosing the combination pattern of openings seniors in a direct-blood relationship exhibit a more conservative way, that is, 'undergreet' as labeled by Hopper and Chen (1996: 307). Through this way, they imply claims of both certainty of recognition of the recipient's status and of the relationship state with juniors. The data also show that the topics of expressive acts used by parents are often centered on expressing happiness (3 out 7), or expressing hope (2 out of7).

178 Table 5.1 0. Frequency and distribution of combination patterns of openings across social roles in reply letters (N=107) Combination Social roles Total patterns Family Non-family Parent Child Sibling Relative Friend Teacher Student (15) (5) (19) (8) (48) (4) (8) (107) . 1. Greeting + 5 4 13 4 41 3 5 75 acknowledge- (33.3) (80) (68.4) (50) (85.4) (75) (62.5) (70.1) ment of receipt of a previous letter + other expressive act 2. Acknowledge- 7 0 5 2 4 0 0 18 ment of receipt (46.7) (0) (26.3) (25) (8.3) (0) (0) (16.8) of a previous letter + other expressive act 3. Greeting + 3 1 1 2 3 1 3 14 acknowledge- (20) (20) (5.3) (25) (6.3) (25) (37.5) (13.1) ment of receipt of a previous letter Total 15 5 19 8 48 4 8 107 (100) (100) {100) (100) (100) (100) {100)_ (100)

More interestingly, no case of Combination 2 is found for child, teacher, or student groups in the data. This phenomenon could be explained by the suggestion that this combination pattern is not appropriate to use by a junior towards senior in family setting, and between seniors and juniors in non-family setting. If a junior in either context uses this combination pattern without greeting, it would lead to an interpretation of impoliteness towards the seniors. The use of such a pattern by seniors in a non-family setting would mean of excessive informality.

5.3.1.3 Discussion Based on the findings described in the above, it can be suggested that social roles do influence the choice of the combination patterns of openings. First, for the asymmetrical role-relationships within the family setting, seniors (excluding older relatives) exhibit a preference for emphasizing authority and intimacy, rather than a display of social recognition of the recipient's status. Juniors (excluding younger relatives) show a strong preference for using the most elaborate opening pattern, explicitly emphasizing respect to the seniors' higher status and the emotional bond with them. Second, for the symmetrical role-relationships (in both family and non-family),

179 the intimate and social equals are found to follow the highly conventionalized combination patterns in the choice of combination patterns. This marks their obedience to the regular course of the opening procedure, and also stresses solidarity and equality. Third, for the asymmetrical role-relationships in the non-family setting, both seniors and juniors' preference focuses either on the recipient status-then-individual feeling-oriented or on the commonly accepted pattern, although there are some variations in the choice of the specific strategy (mainly in choosing the expressive acts) between them.

Such different preferences for the combination patterns among the social roles in the family and non-family letters are related to the influence of authority of the writer and the recipient in age or rank in people's opening behaviour, which is shaped by the Chinese cultural values and beliefs in verbal interaction to a considerable extent. As is discussed in Chapter 4, in the Confucian tradition, Chinese attitudes towards authority is based on father-son relationships, that is the father plays an important role in providing a strict and personalized source of guidance for his son, while the son's verbal behaviors are characterized as a mixture of respect, dependence and affection (Ebrey 1991: 46). The most salient feature of a father's authority is its "rich affective display" (Blum Kulka 1990: 265). Under this influence, therefore, it would be likely for parents to emphasize consideration (personalized) and involvement (authoritative), rather than emphasize their children's social identity, and for the children to convey sufficient social message to show respect, in-groupness, and affection to their parents in the event of openings.

However, the authority based on teacher-student relationships is different from that embedded in a father-son relationship (e.g. care-taking, socializing children). Apparent from the data, the authority between teachers and students seems to involve less affection due to the inherent social distance in a general sense. It is the respectively pure status-oriented authority that enables these social seniors and juniors to equally emphasize both social recognition and social concern.

For the family and social equals, their opening performance does not seem to be consistent with the conventional assumption that Chinese people tend to submit to authority in age or rank, although there exist differences in ages. Brotherliness and

180 friendship seem to override authority in this aspect. These intimate or social equals seem to mutually accept the common social routines. Their valuing of conformity and solidarity is high, but individual creativeness to show differences in age is low. These would be simple reasons why siblings and friends reciprocally prefer the commonly accepted combination types of openings.

It is important to note that social roles integrate the degree of intimacy to influence the choice of the opening patterns. This is typically reflected in the different performance of seniors between the direct-blood (parents) and indirect-blood (uncle, aunt) relationships in the family context. As evident from the findings, for parents, a display of the relationship is intermingled with expressions of affection (e.g. inquiry + expressions of hope; acknowledgement + expressions of happiness). For senior relatives, stressing social recognition is integrated with reaffirming the relationship (e.g. greeting + expressions of good wishes, greeting + acknowledgement + expressions of happiness). The parents' style of openings seems to be informal, which serves to minimize the distance to the least, while the senior relatives' structure appears to be relatively formal, which denotes a level of distance.

This dissimilarity can be seen as a reflection of differentiations in the degree of intimacy in the relationships between father-son and uncle-nephew in the Chinese society. Tian claims that the importance of the kin relationships in ancient China was ordered according to the following: (1) vertical direct blood relationships (nuclear family including father-son, siblings; extended family including grandfather-grandson), (2) horizontal indirect blood relationships (extended family including uncle-nephew), and (3) marital relationships (1996: 222). According to Tian, among Chinese kinship relations, the father-son relationship is dominant, and this kind of relationship is the most intimate (1996: 228). Influenced by this cultural belief, opening patterns adopted by Chinese parents towards their children may show more care-taking and more socializing characteristics, thus more distance-minimizing than that by senior relatives.

The social roles not only impose a significant influence on the choice of the combination patterns, they also affect the linguistic realization even when the same combination type is adopted. This effect is more evident for the relationships between

181 older and younger relatives and between teachers and students. In general, the seniors draw on the language patterns that are characterized as showing in-groupness or familiar, emotional or distancing depending on the relationship contexts. In contrast, juniors prefer to use respectfui patterns in order to convey deference, affection and distance.

The differences in the choice of linguistic patterns to form specific strategies embedded in the combined openings, as with other routine behaviours, can also be seen as a reflection ofthe influence of Confucian ethics 'li' (propriety) in Chinese people's verbal interactions. Repeatedly, while performing a speech behaviour, the addresser's syntactic and lexical choice must be in accordance with the status of the participants. As argued by Lee-Wong, "it is not a personal decision or choice but one based on what is seen to be socially acceptable-appropriacy" (2000: 171).

5.3.2 Openings and age/generation This section investigates the relationship between the opening behaviour and age/generation factor in Chinese personal letters. It aims at finding out to what extent this social factor affects a writer's opening performance in terms of choosing types and patterning the structures.

5.3.2.1 Opening types and age/generation As shown in Table 5.11, there is a difference in the adoption of these two types of openings between the different and same generations in both family and non-family letters. Within the different generation, older writers use single openings more than the younger (family: 23.5% vs. 9.5; non-family: 30% vs. 11.5%). The adoption seems to be a little more complicated within the same generations in these two settings. In the family letters elder and younger writers show almost the same tendency (single: 19.2% vs. 18.5%; combined: 80.8% vs. 81.5%) in choosing the single or combined openings, while in the non-family letters younger writers use the single strategy less commonly compared with the elder and same-age groups (9 .1 %, 13.1 %, 23.3% respectively). It should be noted that this result must be taken as suggestive, as the evidence is scanty.

182 Table 5.11 Frequency and distribution of opening types across ages/generation (N=259) Types Age/Generation Total

Family Non-family DG: 55 SG: 53 DG: 36 SG: 115 0-Y Y-0 0-Y Y-0 0-Y Y-0 0-Y Y-0 SA (34) (21) (26) (27) (10) (26) (61) (11) (43) 259 1. Single 8 2 5 5 3 3 8 1 10 45 openings (23.5) (9.5) (19.2) (18.5) (30) ( 11.5) (13.1) (9.1) (23.3) (17.4) 2.Combined 26 19 21 22 7 23 53 10 33 214 openings (76.5) (90.5) (80.8) (81.5) (70) (88.5) (86.9) (90.9) (76.7) (82.6) Total 34 21 26 27 10 26 61 11 43 259 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: d1fferent generahon, SG: same generation; 0-Y: older to younger; Y-0: younger to older; SA: same age

5.3.2.2 Combined openings and age/generation Due to the limited samples with regards to the choice of single strategy across different ages/generations, the examination mainly focuses on the sequential organization of openings in this respect. In general, age appears to be a factor determining the choices of the structure of openings more or less in different situations between family and non family letters.

5.3.2.2.1 In initial letters As shown in Table 5.12, a significant correlation exists between combination patterns and writers' ages across the different generations in the family setting: older writers exhibit a preference for Combination 4 (60%), while younger writers show a preference for Combination 2 (57.1%). In the non-family setting this correlation still exists, but is not that significant: both the older and younger show a preference for Combination 3 (66.6% and 53.3% respectively). However, within the same generation, age differences appear to be less related or not related at all to the choice of the combination patterns in both family and non-family letters: all writers with both different ages and same ages favour the use of Combination 1 (family: 60%, 50%; non-family 45.8%, 50%, 45.5%).

183 Table 5.12. Frequency and distribution of combination patterns of openings across age/generation in initial letters (N=107) Combination Age/Generation Total patterns Family Non-family DG: 19 SG: 22 DG: 18 SG: 48 E-Y Y-E E-Y Y-E E-Y Y-E E-Y Y-E SA (5) (14) (10) (12) (3) (15) (24) (2) (22) 107 1. Greeting + 1 0 6 6 1 5 11 1 10 41 inquiry (20) (0) (60) (50) (33.3) (33.3) (45.8) (50) (45.5) (38.3) 2. Greeting + 0 8 4 5 0 2 9 1 6 35 inquiry+ (0) (57.1) (40) (41.7) (0) (13.3) (37.5) (50) (27.3) (32.7) other expressive act 3. Greeting + 1 6 0 1 2 8 3 0 3 24 other (20) (42.9) (0) (8.3) (66.6) (53.3) (12.5) (0) (13.6) (22.4) expressive act 4. Inquiry+ 3 0 0 0 0 0 1 0 3 7 other (60) (0) (0) (0) (0) (0) (4.2) (0) (13.6) (6.5) expressive act Total 5 14 10 12 3 15 24 2 22 107 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: different generatiOn, SG: same generatiOn; E-Y: elder to younger; Y-E: younger to elder; SA: same age

5.3.2.2.2 In reply letters As demonstrated in Table 5.13, the preference for combination patterns also varies for different age groups in reply letters. Within the different generations in the family letters, older writers prefer to use Combination 2 (47.4%), while younger writers overwhelmingly use Combination 1 (83.3%). However, as with initial letters, age differences have little effect in the choice of the combination patterns across different generations in the non-family setting, and it is the same within the same generation in both family and non-family settings. All seven groups of different ages prefer to choose Combination 1 (66.7%, 50%, 75%, 62.5%, 84.6%, 83.3%, and 87.5%).

184 Table 5.13. Frequency and distribution of combination patterns of openings across age/generation in reply letters (N=107) Combination Age/Generation Total patterns Family Non-family DG:25 SG:22 DG: 12 SG: 48 E-Y Y-E E-Y Y-E E-Y Y-E E-Y Y-E SA (19) (6) (12) (10) (4) (8) (26) (6) (16) (107) 1. Greeting + 8 5 8 5 3 5 22 5 14 75 aclmowledge- ( 42.1) (83.3) (66.7) (50) (75) (62.5) (84.6) (83.3) (87.5) (70.1) ment of receipt of a previous letter + other expressive act 2. Aclmowledge- 9 0 3 2 0 0 2 1 1 18 ment of receipt (47.4) (0) (25) (20) (0) (0) (7.7) (16.7) (6.3) (16.8) of a previous letter+ other expressive act 3. Greeting + 2 1 1 3 1 3 2 0 1 14 aclmowledge- (10.5) (16.7) (8.3) (30) (25) (37.5) (7.7) (0) (6.3) (13.1) ment of receipt of a previous letter Total 19 6 12 5 4 8 26 6 16 107 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: different generatlon, SG: same generatlon; E-Y: elder to younger; Y-E: younger to elder; SA: same age

5.3.2.3 Discussion From the examination of the relation between the age/generation factor and the choice of opening types and the ways in which the combined openings are sequentially structured, it can be seen that the effect of age/generation is not balanced. For the opening types, the influence of age/generation mainly works for the writers of the different generations in both family and non-family settings, and sometimes occurs in the same generation in the non-family setting. Within the different generations younger writers use single openings less than older writers in both settings. In the same generations younger writers choose the single strategy equally to elder writers in the family setting, while younger writers use single openings less than elder and same-aged writers in non-family letters though this is not very marked.

As mentioned in section 5.2.3, an opening with a single strategy conveys less of social message to the recipient, and involves less face work. Hence it is not surprising to see that younger writers of the different generations use the single opening less than the

185 older writers in both family and non-family letters, though both of them make extensive use of the combined openings. By doing so, the younger people avoid sending insufficient social messages, which may be interpreted as inappropriate in communicating with the older people, thus affecting the maintenance of the relationship. It can be suggested that the social norms in terms of age/generation (e.g. deference to elder people) require younger people to have an awareness and sensitivity towards the orientation of an opening (less polite type vs. more polite type).

However, for the same generation, such a requirement seems to be only restricted to the younger writers in the non-family setting, although the sample numbers are very small. This phenomenon appears to be related to the degree of certainty or uncertainty the writers perceive about their relationships in particular circumstances (Wolfson 1988). In accordance with 'bulge theory' (Wolfson 1988), it would seem that the younger writers in this situation are more uncertain about the relationship with the recipient in the non-family (e.g. normal friends) than in the family context, and thus need to do elaborate work in the opening in order to build up or maintain the relationship. The situation for those in the family circle (e.g. siblings) is different. The blood relation makes the younger writer confident about the stable relationship with the elder recipient, thus perhaps feeling freer to some extent to choose the single opening in the same way as the elder one.

As for the choice of combination structures of openings by writers within the different and the same generations, the findings are two-dimensional. On the one hand, the effect of the age/generation factor in people's structure patterning only focuses on the different generations, and only in the family setting. That is older family writers prefer to use the relationship-emphasizing pattern and younger family writers overwhelmingly use the most elaborate combination pattern. However, both older and younger non-family writers show an equal preference for the combination pattern emphasizing social identification and social concern (greeting + other expressive act; greeting + acknowledgement of receipt of a letter + other expressive act). The choices of different structures by different generations of family members result from the traditional requirements for linguistic behaviours under the hierarchical structure of the Chinese family, which are shaped by Confucian norms of filial piety and deference to elders, and

186 _,;,. .... ,·.~· ...... ~" "' '-'-'"· '--""'',6:.'

warm-heartedness to the younger (Ebrey 1991). These structural similarities by writers from the non-family different generations could be explained that both older and younger writers follow a general or socialized pattern in order to avoid showing personal feelings due to a certain distance between them.

On the other hand, both elder and younger writers of the same generation in both family and non-family letters have a similar choice of combination patterns in initial and reply letters: Both equally make an extensive use of the highly conventionalized combination patterns. This result may suggest that both elder and younger people in the same generation are more attentive to well documented cultural patterns and their individual's actual positions in age are subordinate to desires to follow 'public compliance' concerning 'harmony maintenance' (Bond & Hwang 1989: 257). This reasoning is consistent with Chu's finding that Chinese subjects are less likely to respond independently of the group (1979). According to Bond and Hwang, it may be the 'collectivism', a significant feature in the Chinese culture, that enables both the elder and younger persons to accept group-oriented values with the purpose of maintaining "good human achievements and tradition" (1989: 257). However, this conclusion is only suggestive since the subjects for the same generation in this study are only restricted to certain categories (i.e. siblings and friends).

5.3.3 Openings and gender This section examines the influence of another social factor, gender, on the opening behaviour in relation to opening types and structural patterns in Chinese personal letters.

5.3.3.1 Differences in the choice of opening types Table 5.14 indicates that both females and males show preferences for the use of combined openings to single openings in both family and non-family letters (88.7% vs. 74.5%; 93%, vs. 75% respectively). However, a closer observation of the data reveals that males are more likely to use the single opening than females in both settings (25.5% vs. 11.3%; 25% vs. 7% respectively). The data also show that the gender differences are sometimes intertwined with the age factor to influence the language use by some gender. For instance, the single opening is often appropriate for older male writers (family: seniors: lO,juniors: 4; non-family: seniors: 10, same age: 7,junior: 1).

187 Table 5.14. Frequency and distribution of opening types across gender (N=259) Types Gender Total Family Non-family F: 53 M: 55 F:71 M: 80 259 N % N % N % N % N % 1. Single opening 6 11.3 14 25.5 5 7.0 20 25 45 17.4 2.Combined openings 47 88.7 41 74.5 66 93 60 75 214 82.6

Total 53 100 55 100 71 100 80 100 259 100

5.3.3.2 Differences in the choice of the strategy of single openings Table 5.15 summarizes the gender differences in the choice of individual strategies for the single opening. It indicates that, on one hand, females have fewer choices of individual strategies than males: They are restricted to the use of relationship-oriented strategy (i.e. inquiries), while males use a greater variety of strategies. On the other hand, females use the inquiry strategy considerably more often than men (66.7% vs. 28.6%), while males use the greeting strategy more often than females (50% vs. 33.3%). The gender differences for this respect are mainly found in the family letters.

Table 5.15. Frequency and distribution of the strategy of single openings across gender (N=45) Strategies Gender Total Family Non-family F: 6 M: 14 F:5 M:20 45 N % N % N % N % N % 1. Greetings 2 33.3 7 50 4 80 11 55 24 53.3 2. Inquiries 4 66.7 4 28.6 1 20 3 15 12 26.7 3. Acknowledgement 0 0 2 14.3 0 0 3 15 5 11.1 of receipt of a previous letter 4. Apologies 0 0 0 0 0 0 2 10 2 4.4 5. Expressions of good 0 0 0 0 0 0 1 5 1 2.2 wishes 6. Expressions of 0 0 1 7.1 0 0 0 0 1 2.2 missing Total 6 100 14 100 5 100 20 100 45 100

5.3.3.3 Differences in the choice of combination patterns of openings The following section will investigate the gender differences in the choices of patterns of combined openings in initial and reply letters.

188 5.3.3.3.1 In initial letters As shown in Table 5.16, it appears that no significant difference appears between females and males in the choice of Combinations 2 and 3 in both family and non-family letters, but females differ from males in the preference for Combination 1 and 4 in these two settings: females use Combination 4 (Inquiry + other expressive act) more than males (11.5% vs. 0%; 12.5% vs.O% respectively), while males choose Combination 1 (Greeting + inquiry) more often than females (43.8% vs. 24%; 54.5% vs. 30.3% respectively). Based on these observations, it can be suggested that males show a preference for the highly conventionalized patterns, while females exhibit a stronger preference for the relationship-oriented combination pattern.

A closer observation of the data also shows that the gender differences m the preferences for the structural patterns mainly occur between intimates (i.e. family members, friends), and no significant gender-based difference is found among distants (i.e. teachers, students). On the other hand, the gender differences are intertwined with the age factor, although the degree of the age influence is not balanced in different contexts. For instance, Combination 4, favored by females, appears to be most appropriate for older writers (i.e. mother, elder sister) in the family setting (3 out of 3), or for both older or the same-aged writers under the friendship context (same-aged: 3; elder: 1). No younger female uses this pattern in both contexts. However, Combination 1 which is preferred by males appears to be most appropriate for both older and younger writers (i.e. father, elder brothers, younger brothers) in the family letters (older: 3; younger: 4), and also appropriate for all the age groups among friends (elder: 10, same aged: 7; younger: 2).

Table 5 .16. Frequency and distribution of combination patterns of openings across gender in initial letters (N=107) Combination patterns Gender Total Family Non-family F:25 M: 16 F: 33 M:33 107 N % N % N % N % N % 1. Greeting + inquiry 6 24 7 43.8 10 30.3 18 54.5 41 38.3 2. Greeting+ inquiry+ 11 38.5 6 41.2 10 30.3 8 24.2 35 32.7 other expressive act 3. Greeting+ other 5 20 3 18.8 9 28.1 7 21.9 24 22.4 expressive act 4. Inquiry+ other 3 11.5 0 0 4 12.5 0 0 7 6.5 expressive act Total 25 100 16 100 33 100 33 100 107 100

189 5.3.3.3.2 In reply letters Table 5.17. Frequency and distribution of combination patterns of openings across gender in reply letters (N=107) Combination patterns Gender Total Family Non-family F:20 M:27 F:28 M:32 107 N % N % N % N % N % 1. Greeting + 11 55 15 51.9 22 78.6 27 84.4 75 70.1 acknowledgement of receipt of a previous letter + other expressive act 2. Acknowledgement 7 35 7 29.6 3 10.7 1 3.1 18 16.8 of receipt of a previous letter + other expressive act 3. Greeting + 2 10 5 18.5 3 10.7 4 12.5 14 13.1 acknowledgement of receipt of a previous letter Total 20 100 27 100 28 100 32 100 107 100

Table 5.17 summarizes the gender differences in the use of combination patterns in reply letters. As shown in this table, there is no significant difference in the preference of Combination 1 between males and females in both family and non-family letters (55% vs. 51.9%; 78.6% vs. 84.4% respectively). But females differ from males in the preference for Combinations 2 and 3 in both contexts. Females use Combination 2 (acknowledgement of receipt of a previous letter + other expressive act) more than males (10.7% vs. 3.1%) in non-family letters. No distinct difference is shown for this combination across gender in family letters (35% vs. 29.6%). On the other hand, males use Combination 3 (greeting + acknowledgement of receipt of a letter) slightly more than females (18.5% vs. 10%) in family letters, while there is no significant difference regarding this usage across women and men (10.7% vs. 12.5%) in non-family letters.

Similar to that in initial letters, the gender differences in the preferences for these two patterns are also mainly centered among intimates in both family and non-family letters in reply letters. But it is found that the age impact does not seem to be apparent in the gender differences in this respect. But, this claim is tentative, as the evidence for this result is very limited.

190 5.3.3.4 Discussion Based on the findings, it can be argued that gender differences do influence the ways in which an opening is realized (i.e. selecting types, strategies, and structural patterns) in Chinese personal letters. The first difference between females and males is related to their different orientations in performing openings. The findings show that females appear to use the combined openings more than males, while males use the single opening more than females, although both of them show a strong preference for the use of the combined structure. Given that the combined openings conveys more social message and is thought of as a more polite type of openings, more use of this opening type by females demonstrates that Chinese women are likely to be more linguistically polite than men while organizing the openings in personal letters. Such a gender-related contrast in the choice of opening types could be due to psychological differences in orientation to others (Holmes 1995: 7). As with some findings obtained by some scholars (see Boe 1987; Holmes 1995), compared with men, Chinese women place more focus on making connection with others, and show more consideration and involvement (Tannen 1986) in the performance of openings. In contrast, Chinese men are more concerned with status and the hierarchical relationships when dealing with the start ofwritten communication.

Gender differences in the choice of opening types are also related to the Chinese cultural background. As is well known, each culture has its own prescription of gender based appropriate behaviors (Cameron & Coates 1989; Chamber 1992; Gordon 1997; Trudgill 1983). Chinese culture has traditionally shown a significant gender differentiation (more male centered) in the social interactions (Bond & Hwang 1989: 240). Based on an analysis of gender differences in Chinese request patterns, Wei (1997) finds that Chinese women act more polite than men in several requesting situations. According to Wei (1997: 205), the causes of this gender difference are societal and cultural, such as social factors and Chinese traditional expectations of women's and men's speech. For instance, traditionally, Chinese women's linguistic behaviours should be modest, elegant and polite. In contrast, there is no such restriction for men's linguistic behaviours. As the dominant gender, the man is not generally expected to include many prestige elements in his speech. Given these Chinese

191 traditional norms for women's and men's speech styles, it is no surprise that Chinese women use the combined strategies of openings more often than do Chinese men.

The second evident difference concerns the ways in which Chinese women and men choose a specific strategy for achieving the single opening and construct the sequences in combined openings. Regardless of the similarities in the choice across gender, the findings suggest that females attend to the relationship-oriented way while performing either the single or the combined openings, while males tend to use the status-oriented expression or sequences in achieving these two types of openings in general.

This result could be related to different perceptions from each gender with respect to the appropriateness in social interactions. According to the literature, in general women tend to value and enjoy the intimate and affect-oriented talks (Holmes 1995: 37), which are thought of as a more personal approach to interaction. But men tend to adopt a communicative style that "necessitates no compromise in the maintenance of their own social status" (Lee-Wong 2000: 230), which is a more instrumental and less personal approach to interaction. This view seems relevant and applicable for the gender differences in performing openings for this study. For example, when a woman uses the pattern of 'inquiry+ other expressive act' towards her daughter such as 'How have you been recently? I hope you are studying hard, learning lots and making good friends,' she shows attention to the emotional bond with her daughter by conveying consideration and involvement. The purpose of using such a sequence by this mother is not so much status (senior vs. junior) enhancement, rather to establish or maintain solidarity and connection (Chodorow 1974: 44). When a male adopts the pattern of 'greeting+ inquiry' towards his son as in 'How are you? Have you been well recently?', on one hand, he tends to maintain his own and the recipient's status (i.e. senior, junior), and on the other hand to reveal the relationship between them (i.e. kin relationship). Such styles of openings contributed by Chinese women and men can be labeled as balanced and co-operative communication styles (Maltz & Barker 1982). This evidence confirms the view suggested by Zimin that "the sexes have different speech styles or perhaps different ways of dealing with the world" (1981: 38). For men, it is the world of positions, and for women it is the relationships.

192 What is more interesting is the intersection of intimacy/distance differences with gender differences. As evident from the findings, gender differences in the choice of sequential patterns in both initial and reply letters appear to be restricted to family and social intimates. This finding is interesting since it supports the view that the linguistic behaviour associated with women and men is not just a matter of sex-role demonstration, but also a matter of avoidance (Gordon 1997: 47). For example, in the data women within the family and in the friendship context use the pattern 'inquiry + other expressive act' more than men with the same situation, but women with distant relation to the recipients (i.e. teachers, students) do not show such a preference for this pattern. Rather, they avoid using this pattern in the same way as men do. This is not because these women do not want to appear affectionate or intimate than they really are, but because they want to avoid the misinterpretation attached to the personal-oriented pattern due to a certain of inherent social distance. For them, inquiring about the recipients and then expressing some specific personal feelings could be interpreted as too intimate, which may make the recipient feel uncomfortable according to their relationship (e.g. teacher-student). The answer to this fact seems to lie clearly with the double standard regarding gender linguistic behaviours (Gordon 1997: 61). From this evidence, it may be argued that gender differences cannot be understood without putting them in a context of society, as noted by Aries (1996: 195).

5.4 Concluding remarks This chapter examined the opening behaviour conducted by Chinese people in personal letters. Based on analysis of the empirical data from this study, some concluding remarks can be made. First, a series of expressive acts, or phatic expressions, are used to achieve openings in Chinese personal letters. Such an opening practice demonstrates Chinese people's perception of the need of taking phatic expressions to be just one way of attending to the establishment or maintenance of the relationship and achieving transition in personal written communication. Second, greetings, inquiries and acknowledgements of receipt of a letter are perceived as three key opening strategies in letters, which shows a strong influence of Chinese cultural values of propriety ( 'li '), consideration and sincerity on the written opening behaviour. Third, Chinese people show a great preference for using the combined openings, thus revealing a tendency of conveying maximal social messages to achieve sufficient identifications of the status

193 and relationships. Fourth, various social factors, such as social roles, age, and gender influence Chinese people's opening performance (e.g. choosing types, patterning structures) on different levels, hence providing some management features which differentiate social groups with regard to the opening behavior in this particular society.

194 CHAPTER SIX CLOSINGS IN CHINESE PERSONAL LETTERS

6.0 Introduction · This chapter examines the closings of Chinese personal letters. The first part of the chapter describes the' grammatical and pragmatic features of closings, as well as the manner in which they are organised. The aim is to explore the properties, the sequencing and the operation of sequencing in closings. The second part investigates the extent to which social factors such as role, age, gender, and cultural values and beliefs influence the closing behaviour, presenting cultural ways of interactive written communication in this particular society.

6.1 Closing strategies Within the data set of this study, eleven varieties of closing strategy have been identified. As is noted in Table 6.1, the frequency with which these strategies occur is as follows: expressions of good wishes (35.9%); pre-closings (29.6%); longing for future contact (8.7%); sign-off phrases (7.4%); expressions of concern (5.0%); reassurance about oneself (4.1% ); farewell message (2. 7% ); sending regards to others (2.0%); offering help (1.6%); expressions of gratitude (1.6%); and expressions of apologies (1.1 %). The statistics clearly show that 'expressions of good wishes' and 'pre-closings' are the two most commonly used strategies. The strategies of 'longing for future contact' and of 'sign-off are somewhat common (8.7%, 7.4%), but the other seven strategies have a more limited usage. As is the case with openings, the different strategies are not mutually exclusive, and some strategies are often combined. That is why there are 514 instances of closings, even though there are only 259 letters. The following section will examine the linguistic realisation of these closing strategies, including their linguistic forms and pragmatic functions.

195 Table 6.1. Categories and distribution of closing strategies in Chinese personal letters ( 514 instances) Strategies Functions Patterns N 0/o 1. Expressions Express the writer's 1.1. Concrete good wishes. 180 35 of good wishes desire for good e.g. 'Zhi ni yiqie shunli. '(Wish you all fortune for the right in everything.) recipient 1.2. Conventional phrases. 5 0.9 e.g. 'Zhuhao. '(All the best.)

Subtotal 185 35.9 2. Pre-closings Initiate the 2.1. Indication of the writer's desire to end 136 26.5 appropriate end of the contact. the letter writing e.g. 'Haole,jiuci gebi. Yuyan hou xu.' (Well, stop here, more will be talked about next time.) 2.2. Reasons for stopping+ indication of the 16 3.1 writer's desire to end the contact. e.g. 'Haole, youyu shijian guanxi, zheci xian dao ci, xiaci zaitan. ' (Well, because of time, there is no more to add this time. Talk again next time.)

Subtotal 152 29.6 3. Longing Express desire for 3 .1. Explicit indication of future contact 41 7.9 future contact further connection e.g. 'Wang laixin. ' (Looking forward to hearing from you.) 3.2. Implicit indication of future contact 4 0.8 e.g. 'Xiwang neng gen ni duo tan.' (Hope to talk with you again later on.)

Subtotal 45 8.7 4. Sign-off Conventionally 3.4. Conventional pattern. 38 7.4 phrases indicate termination e.g. 'Cizhi jinli' (With high respect) of writing 5. Expressions Express concern 5.1. Take care. 14 2.7 of concern about the well- e.g. 'Duo baozhong. '(Take care.) being of the 5.2. Hope you look after yourself recipient e.g., 'Xiwang ni nen zhuyi shent.i' 12 3 (Hope you look after yourself.)

Subtotal 26 5.0 6. Reassurances Show consideration 6.1. Describe one's current state+ reassure 21 4.1 about oneself and affection the recipient. e.g. 'Wo zai zheli yiqie dou hao, ni fangxin ba. '(All is well with me here. Please put your mind at ease.) 7. Farewell Affectively indicate 7.1. Goodbye. 14 2.7 termination of e.g. 'Zaijian. '(Goodbye.) messages writing 8. Sending Express good wish 8.1. (Please) give regards to .... 11 2.1 for the third party e.g. 'Zuihou daiwen daye dama hao.' regard to (Finally, give my regards to uncle and others aunt.)

196 9. Offering help Show consideration 9 .1. If there is any problem (or if you need 8 1.6 and involvement anything), please let me know ( + I will give a hand) e.g. You shenme kunnan, qing gaosu wo, wo yiding hui bangzhude. ' (If there is any difficulty, please let me know, I will give a hand.) 10. Expressions Express gratitude 10 .1. Thank you for. .. 8 1.6 for the benefit e.g. 'Xiexie ni dui wo de zhufu. ' of thanks obtained (Thank you for your good wishes.) 11. Apologies Provide an apology 11.1. Explicit apologies. 1 0.2 for an offence to e.g. 'Hen baoqian, nide huixin tuo le restore a good yige xinqi. '(An apology for having relationship delayed in writing back to you for one week.) 11.2 Admitting offence and requesting 5 0.9 forgiveness e.g. 'Dui chifu nide xin, zaicijingqing yuanliang. ' (Please forgive me for the delay of writing back again.)

Subtotal 6 1.1 Total 514 100

6.1.1 Expressions of good wishes This is the strategy of expressing good wishes for the recipient's health, life, enjoyment, and other aspects. It is used to create a happy ambience (Fong 2000: 224), serving to consolidate the relationship between writer and recipient. As is evident from the data, this strategy occurs not only in openings, but also in closings. The difference is that in the case of the former, it is infrequently used while in the case of the latter it is the most frequently used of all strategies. Also, in the openings the well-wishes are often achieved through a sentence-level structure, such as 'wish you good health and all is going well.' In the closings, however, they can be realised either through sentences or through conventional phrases such as the following:

1.1 'Concrete good wishes' Zhuni xuexi jinbu, ytqte shunli. wish you study progress everything smooth. Good luck and progress in your studies.

1.2 'Conventional phrase' Zhuhao wish good All the best.

197 There are various contents that can be used in expressions of concrete good wishes. The common ones include 'good health' (18.4%), 'good luck' (17.3%), 'happiness' (15.7%), 'improvement' (12.4%), 'smoothness' (9.7%) and 'enjoyment' (3.2%). In a similar way to those of the openings, most good wishes in the closings are expressed by a combination of two or three wishing aspects (74.1 %), such as 'Good luck and progress in your studies.' In addition, the choice of the content of the combination appears to be related to the social roles of the writers and recipients. For example, children often express good wishes to their parents by combining 'good health' with 'good luck'; friends often perform this act by combining 'improvement' with 'good health' or 'happiness'; students often send wishes to their teachers by combining 'good health' with 'no troubles.' Thus, it appears that combinations of good wishes are situation-specific, especially in regard to the social roles of the recipients.

To convey good wishes in general, a number of writers use a conventional phrase, such as 'All the best' or 'Good luck this spring.' This is an infrequent strategy, which only accounts· for 0.9% of the total instances, and it also only occurs in letters between friends. It is likely that the conventional well-wishing versions are employed in the letters exchanged between writers who had similar educational experiences and thus learned the same traditional phrase patterns.

It is interesting to find that good wishes can be duplicated in the opening and closing sections. From the rhetorical point of view, the double use of good wishes separated in these two sections can be seen as a way of producing an echoing effect, emphasising the writer's strong intention of bringing good fortune to the recipient. It is worth noting that in these cases these expressions are often distinguished in form by some preceding devices. When they occur in the opening, they are often preceded by an adverb 'shouxian' (at first) in the initial position, such as 'At first, wish you well in health and all right in everything.' When they occur in the closing, the writers often add the adverb 'zuihou' (finally) before the wishes, such as 'Finally, wish you happy and all right in everything.' The data shows that the duplication of good wishes appears to be an appropriate practice for juniors (children, young relatives, students). By duplicating good wishes, these writers attempt to emphasise their respect for seniors.

198 6.1.2 Pre-closings Pre-closing phrases refer to the expressions that signal the writer's desire to end the letter. They serve to represent 'a warrant' for smoothly undertaking the end of the contact (Schegloff & Sacks 1973: 303-305). In addition to their function of terminating the letter, these devices can also be seen as a way of offering some reassurance as to "the continuation of the relationship by emphasising group membership" (Placencia 1997: 60). The pre-closing devices are the second most frequently occurring strategy (29.6%), and they are often realised through these two patterns:

2.1 'Indication ofthe writer's desire to end the contact' a. Bie bu duo tan. something else not more talk There is no more to say. b. Jiuci ge bi, yuyan hou xu. here lay down pen rest after talk I'll stop writing here. The rest will be talked about next time. c. Haole, zai me shenme shi le, JIUCI weizhi ba. good again no what thing (now) here stop (do you agree) Well, there is no more to add. So I'll stop here for now.

2.2 'Reasons for stopping+ indication of the start of the closing' a. Youyu shijian guanxi, jintian zhan dao c1. because time relation today temporary come here Because oftime, I'll stop here for now. b. Haole, xianzai kuai yao shangke le, jiuci zhi bi. good now soon go to class (now) here stop pen Well, because it is time to go to class, I'll stop writing here for now.

As can be seen in examples 2.1 a and 2.1 b, the pre-closings are realized through an explicit indication of the writer's desire to finish the present phase of writing. Common phrases are 'There is no more to add' or 'I'll stop writing here. The rest will be talked about next time.' This direct initiation of the end is not popular with Chinese writers, and is restricted to letters written by family or non-family seniors (i.e. parents, older relatives, and teachers). It is evident from the data that Chinese writers often frame the initiation of the end of writing by preceding the phrase with a discourse marker 'haole' (well) , such as with 'Well, there is no more to add, so I'll stop here for now.' in example 2.1 c. On the surface this marker seems to indicate that the present topic has

199 reached the end (Schegloff & Sacks 1973: 309-313). Underneath this, it serves as a means of mitigating the abruptness which may arise from the closing phrases, and also serves to bring about a tactful closing, hence assuring the continuation of the relationship by emphasising in-groupness and politeness (Placencia 1997: 60). In general, the mitigating pre-closing expression is often used by juniors and equals.

In several cases (3.1 %), the initiating-end statement is preceded by a reason for stopping the writing, as in example 2.2a: 'Because of the time, I'll stop here for now.' The provision of the reason for ending the contact (e.g. lateness, desire to do some other thing) can be seen as another means of mitigating the directness or abruptness embedded in the initiating expression. Reasoning serves to assure cooperation and increase acceptance of the ending from the point of view of the writer. By doing so, the writer tends to pay homage to the recipient's negative face, showing deference and affection as well (Brown & Levinson 1987). Sometimes, writers may precede the reasoning phrase with the discourse marker 'haole' (well), as in pattern 2.2b. By combining the discourse marker with a reason for stopping, the writer tends to further mitigate the directness of the initiation of the end of writing, thus serving to signal familiarity and in-group membership. This pattern appears to be appropriate for teachers, students and friends.

6.1.3 Longing for future contact The fourth commonly used closing strategy is that of longing for future contact, which accounts for 8.7% of the total instances. This strategy refers to the expression of the writer's desire to interact with the recipient again in future. It demonstrates the writer's attention to the maintenance of the relationship and to further connection with the recipient. This act is often realised through the following structures:

3.1 'Explicit indication of future contact' a. Duo xie xm more write letter. Write more. b. Jian xin hui yin see letter back voice. Write back when you receive this letter.

200 c. Wang lai xi. hope come letter. Hope you write back. d. Dendai nide hui . xin wait for your reply letter Looking forward to hearing from you.

3.2 'hnplicit indication of future contact' a. Panwang zaori dedao ninde zhijiao. look forward earlier get your advice Looking forward to your advice at your earliest convenience.

As seen in examples 3.1 a and 3.1 b, the writers directly demonstrate their desire to continue their interaction by asking the recipient to write back. The literal meanings of these expressions, such as 'write more' or 'write back when you receive this letter', can be treated as requests, but they do not actually function here as requests. Rather, they serve to convey that the writer regards the recipient as an in-group member and hopes to continue the relationship. In examples 3.lc and 3.1d the writers express their hope for the continuation of interaction in a milder way: They use some linguistic devices like 'wang' (hope) and 'dendai' (wait for) to express their desire for future contact, as in the phrase 'looking forward to hearing from you'. This style seems to be softer or more emotional, which reflects the writer's strong desire for the continuation of the relationship. The data shows that the pattern in examples 3.la and 3.1.b seems to be appropriate for seniors in the family (i.e. parents), while the pattern shown in examples 3.lc and 3.1d seems to be suitable for equals in both family and non-family settings.

Aside from the explicit expressions, longing for the future contact can be also realised in an implicit way, as in the pattern of 3.2. In this case, instead of directly or mildly asking for a reply from the recipient, the writer implies desires to continue the connection by indicating an eagerness to get the recipient's suggestions. This can be seen as a respectful way to keep the connection going in future, since it may be thought of as less imposing than the explicit expressions. It is used to stress deference to the recipient. On the other hand, it may imply some social distance. This pattern is appropriate for non-family juniors (i.e. students).

201 6.1.4 Sign-off phrases Sign-off phrases such as 'Cizhi jinli' are formal and conventional tokens that signal the ending of written interaction. As formal-style terminating indicators, they no longer carry their original meaning (I hereby submit salutation), but they are roughly equivalent to 'with high respect.' They are treated as polite devices. The following are some examples found in the sample used for this study:

4.1 'Conventional pattern' a. Ci zhi - j inli here submit salute With high respect b. Zhi - jinli submit salute With high respect c. Ci - li here salute With high respect d. Shun zhi by the way submit With high respect

The pattern in the first three cases (4.la, b, c) consists of two parts (divided by '-'). The first one is placed either immediately after the last paragraph of a letter (e.g. closings), or in the middle of the line below the last paragraph. The second part is put at the beginning of the next line. The fourth pattern (4.1d) is made up of one phrase, which is often placed in the middle of the line below the last paragraph. In general, the sign-off strategy is more prevalent for juniors in both family and non-family settings (i.e. children, students) to show deference. Occasionally, this strategy can be used by persons of equal status (e.g. friends), in which case it serves to convey deference and in group membership.

6.1.5 Expressions of concern This is the strategy of using expressiOns that denote the writer's concern for the continued well being of the recipient. It emphasises the writer's consideration and affection, and is intended to reaffirm and consolidate the relationship between writer

202 and recipient. It is interesting to note that this strategy can also be used in the opening section. In closings, however, this strategy has a different form of linguistic realisation. In openings, this strategy is often realised either by a simple sentence, or a by complex structure in which a precondition precedes the core act, as in 'Chunhan douqiao, wang ni duo zhuyi shenti.' (In the chilly air of the spring period, please take care of your health.) In closings, however, it is often realised through a formulaic phrase or a simple sentence, as in the following two patterns:

5.1 'Take care' Duo baozhong. more take care Take care.

5.2 'Hope you look after yourself Xiwang nimen duo zhuyi shenti. hope you more pay attention to body Hope you look after yourself.

The phrase in example 5.1 is a minimal version that conveys the writer's positive attitude towards the recipient's life in general. The expression in example 5.2 is a syntactically marked formula which emphasises the writer's concern about the health of the recipient. The minimal version can be regarded as simple and conclusive in displaying general consideration, while the syntactic pattern can be seen as an expression emphasising the writer's specific concern. As is shown in the data, the syntactic pattern appears to be appropriate for seniors or equals in family (i.e. parents and elder siblings). This pattern serves to convey the writer's consideration and affection, and to emphasise their emotional bond with the recipient. The simple token seems to be used most often by juniors (i.e. children, students) or equals in both family and non-family letters, and serves to signal deference and in-groupness.

6.1.6 Reassurances about oneself 'Reassurances about oneself can serve as an opening strategy as well as a closing strategy. In this strategy, writers give a positive description of their situation, with the intention of comforting recipients. This strategy signals and reaffirms the relationship by paying homage to the recipient's positive face need (e.g. to be noticed, cared about) (Brown & Levinson 1987: 103), thus emphasizing in-group membership. It is worth

203 noting that when this strategy occurs in the closing, it is always placed at the beginning of closings, acting as an initiator of the end of contact. In this case, the 'reassurance' carries somewhat 'summarising' characteristics (Albert & Kessler 1978: 542), which may create an affective 'historical record' of an encounter so that the writer can seek the recipient's psychological agreement to close the writing (Placencia 1997: 60). In addition, while acting as a closing strategy, the reassuring expression is often preceded by a concluding marker 'zongzhi' (on the whole), which, like the discourse marker 'haole' (well) in the pre-closing strategy, serves to mitigate the directness and haste of ending the writing. The following are common 'reassurance' patterns:

6.1 'Describe one's current state+ reassure the recipient' a. Wo zai zheli yiqie dou hao, qing fang xin. I in here everything all good please put down heart All is well with me here. Please put your mind at ease. b. Zongzhi, jiali )'lqie dou hao, wu nian. on the whole family everything all good not think over. On the whole, all is well with the family, so don't worry.

In these two examples, the writers use a similar pattern: positively describing one's own or another's present situation, then reassuring the recipient. The difference is that in pattern 6.1 b the writer adds a concluding marker before the main body of the discourse, such as 'Zongzhi, jiali yiqie dou hao, wu nian. ' (On the whole, all is well with the family, don't worry.) The addition of this phrase may serve to provide a warrant for smoothly initiating the end of the present writing. From observing the data, it appears that this is a family-oriented strategy and is often used by parents, children or siblings. It conveys consideration and involvement, emphasising in-group membership.

6.1.7 Farewell messages Following Clark and French (1981), farewell messages such as 'zaijian' (goodbye) in personal letters seem to function as an intimate means of a 'terminal exchange.' The use of these messages reflects the writer's concern over 'the survival of the relationship' (Wolfson 1988: 34). Based on the observation of the data, it appears that farewell messages seem to occur only in letters between siblings and friends, serving to reaffirm solidarity and in-group membership. They, however, are not used in cases where the

204 relationship between the writer and recipient is asymmetrical and either very intimate or relatively distant (e.g. parent-children, teacher-student). From this, it would appear that the farewell message is closely related to role-relationships while acting as a closing strategy.

6.1.8 Sending regards to others Sending regards to others can be seen as another cue for closings in Chinese personal letters. As in the opening section, it refers to a request for passing the writer's salutation to a third party, serving to consolidate a wide network of in-groupness. This strategy is often linguistically realized in closings though the same pattern as it is realized in openings, such as 'Daiwen yonghui tam en hao. ' (Remember me to Y onghui and his fellows.), or 'Qing daiwen nimen jia hao.' (Please give my regard to your family.). The former is a general form, while the latter is a polite expression marked by the polite device 'qing' (please). At times, the closing function of this strategy is signalled through the use of a preceding device like 'zuihou' (finally) or 'bin' (and), such as 'Zuihou, dai wo xiang shushen wenhao. '(Finally, give my regards to uncle and aunt.). Based on the data from this study, sending regards to others seems to be most appropriate for equals in both family and non-family settings.

6.1.9 Offering help In this strategy the writer shows consideration and involvement with the recipient by offering help. The main purpose of this offer is not to actually do something for the recipient. Rather, it is intended to convey sincere feelings, reinforcing the established relationship. This strategy is often realized through the following patterns:

9.1 'If there is any problem (or if you need anything)+ please let me know(+ I will give a hand)' a. You shenme xuyao bangzhude, jingguan gaosu wo. have what need help feel free tell me Ifyou need anything, don't hesitate to tell me. b. You shenme kunnan, qing gaosu wo, wo yidin hui bangzhu de. have some difficulty please tell me I must help If there is any difficulty, please let me know, I will give a hand.

205 It becomes clear that this strategy is realized through an indication of difficulties the recipients may have or help they may need, followed by a request for further information, such as 'You shenme xuyao bangzhu de, jingguan gaosu wo.' (If you need anything, don't hesitate to tell me.), as in Pattern 9.1a. Sometimes a writer may explicitly offer help after first presuming difficulties and asking to be notified, as in Pattern 9.1b 'You shenme kunnan, qing gaosu wo, wo yidin hui bangzhu de.' (Ifthere is any difficulty, please let me know, I will give a hand.). By doing so, the writer tends to emphasize the close relationship with the recipient. The data shows that this strategy is used most frequently among friends.

6.1.1 0 Expressions of thanks As examined in the chapter on openings, expressing thanks is a speech act that indicates the writer's indebtedness for what the recipient has done. This strategy appears to be used not for the purpose of factual confirmation, but rather to assure the writer's warm and friendly feelings towards the recipient (Coulmas 1981: 71). Expressing thanks in closings is often linguistically realized through the common pattern of 'thank you for. .. ', as in 'Xiexie nide wenhou. ' (Thank you for your greeting.). Sometimes the expression of thanking is preceded by the adverbs 'zuihou' (finally) or 'zaici' (again), such as 'Zaici ganxie nide bangzhu' (Thank you for your help again.). It is worth noting that 'zaici' (again) is often used in the case where the thanking is a duplication. In this case, the thanking strategy seems to convey the feelings that the writer cares about their indebtedness to the recipient, thus signalling the writer's attention to the relationship. As is evident in the data, this strategy is more prevalent among students, and it serves to show deference and as well as to imply a certain degree of distance.

6.1.11 Apologies As in openings, apologies in closings act as social and affective speech acts. They can also be thought of as 'face-supportive acts,' which are primarily aimed at providing a remedy for an offence and at restoring social equilibrium and harmony (Holmes 1995: 155). As a closing strategy, apologies are often used by students, and sometimes by children and friends, and they are a general signal of deference. Like those in openings, closing apologies are linguistically formed in the following patterns: 'explicit expression of apologies,' 'admitting an offence + requesting forgiveness,' and 'an

206 explanation + admitting an offence + requesting forgiveness.' The second pattern is used more commonly than other two patterns.

It is interesting to fmd that, as a closing strategy, apologies provide a remedy not only for the offence of delaying one's reply to letters, as was the case with opening apologies, but also and even more so for a letter being written inappropriately (i.e. too hastily, too briefly, or too-long-windedly). Such is the case with the apologetic sentences 'Xin xiede liaocao, qing laoshi yuanliang' (Please forgive me for this hasty writing, teacher.) and 'Langfei le bushao shijian, qing yuanliang' (I'm costing you much time to read through this letter, please forgive me.). It would seem that apologies in closings are more concerned with the actual aspects of an offence. This might suggest that writers' motivations for apologising are related to more than simply their perceptions of what is necessary to maintain the relationship (i.e. apologising for the delay in answering). Writers are motivated also by their desire to impress upon the recipient how much they care for and value the relationship by being sensitive to the language style.

Another feature of apologies in closing is their duplication. The data shows that at times writers may use apologies in the closing after they have already been used in the opening. In this case, the expression of apology is often preceded by the adverb 'zaici' (again), such as with 'Dui chifu ninde xin, zaici jingqing yuanliang.' (Again, please forgive me for the delay in answering your letter.). With the repetition of apologies, writers tend to emphasise their "one-down position" relative to the recipient (Holmes 1995: 163), thus emphasising their deferential intention. The data shows that this strategy is often used by students.

6.1.12 Discussion In light of this description of the linguistic forms and social meanings of closings in Chinese personal letters, some points are summarised and discussed here. First, as in openings, a writer achieves a closing in personal letters through a range of expressive speech acts (excluding the pre-closing phrase), such as good wishes, longing for future contact, and sending farewell messages. Similarly, these expressive acts basically perform the phatic function (Manlinowski 1972). That is, following Laver's approach to phatic communion in conversational closings (1981), they are used to mitigate

207 potential feelings of rejection that may arise due to the initiation of the end of communication, and to consolidate the relationship between writer and recipient. The importance ofthis becomes obvious if any of these strategies is omitted. For example, if a writer ends the writing without using any phatic expression, without including, for example, an indication of the desire to finish the current written contact, a reference to the continuation of contact, or the expression of good wishes, the relationship between writer and recipient would be affected, since the recipient's 'psychological comfort' is most at risk "in the crucial marginal phrases of encounters" (Laver 1975: 236). Therefore, the use of closing strategies can be thought of "interactively necessary work" (Pavlidou 2000: 137).

The second feature of Chinese written closings is the extensive use of both expressions of good wishes and pre-closings, which emphasize a mutual belief in good luck and an offer of reassurance for the continuation of the relationship. The overwhelming preference for good wishes appears to reflect the high value placed on 'luck talk' in Chinese culture. According to Fong, a norm for Chinese interaction is that of expressing good wishes with respect to health, prosperity, happiness and improvement to one another, which is said to bring good luck to the addressee. Such luck talk can be seen as "a cultural-linguistic system" and it is used to "create balance and harmonize communicative behaviours" (2000: 224).

Chinese people's belief in good luck can be seen as evidence of the widespread popularity of the Taoist philosophical principles of 'yin' and 'yang' (Fong 2000: 233). Following this philosophy, participants see bad luck as a negative energy force that creates unfavourable life events (yin), and good luck as a positive energy force that creates favourable life events (yang). The belief in these two elements is a factor that motivates the widespread practice of expressing good wishes in letters. In the context of personal letters, this practice not only reflects the writer's belief that good wishes can bring good fortune and help maintain harmony and balance in the future life of the recipient, but it also conveys the writer's strong intention of consolidating their relationship with the recipient.

208 The use of pre-closings affirms Brown and Levinson's theory of face (1987). According to Placencia, the inappropriate closing of an interaction can be a face threatening act, resulting in damage to the relationship. Therefore, in the way they express their desire to end the current interaction, participants need to be tactful. They need use some devices to seek mutual agreement to close the interaction (1997: 60). The use of pre-closings in letters appears to serve this purpose. When writers use such phrases as 'Well, I'll stop writing here. The rest will be talked about next time,' or 'Because it is time to go to class, I'll stop writing here for now', they seem to imply that they have to close the letter against their will due to the length of writing or some time consuming activity. This can be regarded as a way of 'attending to the recipient's want' (e.g. their psychological comfort). This is a positive politeness strategy, ultimately aimed at claiming common ground (Brown & Levinson 1987: 102), and offering a warrant for the initiation ofthe end ofwriting (Schegloff & Sacks 1973: 305).

The third feature is that some strategies occurring in openings also can be used in closings. This allows for parallel strategies to be used in openings and closings. This issue is two-dimensional. One dimension is that some common strategies only occur in the closing. Such strategies include expressions of concern, expressions of thanks, apologies, reassurances about oneself, and sending regards to a third party. The occurrence of these strategies in closings appears to be motivated by 'conversational,' 'structural' and 'situational' requirements, aside from the general phatic properties of these strategies. For instance, expressing thanks, concern, or sending regards to others can be considered to be 'conversationally required' or at least expected (Pavlidou 2000: 133). When recognising that they have failed in the opening to thank the recipients for something, writers may then express their gratefulness somewhere in the closing by using a phrase such as 'Thank you for your card.' A writer thanks the recipient in the closing, as if making up for the failure to do so in the opening. Reassurances about oneself in the closing can be regarded as a 'structural' requirement. For example, a 'reassurance' phrase such as 'On the whole, all is going well with the family, so don't worry', is often taken as a 'summarising statement' (Albert & Kessler 1978: 542), which serves to signal the closing of the letter. The use of this strategy in the closing not only confirms membership of a social group, but also creates "a concise historical record of an encounter" (Albert & Kessler 1978: 542). This allows for the recipient to

209 be comfortably convinced that the closing of the letter has begun. The provision of apologies seems to be a situational necessity. As mentioned in chapter five, apologies in openings usually refer to a delay in answering the previous letter, but in closings they usually refer to some inappropriate feature of the present letter. Probably, giving such apologies at the end of a letter may not seem to make as much sense as the apologies given for delaying one's answering a letter in the opening, as the former appears to be pure 'formulae.' It is observed, however, that Chinese may even refer to this kind of apology, because it is situation-specific. Giving this kind of apology may be a way for a writer to demonstrate particular attention to a recipient's face needs (e.g. to be cared) (Brown & Levinson 1987: 103).

The second dimension of this issue is the repetition of one common strategy in both the opening and the closing. As mentioned earlier, some writers duplicate a certain strategy in both opening and closing sections, such as expressing good wishes, apologies, or thanks. This duplication can be seen as an emphatic necessity for conveying a writer's phatic and politeness meanings, as opposed to the repetitions made for the purpose of minimal structural necessities. Tannen (1989) discusses various types of repetition and the functions they may accomplish in conversation. One function which she considers to be universal, and which seems to be applicable to Chinese context of this study, is that of creating personal involvement. Hence, the duplication of certain strategies in these two sections might, in fact, serve to keep the recipient focused on the writer's specific feelings or attitudes, and convey the writer's intention of emphasising a particular meaning.

The final point is that the linguistic realizations of closing strategies are less elaborate than those of opening strategies. As discussed in chapter five, the majority of opening strategies are comprised of a compound structure, such as ' precondition/pre-ground + a core act' (e.g. It has been five months since I left home. I am wondering whether all is going well with the family.). By using this structure, writers tend to achieve the relevance and clarity of the whole discourse, emphasising the importance of particular elements (Lee-Wong 2000), and also emphasising the writer's sincerity (Shih 1986). However, many closing strategies are manifested through a simple phrase or sentence. Why are the strategies structured in such a different way in these two sections? From

210 observing the data, it appears that the lack of elaboration in closings can be considered a result of formulaic requirements. The closing strategies manifested through simple structures are always those that are linguistically formulaic (e.g. goodbye, with high respect, take care) .. When they appear in a fixed formulaic manner, their lexical and semantic structures seem to be short and precise.

On the other hand, the less elaborate structure of closing strategies can also be seen as a result of rhetorical requirements. The closing is the section in which interaction is discontinued and it normally requires a concise linguistic structure. It is appropriate in an opening for a writer to convey concern for the recipient by saying 'The cold winter is approaching. I hope you are looking after yourself.' However, the letter would seem somewhat wordy if the writer were to structure the closing strategy in the same style. From an organisational point of view, 'I hope you are looking after yourself would have been sufficient at the end of a letter. It can thus be suggested that both the structure and the organisation of the segments in an interactional location (e.g. closings of a letter) influence the way in which this interaction is performed.

In summary, the closings of Chinese personal letters use a range of expressive acts or routine formulae to serve phatic functions. Good wishes and pre-closings are the two key components of the formulation of closings. Parallel strategies are used in openings and closings for emphatic purposes. The linguistic features of closing strategies are characterised by their simple lexis and semantics.

6.2 Closing types As with openings, there are two types of closings in Chinese personal letters: single closings and combined closings. The single closing is made up by one strategy, while combined closings consist of two or more strategies. As shown in Table 6.2, Chinese writers exhibit a strong preference for using combined closings rather than single closings (76.4% vs. 23.5%).

211 Table 6.2. Frequency and distribution of closing types in Chinese personal letters (N=259) Types N % 1. Single type closing 59 22.8 2. Combined type closing 200 77.2 Total 259 100

The following section examines these two respective types of closings. There are two main questions addressed in this section: Which strategies are appropriate for single closings? and how are individual strategies sequentially organised to realise a combined type?

6.2.1 Single closings As evident in Table 6.3, eight strategies can be used as a single strategy in the closing section. Among the eight, the expression of good wishes is used most frequently (59.3%). The second most common strategy is the pre-closing statement (15.3%). The six other strategies include sign-off phrases, longing for future contact, expressions of concern, apologies, expressions of thanks, and offering help. However, these six strategies occur infrequently, making up less than 7% of the total number single types in the study.

Table 6.3. Frequency and distribution of the strategy of single closings (N=59) Stratel!;ies N % 1. Expressions of good wishes 35 59.3 2. Pre-closings 9 15.3 3. Sign-offphrases 4 6.8 4. Longing for future contact 4 6.8 5. Expressions of concern 2 3.4 6. Apologies 2 3.4 7. Expressions of thanks 2 3.4 8. Offering help 1 1.7 Total 59 100

One noticeable feature observed from Table 6.3 is that in no case is a single strategy in closings comprised of farewell messages, reassurances about oneself, or sending regards to others. The explanation for this might be related to the features of each of these three strategies. Farewell messages, such as 'zaijian' (goodbye), are seen as a sort of

212 'terminal exchange' (Clark & French 1981: 3), which bring the writing to completion. This strategy is often accompanied by other strategies, since by itself it fails to convey any desire to continue the relationship, which is one of the major aims of closings. If a letter simply ends with a 'goodbye' and without any other preceding or following strategy, this closing can be interpreted as meaning that the writer intends to take a long break from contact with the recipient, a message which both parties tend to avoid. Reassurances about oneself, such as in the phrase 'Wo zai zhe yiqie dou hao, qing fangxin. '(All is well with me here. Please put your mind at ease.), normally function as a summarising statement which initiates the end of the closing. In general, it is assumed that reassurance about oneself will be followed by other strategies. Sending regards to others normally functions as a means of reaffirming a wide network of social relations. However, if this alone is used in the closing section, it does not make a strong reference to the continuation of the relationship between writer and recipient. Considering all these arguments, it appears again that the emphasis of closings is on the consolidation of the relationship and the continuity of future contact, despite the differences in the type of closing chosen.

Table 6.4. Frequency and distribution of single closings in initial and reply letters (N=259) Strategies Initial letter: 145 Reply letter: 114 N 0/o N % 1. Expressions of wishes 26 19.3 9 7.9 2. Pre-closings 6 4.1 3 2.6 3. Sign-offphrases 3 2.1 1 0.9 4. Longing for future contact 0 0 4 3.5 6. Expressions of concern 0 0 2 1.8 5. Apologies 1 0.7 1 0.9 7. Expressions of thanks 1 0.7 1 0.9 8. Offering help 1 0.7 0 0

Total 38 26.2 21 18.4

A close examination of the data shows that the occurrence of the single strategy in closings varies between the different types of letters. As suggested in Table 6.4, the single strategy occurs more frequently in initial letters than in reply letters (26.2% vs. 18.4%). The possible explanation for this might be that writers of initial letters usually have less common ground to share with and refer to due to their infrequent contact with

213 recipients. As this issue has been addressed in chapter five, detailed explanation is omitted here.

Apart from the difference in the frequency of occurrence, the focus of the content of single closings in initial letters also differs from that of reply letters, although the evidence for this is not strong. For instance, the expression of good wishes occurs more frequently in initial letters than in reply letters (19.3% vs. 7.9%), while longing for future contact and expressions of concern occur only in reply letters (3.5% and 1.8% respectively). Such differences are also related to the frequency of contact between writer and recipient. In the case of infrequent contact, it is reasonable for a writer to employ a more general or conventional strategy (e.g. good wishes) to end the writing. In a context of frequent contact, however, it will be appropriate for a writer to use context-specific strategies (e.g. expressing concern, longing for future contact) to remind the recipient of their emotional ties or acquaintance. Therefore, it is frequent contact that provides a writer with the basis (i.e. knowledge of the recipient's state, confidence about the relationship) for using either of these single-strategy closings.

6.2.2 Combined closings Since no significant difference is found in the organising structures of combined closings between initial and reply letters, these letters will not be analysed separately. As shown in Table 6.5, four main sequential patterns are identified from these data. The overall preference for these patterns is in the following order: 'pre-closings + expression of good wishes/sign-off (50%), 'pre-closings + (other expressive act) + expression of wishes/sign-off+ other expressive act' (25%), 'other expressive act + expression of wishes/sign-off (15%), and 'pre-closings + other expressive act' (1 0%). The following section investigates how these patterns are sequentially organised.

214 Table 6.5. Combination patterns of closings and their distribution in Chinese personal letters (N=200) Combination patterns N % 1. Pre-closings + expressions of good wishes/sign-off phrases 100 50 2. Pre-closings + other expressive act 20 10 3. Pre-closings + other expressive act+ expressions of good wishes/sign-off 50 25 phrases(+ other expressive act) 4. Other expressive act+ expressions of good wishes/sign-off phrases 30 15 Total 200 100

Combination 1: Pre-closings + expressions of good wishes/sign-off phrases Half of the Chinese writers (50%) in the data use this combination pattern of closings. Within this pattern, closings are initiated with a pre-closing phrase and are terminated by an expression of good wishes or a sign-off formula. Here are some examples of this pattern:

Example (1) (parents to son) (1) Bie bu duo tan. (2) Zhu m yiqie pmgan, xueyie youchen. something else not more talk. wish you everything peaceful study successful There is no more to add. Wish you peace in life and success in your studies.

Example (2) (younger brother to elder brother)

(1) Hao le, (2) jiuci ge bi, yuyan hou xu. (3) Zhu ni shenti jiankang good (now) here put pen the left later talk wish you body healthy

gongzuo shunli. work smooth Well, I'll stop here for now. I'll talk about the rest later. Wish you good health and I hope things go smoothly at work.

Example (3) (teacher to student)

(1) Shijian guanxi, zhan ge ci. (2) Zhu yiqie shunli. time relation temporarily put here wish everything smooth Because oftime, I'll stop here temporarily. All the best.

As is evident from these examples, the writers initiate the closings by using pre-closing statements, such as 'Bie bu tuo tan ' (There is no more to add), 'Haole, jiuci gebi, yuyan houxu.' (Well, I'll stop here for now. I'll talk about the rest later.), or 'Shijian guanxi, zhan ge ci.' (Because of time, I'll stop here temporarily.). Borrowing Schegloff and Sacks' approach to closings in telephone conversation, these pre-closing expressions

215 can be seen as announcements that clearly state the writers' desire to terminate the present writing. As is apparent from the examples, the pre-closing expressions are preceded by some mitigated devices, such as the discourse marker 'haole' (well) as in Example (2), or a reason for stopping writing 'youyu shijian guanxi' (because of time) as in Example (3). These devices serve to mitigate the directness embedded in the explicit announcement of ending the writing, thus providing a warrant for undertaking this closing procedure. Under Brown and Levinson's framework of politeness, these devices can be also regarded as measures of politeness, as they are "attending to the hearers' interest", and constitute positive politeness strategies ultimately aimed at claiming common ground (1987: 102).

After initiating the closing with the pre-closing, the majority of the Chinese writers (70 out of 100) terminate their writing with expressions indicating concrete good wishes, such as 'Zhu quanjia xingfu. '(Wish the whole family well.), as in Example (1), or 'Zhu ni shenti jiankang, gongzuo shunli. ' (Wish you good health and I hope things go smoothly at your studies.), as in Example (2). The data reveals that a concern for achieving a proper termination is the underlying motivation for attaching expressions of good wishes to the pre-closings. This confirms the findings of Schegloff and Sacks (1973: 318). By using expressions of good wishes relating to good health or improvement, writers tend to convey their concern for the recipient's future well being, a concern which has much to do with "the desire for continuity" (Albert & Kessler 1978: 542).

Occasionally (5 out of 70), a writer may use an expression of good wishes which refers to general aspects of the recipients, such as 'Zhuhao' (All the best), or 'Chunqi' (Good luck this spring). These kinds of good wishes are a conventional form of well-wishing, and they are often used among friends and siblings.

Besides the overwhelming use of good wishes as a terminating device in this pattern, sign-off phrases are sometimes used to end the closings (20 out of 100), as shown in the following examples:

216 Example (4) (student to teacher) (1) Jintian jiu xie dao zheli ba. (2) Ci zhi jingli. today just write to here (do you agree) here submit salute I have to stop here today. With high respect.

Example (5) (son to parent) (1) Haole, jiu ci ge bi. (2) Zhi jingli. good just here put pen submit salute Well, I have to stop here for now. With high respect.

Generally, conventional sign-off formulae like 'Cizhi jingli' or 'Zhi jingli' (With high respect) are used to signal that the end of written interaction has been reached. They are treated as traditional-style politeness devices for marking the termination of writing. Initiating the end of interaction may to some extent present a potential threat to the recipient's psychological comfort, even though the initiating statements (i.e. pre-closing) are often modified by some sort of mitigation. In such instances, the pre-closings should be followed not only by general positive statements, such as good wishes, but also by conventional deferential formula, which would repair the potential damage. From the writer's side, terminating the writing with the sign-off phrase, which connotes a certain level of formality and deference, would prevent the recipient from experiencing the potential psychological discomfort produced from the pre-closings. Thus, contact would be continued and guaranteed, although there is no explicit reference to future contact.

It is interesting to find that Chinese writers sometimes (1 0 out of 100) like to terminate the writing by combining specific good wishes with the sign-off phrase, as the following examples show:

Example (6) (daughter to father) (1) Hao, dao ci, xmc1 zai tan. (2) Zhu jiankang shunxin. good come here, next time again talk wish health happy

(3) Ci zhi jingli. here submit salute Well, I'll stop here and talk to you next time. Wish you good health and happiness. With high respect.

217 Example (7) (friend to friend) (1) Haole, xiaci zai tan. (2) Zhu xueyie youchen. (3) Ci zhi Jingli. good next time again talk. wish study successful here submit salute Well, I'll talk to y.ou next time. Wish you success in your studies. With high respect.

In these two cases, the writers terminate the written interaction with a combination of expressions of wishes and the sign-off phrases, such as with 'Zhu jiankang shunxin. cizhijinli' (Wish you good health and happiness. With high respect.). The combination of the informal expression of good wishes with the formal sign-off formula can be seen as a way of strengthening the relationship. That is because this mixed terminal expression not only reflects the writer's positive attitudes towards the recipient, but it also conveys some degree of deference. The evidence here suggests that the terminating expression can incorporate one or two ritual formulae of the same class that function independently. However, this way of terminating the letter is used infrequently (10%).

Combination 2: Pre-closings + other expressive acts In Chinese letters, a writer may terminate the closing section with some sort of expressive act after having first initiated the end with a pre-closing. Ten percent of the combined closings follow this sequential pattern. A typical characteristic of this pattern is that a context-specific expressive act is added after the pre-closing. Specifically, it may rely on the background information of previous letters from the recipient, as in the following cases:

Example (8) (parents to son) (1) Bie wu ta shi. (2) Xiwang duo gei jiali xie xin. other no other thing hope more give family write letter There is nothing else. Hope you write more to the family.

Example (9) (friend to friend) (1) Haole, xin dao ci. (2)Yihou yougong chang lai xm. good letter come here later have time often come letter Well, I'll stop here. Write me often when you have time.

The data shows that the most commonly used expressive act in this pattern is 'longing for future contact' (13 out of 20), which is very often expressed through the formula

218 'Wang laixin. ' (Hope you write me.), or 'Yihou yougong chang laixin' (Write me often when you have time.). By asking for future contact, the writer appears to pave the way for future encounters. This act can also demonstrate the writer's affection for the recipient, signalling that the relationship will continue after this particular written communication.

In addition to 'longing for future contact,' 'sending regards to other' is also commonly employed as a terminating device as in:

Example (1 0) (friend to friend) (1) Jiu dao zheer ba. (2) Dai wo xiang zhang tong wen hao. just arrive here (do you agree) behalf me to zhang tong ask good I'll stop here. Give my regards to Zhang Tong.

The writer in this example terminates the writing by sending regards to someone with whom both writer and recipient are familiar. Using this expressive act, the writer appears to reflect implicitly personal bonds with the recipient by confirming a social network they are both involved in. Through reminding the recipient of their "ties of common acquaintance" (Pavlidou 1997: 160), the writer demonstrates a share of membership of a social group, thus signalling a close relationship with the recipient.

Apart from expressing continuity and giving regards to a third party, a few writers may terminate the writing by expressing gratitude or concern after the pre-closings, as in:

Example (11) (friend to friend) (1) Haole, bu duo shuole, xiaci zai tan. (2) Xiexie nide wenhou. good not more talk next time again talk. thank your greeting Well, there's no more to say. Thanks for your greeting.

Example (12) (younger brother to elder brother) (1) Bie wu ta shi. (2) Wang baozhong. other not other thing hope take care There's nothing more to say. Hope you take care.

Following Placencia, thanks and the expression of concern can be regarded as other signs of 'connectedness' (1997: 74), which have the similar function of reinforcing the relationship and enhancing future contact with the recipient (Laver 1975: 230). In

219 Example (11 ), where after the pre-closing the writer adds phrases of thanks like 'Xiexie nide wenhou. ' (Thank you for your greeting.), he or she conveys indebtedness to the recipient for the well-wishing, thus signalling appreciation and respect. Similarly, in Example (12), by ending the contact with an expression of concern such as 'Wang baozhong.' (Hope you take care.), the writer explicitly displays consideration and affection for the recipient, thus emphasising in-groupness.

From the above discussion, it may be suggested that the expressive acts used in this pattern appear to reassure the continuation of the relationship in a more personal and specific manner than expressions of good wishes or sign-off phrases. Goffman (1971) refers to such expressive acts as 'supportive exchanges,' which function to pave the way for future interaction.

Combination 3: Pre-closings + other expressive act + expressions of good wishes/sign-off phrases(+ other expressive act) This is the most elaborate combination pattern, and the second most commonly used (25%) sequential pattern in these data. This structure can be seen as an extended pattern developed from the basic combination 'pre-closings+ expressions of good wishes/sign off phrases', in which an expressive act or some expressive acts are inserted either before or after the expression of good wishes (or only before sign-off phrases). The intensified use of constitutive components for closings may reflect Chinese writers' perception that closing sequences need to be lengthy for strengthening the reaffirmation and consolidation of the relationship. The data shows that commonly adopted expressive acts include 'longing for future contact' (18 out of 50), 'farewell messages' (7 out of 50), 'offering help' (7 out of 50), 'expressions of concern' (6 out of 50), 'apologies' (5 out of 50), 'expressions of thanks' ( 4 out of 50), and 'sending regards to others' (3 out of 50). The following examples illustrate the representative structures involved in this pattern:

Example (13) (younger sister to elder sister) (1) Shijian bu zao le, womenjiu tandao zhe. (2) Panwang lai xin. time not early (now) we just talk here look forward come letter

220 (3) Zhu xuexi jinbu, xinqing yukuai. wish study progress mood happy It is late now, so I'll stop here. Waiting for your reply. Wish you improvement and happiness.

Example (14) (friend to friend) (1) Haole, youyu shijian guanxi, xian dao ci. (2) You shenme kunnan, good because time relation now arrive here have what difficulty

qing gaosu wo, wo yiding hui bangmang de. please tell me I must can help

(3) Zhu jieri kuaile. wish festival happy Well, because of time, I'll stop here. If there is any difficulty, please let me know, I will give a hand. Wish you a happy festival.

In Example (13), the writer employs the continuity statement 'Panwang laixin. ' (Waiting for your reply) before expressing good wishes 'zhu xue xi jinbu, xinqing yukuai. ' (Wish you improvement and happiness.). It is likely that making explicit reference to the continuation of the relationship before sending concrete good wishes may emphasise ties of personal connection and create personal involvement. In Example (4), the writer offers help to the recipient with the phrase 'You shenme kunnan, qing gaosu wo, wo yiding hui bangmang de. ' (If there is any difficulty, please let me know, I will give a hand.) before expressing good wishes. The provision of help here seems to serve the same function as the 'longing for future contact', that is, emphasizing that they share membership of a social group. Thus, these examples suggest that the overarching purpose of the expressive act is to please recipients by attending to their need to be of interest, although the choices of the contents are varied and context specific.

It is worth noting that the choice of the combination of an expressive act with expressions of good wishes or sign-off phrases is related to the type of letter. For instance, a 'longing for future contact' or a 'farewell message' accompanied by good wishes or sign-offphrases often occurs in initial letters (66.7% and 100% respectively). The combination of good wishes with 'offering help', 'expressions of concern', 'apologies', or 'expressions of thanks' appears to be appropriate for reply letters (75%, 66.7%, 66.7% and 100% respectively). Sending regards to others combined with good

221 wishes occurs equally in both initial and reply letters (50%, 50%). A possible explanation for these results is that writers have different concerns when writing initial and reply letters. Within initial letters, there is more concern over the continuation of contact, while in reply letters, there is more concern over specific matters arising from previous correspondence (e.g. thanking the recipient for having done something) or from the present writing (e.g. apologising for the improper writing style). The different focuses of initial and reply letters in their choice of particular expressive acts reflect to some extent the influence of contact frequency upon language use.

As well as adding one expressive act, a few writers (6 out of 50) employ two expressive acts before expressions of good wishes or sign-off phrases. This practice may reflect another feature of closings in Chinese letters: Before the terminating device there appear to be intensified additions, which may serve as a stronger way of enhancing the relationship. As is evident from the data, 'longing for future contact' is often ( 4 out of 6) accompanied by farewell messages, as in the following example:

Example (15) (friend to friend) (l)Haole,jiu dao ci ba. (2)Wangyou gonglai xin. (3)Zai Jian. good just come here (do you agree) hope have time come letter again see

( 4) Ci zhi jingli. here give salute Well, I have to stop here. Hope you write when you have time. Goodbye. With high respect.

In this case, before the sign-off phrase the writer uses a 'longing for future contact' and a 'farewell message,' 'Wang yougong laixin. zaijian' (Hope you write when you have time. Goodbye.) The statement of continuity of contact can be seen here as a means of affirming the relationship with the recipient, and the farewell message seems to be related to the writer's specific desire of reaffirming the relationship (acquaintance or in groupness). By saying 'zaijian' after the statement of continuity, the writer tends to further imply that their acquaintance or closeness will continue and endure, although the current contact is nearly finished 'here for now.'

In addition to the combination of 'longing for future contact' with 'farewell messages', other combinations may sometimes precede the expression of good wishes. There may

222 be combinations of 'longing for future contact' with 'sending regards to other' (1 out of 6), and 'expression of concern' with 'offering help' (1 out of 6), as in the following:

Example (16) (friend to friend) (1) Haole, jiu shuo zhexie ba. (2) Xiwang duo lianxi. good just talk these (do you agree) hope more contact

(3) Xiang wu feng wen hao. to Wu Feng ask good

(4) Zhu xuexi jinbu, shiye youcheng. wish study progressive course successful Well, this is all there is to say. Hope for more contact. Give my regards to Wu F eng. Wish you progress and success in your studies.

The occurrence of sending regards to someone after a statement of continuity, as with 'Xiwang duo lianxi. xiang wu feng wenhao' (Hope for more contact. Give my regards to Wu Feng.), can also be seen as a measure of the consolidation of the relationship. Through emphasising the personal bond and the shared memberships of a social group, the acquaintance or closeness can be reinforced.

Occasionally (4 out of 50), even more than two expressive acts can precede the expression of good wishes, as shown in Example (17):

Example (17) (friend to friend) (1) Haole, jiu shuo zhexie ba. (2) Xiwang ni duo baozhong. good just say these (do you agree) hope you more take care.

(3) Xiang liu lan wen hao. to liu lan ask good.

(4) Xiwang duo lianxi. (5) Zhu long nian hao yun, shiye youcheng. hope more contact wish dragon year good luck cause successful Well, this is all I have to say. Hope you take care. Say hello to Liu Lan. Keep in touch. Wish you good luck and success in the Year ofthe Dragon.

Obviously this is a more elaborate closing combination. By conveying consideration, stressing common membership, and referring to future connection before terminating the closing with the expression of good wishes, the writer tends to convey as much of a social message to the recipient as possible. This is an utmost manner of attending to the

223 recipient's interest, and thus laying a strong base for the continuation of the relationship. However, due to the small number of examples ( 4 out of 50), it is difficult to make any generalisations from such a combination with regard to topic selection and positioning.

Apart from inserting one or more expressive acts before the good wishes or the sign-off phrase, some writers (1 0 out of 50) may end the closing by adding one or more expressive acts after the good wishes. Attaching some expressive acts to the good wishes appears to result from the writer's purpose of achieving a special way of showing "attention and interest" to the recipient (Schegloff and Sacks 1973: 22). The placement of the expressive act after the good wises appears to show that, just after sending the good wishes, a certain feeling is 'going through the writer's mind'. Using Schegloff and Sacks' terms, the 'going through' feelings are treated as 'afterthoughts' (1973: 321). They serve to emphasise the writer's specific feelings towards the recipient. The following excerpts provide examples of this:

Example (18) (friend to friend) (1) Haole, jiu xie dao zhe, xiaci zai tan. (2) Zhu kuaile jiankang. good just write to here, next time again talk. wish happy healthy

(3) Qing dai wen nimen jia hao. please behalf ask your family good Well, I'll stop here, I'll talk about the rest next time. Wish you good health and happiness. Please give my regards to your family.

Example (19) (younger brother to elder brother) (1) Ci xin jiu xian xie dao zhe ba. (2) Zhu ni gongzuo shunli, this letter just first write to here (do you agree) wish you work smooth,

wan shi ruyi. (3) Panwang nide hui xin. million thing satisfactory look forward to your back letter I'll stop writing here. Wish you good work and all the best. Looking forward to your reply.

In these two examples, the writers begin the closings with pre-closings, then express good wishes concerning the recipients' good health, good feelings and improvement. After this, they either send regards to a third party as in Example (18) or express a desire for future contact as in Example (19). These final acts can be seen as a way of marking the expressive acts in a distinctive way. The respective interpretations these

224 expressions might be 'there is something I almost forgot, please remember me to your family,' and 'there is something else I want to tell you: I am waiting for your reply.' Thus, it is clear that the positioning of these two acts in the final places displays the writers' deliberate attention to tie the personal bonds with their recipients.

The data shows that expressive acts that are often placed after expressions of good wishes include longing for future contact (3 out of 10), sending regards to others (2 out of 10), and farewell messages (1 out of 10). The data also shows that the pattern of a combination with an afterthought is usually used in reply letters, and there are only two cases (in the pattern of good wishes plus longing for future contact) occurring in initial letters.

Besides attaching one or more expressive act before or after the 'expression of good wishes', sometimes writers may place an expressive act on either side of an 'expression of good wishes' (4 out of 10). The underlying motivation for doing so might be that the first expressive act plus the good wishes might not be sufficient to reaffirm the relationship, and therefore a second expressive act is needed. This combination appears to display a writer's orientation towards achieving a proper closing in a special way. However, this type of pattern is found infrequently in the data. The following are some representative examples:

Example (20) (friend to friend) (1) Youyu shijian guanxi, xian dao ci. (2) Zhuyi duanlian shenti. because time relation first come here notice do exercise body

(3) Zhu kaixin. (4) You shenme kunnan, qing gaosu wo. wish happiness have what difficulty please tell me Because oftime, I'll stop here temporarily. Hope you do more physical exercise. Wish you happiness. If there is any difficulty, please let me know.

Example (21) (elder sister to younger sister) (1) Yijing bu zhao le, zhan ge ci. (2) Lai xin! already not early (now) temporarily put here. come letter

(3) Zhu ping' an kuaile. (4) Dai wen liu de tamen hao. wish peace happy. behalf ask liu de they good It is late now, so I have to stop here. Write me! Wish you peace and happiness. Say hello to Liu Min and his friends.

225 In Example (20), after initiating the end by a pre-closing statement preceded by a reason, the writer expresses his concern about the recipient's health with the phrase 'Zhuyi duanlian shenti.' (Hope you do more physical exercise.), and then sends his good wishes 'Zhu kaixin. '(Wish you happiness.). As this is a combination type of closing, it has already achieved a proper initiation and termination, and has thus sufficiently reaffirmed and consolidated the relationship. However, it seems to the writer that such a structure is still insufficient. There appears to be something about their relationship with the recipient that is yet to be mentioned. The writer therefore adds another expressive act after the good wishes: offering help with 'you shenme kunnan, qing gaosu wo. '(Ifthere is any difficulty, please let me know.). On the basis of quantity, the addition of the offer of help apparently increases the force of reaffirmation and consolidation. From the recipient's view, the provision of help after the good wishes may display a special attention, not just by the addition of a phatic expression, but also by its conspicuous position.

Combination 4: Other expressive act+ expressions of good wishes/sign-off phrases The fourth pattern is made up of the combination of an expressive act with the expression of good wishes. Fifteen percent of combined closings in the data belong to this category. The use of the expressive act as the initiator seems to have two functions. One is to initiate the whole closing discourse, and the other is to accomplish its own interaction-relevant activity. The most commonly selected expressive act in this pattern is the 'reassurances about oneself (21 out of30), shown in the following examples:

Example (22) (father to son) (1) Jiazhong yiqie dou hao, wunian. (2) Zhu jiankang jinbu. family everything all good, not worry wish healthy progressive Everything has been well with the family, so don 't worry. Wish you good health and progress.

Example (23) (daughter to father) (1) Wo zaizhe ytqie dou hao, ni bie danxin. I in here everything all good you not worry

(2) Zhu ni shenti jiankang, shishi ruyt. wish you body health everything smooth. I am ok in everything here. Don't worry about me. Wish you good health and no trouble in anything.

226 In each these two cases, the writer begins the closing with reassurances about oneself and terminates it with expressions of good wishes. Reassuring the recipients in the initial place of the whole pattern can be taken as a pre-closing statement, indirectly signalling that there is nothing further to add (Placencia 1997: 60). As with normal pre closings, it serves to prepare a warrant for undertaking the closing. On the other hand, this act also bears its own function: expressing involvement and consideration, thus attempting to meet the recipient's face needs. In this regard, using the reassurances of the recipient as an initiator helps to build up both structurally and emotionally the warrant for the closing.

Some other expressive acts can also function as an initiator of closings, such as 'longing for future contact' (3 out 30), 'expressions of concern' (2 out of 30), 'sending regards to other' (1 out of 30), 'offering help' (1 out of 30), 'apologies' (1 out of 30), and 'expressions of thanks' (1 out of 30). These are infrequent, however.

Example (24) (elder sister to younger sister) (1) Shijian guode hen kuai, bie wangle gei women chang xie xm. time pass very fast not forget give we often write letter

(2) Zhu shenti jiankang. wish body health Time is passing very quickly. Don 't forget to write us often. Wish you good health.

Example (25) (younger sister to elder brother) (1) Tianqi liang le, duo baozhong shenti. (2) Zuihou zhuni gongzuo shunli , weather cool (now) more take care body at last wish work smooth xuexi jinbu. study progress. It is turning cool now. Take care ofyourself. Wish your work is running smoothly and your study is progressing well.

Example (26) (friend to friend) (1) Dai wen nide pengyou hao. (2) Zhu hao. behalf ask your friend good wish good Give my regards to your friends. All the best.

227 Example (27) (friend to friend) (1) Ru you shenme wenti, Jlnguan lai xin gaozhi. if have what problem feel free coming letter inform. (2) Zhi yiqie ~hunli. wish everything smooth If there are any problems, feel free to tell me. Wish you no trouble in anything.

Example (28) (friend to friend) (1) Hen baoqian, nide hui xin duole yi xinqi. very apology your reply letter delay one week

(2) Zhu shujia yukuai. wish summer vocation happy Many apologies for my delay ofone week in answering your letter. Wish you a good summer vacation.

Example (29) (student to teacher) (1) Zuihou xiexie ninde bangzhu. (2) Cizhi jing li. Finally thank your help hereby give salutation. Finally, thank you for your help. With high respect.

Similarly, the use of these expressive acts as the initiator of closings can be regarded as an indirect way of indicating the writer's desire to finish the present writing. For instance, indicating a desire for future contact may imply that there is no more to add, as a reference to a future issue implies that the present issues are concluded. Exhibiting concern about the weather and the recipient's health also signals that the writer has no more to say on the present topic by referring to an apparently irrelevant topic. Sending regards to a common acquaintance seems to implicitly mark the closure of the present topic by referring to the third party. By using such an indirect way of initiating the closing, the writer appears to attend to the recipient's interest (e.g. the recipient might not like an abrupt ending). This makes the beginning of closing process a certainty.

Based on the above discussion, it would seem that using certain expressive acts as initiators of closings not only serves to lay a warrant for undertaking the closings smoothly, as the conventional pre-closings do, but also serves to exhibit the writer's specific feelings or attitudes toward the recipient. It is this specific warranting and personally-oriented function that accounts for the restricted use of expressive acts at the beginning of closings: They are used only in situations of either great intimacy or great distance.

228 It is also worth noting that when expressive acts are used as initiators of closings, some of them, such as 'longing for future contact,' 'expression of concern,' or 'sending regards to others' are often realised through a complex structure or an informal pattern. This is despite the overall simplicity of closing strategies. For example, the statement of continuity is often structured by the expression of a precondition before the core act, such as with 'Shijian guode henkuai, bie wangle gei women chang xiexin. ' (Time is passing very quickly, so don't forget to write us often.), as in Example (24). Other such closings use an oral-like expression, such as 'Yougong xiexin liaoliao' (Write and chat when you have time.). Similarly, the expression of concern is also realized through a complex sentence that begins with a precondition, such as with 'Tianqi liang le, du baozhong shenti.' (It is turning cool, take care of yourself.), as in Example (25). At times, the expression of concern is accomplished by a casual expression such as 'Zaiwai duo zhuyi shenti.' (Be careful about yourself when you are outside.). The use of a complex structure and an informal pattern in the initial part of closings seems to further the purpose of mitigating the directness and hastiness resulting from the ending of topics. This may make the expressions sound softer and more intimate, which is a significant way of warranting the closing. Furthermore, it also vividly reflects the writer's attitudes or emotions, emphasising their relationship with the recipient.

6.2.3 Discussion This section has explored how closings are organised in Chinese personal letters, with a particular focus on the sequential organisations of closing strategies. Based on the findings, several points can be made in regard to the linguistic features of the structures and their underlying social meanings.

First, as with openmgs, Chinese writers exhibit a strong preference for combined strategies in closings. This trend reflects the basic communicative requirements of closings in human interaction. According to Laver (1975, 1981), closings in conversation serve two basic functions: To mitigate potential feeling of rejection that may arise from initiating the closing action, and to consolidate the relationship between the participants. Laver also adds that the polite norm embedded in closings, especially in communication between those who do not see each other often, is the requirement of using at least one mitigating or consolidating phrase, together with some appropriate

229 formulaic phrase of parting (1975: 236). Laver's proposal seems to be relevant and helpful for interpreting the finding that Chinese writers prefer combined strategies for closing letters. The first task for writers constructing closings in personal letters is to use pre-closings, either by offering a declaration of the start of the closing (i.e. pre closing), or by indirectly initiating the end of contact through reference to the writer's . specific feelings (e.g. reassurance about oneself). These closing initiators serve to establish a warrant for undertaking the closing smoothly (Schegloff & Sacks (1977: 303).

The use of some terminating expressions after pre-closings can be seen as a way of consolidating the continuity of the relationship, which is the second function of closings. Actually, the employment of a pre-closing expression alone seems socially and ritually insufficient for providing a 'warrant', and is thus insufficient for the task of reassuring future connection. According to Albert and Kessler, "to end an encounter it is necessary to affirm a non-ending", that is, to propose connection in future (1976: 165). Thus, the use of some ritual and phatic expressions ensuring the continuation of the relationship is also needed. Through using these expressions, along with the pre-closing devices, writers might be able to demonstrate their hope that the break in the present contact will not threaten the relationship in future (Schiffrin 1977: 680). In this case, Chinese writers' preference for the combined type of closings further supports the notion that human ritual is both formally and technically efficient (Huxley 1966), and symbolically meaningful (Goffman 1967).

The favouring of combined closings by Chinese writers demonstrates the perceived need to be polite in closing performances. According to Goffman (1967) and Brown and Levinson (1987), one of the central ideas of face work and politeness is that many acts in social interaction can put people's face at risk. To prevent this from happening, people may carry out face-saving actions (Brown & Levinson 1987) or face-satisfying strategies (Shih 1986). The adoption of combined closings can be seen as one of these measures. By using this measure, Chinese writers not only show tact by using mitigatory expressions to express their desire to end interaction, but by using the consolidating phrases also reinforce the relationship with the recipient. By combining these two kinds of strategies or by including some other appropriate formulae, writers

230 reassure their further connection with the recipient by claiming more common ground and sharing more personal experiences (Brown & Levinson 1987: 103).

A second generalisation is that closings in Chinese personal letters exhibit more or less fixed structural patterns with regards to their organisation. Based on the findings, a model is drawn to summarise the major sequential patterns of closings, as shown in figure 6.1. From this model several conclusions can be inferred. First, closings in Chinese personal letters are sequentially organised in an order based on the principles of proper initiation and proper termination (Schegloff & Sacks 1973: 318). These principles are motivated by the two basic functions of mitigation and consolidation, as discussed above.

Wishes/sign-off

Pre-closings Other expressive acts

Other expressive acts H Wishes/sign-off

Other expressive act •I Wishes /sign-off

Figure 6.1 Summary diagram of the sequential combination of strategies of closings in Chinese personal letters

In general, the initiating work is done by the pre-closing statement and occasionally by some other expressive acts (excluding expressions of good wishes, sign-off phrases, and farewell messages). The choice of initiator will affect the orientation of closings (e.g. conventional, intimate, friendly, deferential). The use of pre-closing statements as initiators can be seen as a conventional way of indicating the writer's desire to end the written contact. They are normally shaped by an unmitigated or mitigated

231 announcement of the start of ending, which serves to pave a warrant for undertaking a proper initiation. The use of some other expressive act as an initiating device (e.g. reassurances about oneself, longing for future contact, expressions of concern, sending regards to others) is treated as an indirect way of initiating the closing. In comparison with the conventional pre-closings, these acts seem to offer a more personal and context-specific manner of warranting a smooth closing.

The terminating part is fulfilled by three measures: expressions of good wishes or sign off phrases; other expressive act; and a combination of an expressive act or acts with expressions of good wishes or sign-off phrases. The first one is placed after the pre- . closings or expressive act that initiates the closing. The second and the third are only located after the pre-closing phrase. In general, the choice of the terminating phrases depends on the social context (see section 6.4) and the amount of communicative efficiency and politeness it requires. Expressions of good wishes or sign-off phrases are a highly conventionalised terminating measure in Chinese letters, which, in a very general sense, creates happy ambience and emphasise a harmony of interpersonal relationships. The use of other expressive act is seen as a more personally-oriented terminating device, which may assure the continuity of the relationship and further contact through displaying the writer's specific feelings or attitudes towards the recipient (e.g. intimacy, friendliness, deference) in that particular circumstance.

Terminating the closing by a combination of an expressive act or acts with expressions of good wishes or sign-off phrases can be seen as a lengthier terminating measure, which serves to dramatise the reaffirmation of the continuity and future of the relationship. It is worth noting that the positioning of the expressive act(s) in this combination is to some extent flexible: There could be an addition of an expressive act or acts before, after or around the good wishes (but not after sign-off phrases), according to the individual's communicative purpose. On the other hand, as with leave taking devices in conversational closings (Clark & French 1981: 4), the expressive acts added before, after, or around the good wishes seem to be independent of other strategies of this combination, and are thus optional.

232 In brief, closings in Chinese personal letters are constructed around two basic requirements: mitigation and consolidation. Correspondingly, they are structured through the initiation of the closing and the termination of the closing. These procedures are mostly realized by certain expressive acts (with few exceptions). Moreover, there can be some rather lengthy stretches of expressions during the production of the terminal strategy. This would appear to be a characteristic of closings in Chinese personal letters.

6.3 Combined closings and letter types As mentioned earlier, since there seems to be no significant influence of letter type (initial vs. reply) on the structural patterning of particular closing strategies, the letters have been taken as a whole in the examination of closing sequences. However, the data shows that although there is no significant difference in sequential patterns, the occurrence of these patterns does differ between letter types. The following section will investigate how the letter types affect sequential organisation.

Table 6.6. Frequency and distribution of combination patterns of closings across letter types (N=200) Combination patterns Initialletter:107 Reply letter:93 N % N % 1. Pre-closings + expressions of good wishes/sign-off phrases 53 49.5 47 50.5 2. Pre-closings + other expressive act 13 12.1 7 7.5 3. Pre-closing+ other expressive act+ expressions of good 32 29.9 18 19.4 wishes/sign-off phrases (+other expressive act) 4. Other expressive act + expressions of good wishes/sign-off 9 8.4 21 22.5 _phrases Total 107 100 93 100

As shown in Table 6.6, combination 1 (pre-closing+ expressions of good wishes/sign off phrases) occurs almost equally in both initial and reply letters (49.5% vs. 50.5%). Combination 2 (pre-closings+ other expressive act) occurs slightly more often in initial letters than in reply letters (12.1% vs. 7.5%). Combination 3, the most elaborate pattern, is used more often in initial letters than in reply letters (29.9% vs. 19.9%). Finally, Combination 4 (other expressive act + expressions of good wishes/sign-off phrases) appears to occur more often in reply letters than in initial letters (22.5% vs. 8.4%).

233 As revealed in these findings, it appears that there is no significant difference across letter types in the occurrence of the highly conventionalised combination pattern (Combination 1). However, there is a clear differentiation between letter types in the frequency of Combinations 3 and 4, although to varied extents. To explain the prevalence of Combination 3 in initial letters, one might hypothesise that this most elaborate pattern appears to correspond to what Knapp refers to as the pattern of "dramatisation of relationship features" (1978: 107). Its use in initial letters may denote writers' perception of the need to use lengthier closings to establish or re-establish their relationships with recipients with whom they have had no previous correspondence or have not had any contact for a long time.

With respect to the preference for Combination 4 in reply letters, the explanation could be because this pattern is treated as more personally oriented, due to its feeling-focused initiator. Due to their high degree of closeness and acquaintance, it seems reasonable and appropriate for writers who have frequent contact with the recipient to indicate their specific feelings and attitudes at the very beginning of the closing. It seems that where there is a frequent contact between writer and recipient, the writer may feel freer to use a certain expressive act to initiate the end of written contact.

6.4 Closings and social variables This section examines how social factors such as roles, age, gender and cultural norms affect closings of Chinese personal letters. More specifically, the intention is to examine the extent to which linguistic variations interact with these social and cultural factors in both the choice of closing types and the ways in which closings are sequentially structured.

6.4.1 Closings and social roles It can be observed from Table 6. 7 that persons of different social roles have different preferences in their choice of single or combined types of closings. Parents and teachers use the single closing the more regularly (57.7%, 50%) than all other social roles, while children and students use this type the least often (11.1 %, 11.5%). Compared to the single type, the combined closings enjoy a higher preference among most groups, including children (88.9%), students (88.5%), siblings (85.4%), relatives

234 (75%), and friends (80%). It can be concluded from these findings that the single closing is used more by seniors, while the combined type is used more by juniors and equals.

Table 6. 7. Frequency and distribution of closing types across social roles (N=259) Types Social roles Total Family Non-family Parent Child Sibling Relative Friend Teacher Student (259) (26) (18) (48) (16) (115) (10) (26) 1. Single 15 2 7 4 23 5 3 59 closings (57.7) (11.1) (14.6) (25) (20) (50) ( 11.5) _(22.8) 2. Combined 11 16 41 12 92 5 23 200 closings (42.3) (88.9) (85.4) (75) (80) (50) (88.5) (77.2) Total 26 18 48 16 115 10 26 259 (100) (100) (100) _(100}_ JlOO}_ JlOO}_ JlOO}_ _(100}_

6.4.1.1 Single closings and social roles Apart from the overall differences shown in the preferences for the two types of closings, the data also shows that the choice of specific strategies and the organisational patterning of closing sequences also vary between different social roles. The following section will investigate the relationship between social roles and the choice of strategy, beginning with single closings.

It can be summarised from Table 6.8 that the 'expression of good wishes' is preferred by most social roles (i.e. excluding students) examined in this study (parents: 53.3%, children: 50%, siblings: 57.1 %, relatives: 50%, friends: 69.6%, teachers: 60%). The sign-off phrase is preferred by students (66.7%) and also children (50%) in forming the single type of closings. In addition to the overall preference for sending good wishes, parents, siblings and relatives are likely to choose the pre-closings (parens: 26. 7%, siblings: 28.6%, relatives: 25%) and some personally-oriented strategy as the single closing. For instance, parents and siblings tend to use the strategy 'longing for future contact' (13.3% and 14.3% respectively). Relatives are likely to choose 'expressions of concern' (25%). The different preferences of social roles for a specific strategy in the single closing indicate that the strategy typically harmonizing the relationships through luck talk is appropriate for seniors and equals in both family and non-family settings, while the phrase emphasizing deference and formality is appropriate for juniors in these

235 two settings. Furthermore, personally-oriented expressive acts are more appropriate for seniors and equals in the family setting. These findings show that the selection of the single closing strategy is related to the role relationship and the degree of intimacy or distance between the writer and the recipient.

Table 6.8. Frequency and distribution of the strategy of single closings across social roles (N=59) Strategies Social roles Total Family Non-family Parent Child Sibling Relative Friend Teacher Student (59) (15) (2) (7) (4) (23) (5) (3) 1. Expressions of 8 I 4 2 16 3 I 35 (53.3) (50) (57.1) (50) (69.6) (60) (33.3) (59.3) good wishes 2. Pre-closings 4 0 2 I 2 0 0 9 (26.7) (0) (28.6) (25) (8.7) (0) (0) (15.3) 3. Sign-off 0 I 0 0 0 I 2 4 phrases (0) (50) (0) (0) (0) (20) (66.7) (6.8) 4. Longing for 2 0 I 0 I 0 0 4 future contact (13.3) (0) (14.3) (0) (4.3) (0) (0) (6.8) 5. Expressions of I 0 0 I 0 0 0 2 concern (6.7) (0) (0) (25) (0) (0) (0) (3.4) 6. Expressions of 0 0 0 0 2 0 0 2 thanks (0) (0) (0) (0) (8.7) (0) (0) (3.4) 7. Apologies 0 0 0 0 2 0 0 2 (0) (0) (0) (0) (8.7) (0) (0) (3.4) 8. Offering help 0 0 0 0 0 I 0 I (0) (0) (0) (0) (0) (20) (0) (1.7) Total 15 2 7 4 23 5 3 59 (100) (100) (100) (100) (100) (100) (100) (100)

6.4.1.2 Combined closings and social roles Compared with the single closing, the influence of social roles on the choice of the sequential patterns in closings appears to be more complex, and thus deserves more attention. As shown in Table 6.9, there are differences between social roles in their choice of combination patterns. In general, parents tend to use Combination 4 (other expressive + expressions of good wishes/sign-off phrases) (45.5%). Given that Combination 4 can be interpreted as a personally oriented pattern, its preference by parents suggests that parents appear to give more weightto personal involvement. This appeals to the recipient's (children's) interest, thus establishing a personal warrant for undertaking the closings smoothly. On the other hand, the parents' pattern for organising the closing sequence also reflects their attendance to status differences and the seriousness of closings. This is evident from parents' frequent use of 'reassurances

236 about oneself to initiate the closing and their completion of the closing with an expression of good wishes such as 'Zongzhi, jiali yieqie dou hao, wunian. zhu jiankang jinbu. ' (On the whole, everything is well with the family, so don't worry. Wish you good health and progress.). The combination of the 'summarising' reassurance with some concrete good wishes not only emphasises parents' affection for their children, but also indicates a level of authority over them. The 'summary statement' here appears to stress parents' dominant position by presuming that they are responsible for the whole family.

Table 6.9. Frequency and distribution of combination patterns of closings across social roles (N=200) . Combination Social roles Total patterns Family Non-family Parent Child Sibling Relative Friend Teacher Student (11) (16) (41) (12) (92) (5) (23) 200 1. Pre-closing + 2 9 22 7 42 5 13 100 expressions of (18.2) (56.3) (53.7) (58.3) (45.7) (100) (56.2) (50) good wishes /sign-off phrases 2. Pre-closings + 2 0 3 1 14 0 0 20 other expressive (18.2) (0) (7.3) (8.3) (15.2) (0) (0) (10) act 3. Pre-closings + 2 3 10 4 29 0 2 50 other expressive (18.2) (18.8) (24.4) (33.3) (31.5) (0) (8.7) (25) acts+ expressions of good wishes /sign-off phrase (+other expressive act 4. Other expressive 5 4 6 0 7 0 8 30 act + expressions (45.5) (25) (14.6) (0) (7.6) (0) (34.8) (15) of good wishes /sign-off phrases Total 11 16 41 12 92 5 23 200 (100) (100) (100) (100) (100) (100) (100) (100)

However, Combination 1 is used by most social roles, including children (56.3%), siblings (53.7%), relatives (58.3%), friends (45.7%), teachers (100%), and students (56.2%). That 'pre-closings + expressions of good wishes/sign-off phrases' is the prevalent combination pattern among most social roles appears to indicate that Chinese people have a strong tendency towards using highly conventionalised patterns for organising the closings of their personal letters.

237 Despite the overall preference for Combination 1 by most social roles, the data shows that the linguistic realisation of each component (i.e. pre-closings, good wishes or sign off phrases) in this sequential pattern varies between different social groups. Pre closings can be taken as the first example. The data shows that relatives (5 out of7) are more likely to start the closing with a direct announcement and without any mitigatory device, such as 'Biewu tashi, xibi.' (There is nothing else. I'll stop writing.). Children (5 out of 9), siblings (13 out of 22), friends (26 out of 42), teachers (3 out of 5) and students (10 out of 13) often initiate the ending with a mitigated declaration. For example, friends often use 'Haole, jiuci tingbi.' (Well, I have to stop here for now), while a teacher is likely to use 'Shijian guanxi, zhan ge ci. ' (Because of time, I have to stop here temporarily.).

It appears that the choice of different linguistic realisations for the pre-closing strategy is dependent on the social role-relationship and the degree of intimacy or distance between writer and recipient. When a bald announcement is used between relatives who are not very intimate in their kin relationship, it may emphasise explicitness and to some extent imply distance. When an initiating-ending statement mitigated by a discourse marker, such as 'haole' (well), is employed between siblings or friends, it seems to reveal the writer's intention of avoiding abruptness, thus ensuring that the recipient's psychological comfort is not at risk. By starting a closing with reasons, which are often related to lateness or lack of time, teachers and students seem to signal to their recipients deference and different levels of social distance. Borrowing Placencia's words, providing a reason for stopping contact may allow the closing to begin "without causing any embarrassment or ill feelings" (1997: ()7).

The influence of social roles and the degree of intimacy or distance upon the choice of closing strategies can also be observed in the terminating part (good wishes or sign-off phrases) of Combination 1. For instance, the sign-offphrase, such as 'cizhijingli' or 'ci li, 'is preferred by children (5 out of 9) and students (7 out of 13). 'Expressions of good wishes' are prevalent among siblings (20 out of 22), friends (39 out of 42), relatives (7 out of 7), and teachers ( 4 out of 5). It appears that the conventional sign-off formula is often used by intimate and distant juniors who are in an asymmetrical relationship, while good wishes are often used by distant seniors, intimate equals, and both seniors

238 and juniors in the relative circle. Given that the conventional sign-off phrase constitutes a formal ritual that conveys some deference, its preference by children and students seems to indicate a deferential orientation and status difference. Good wishes are treated as a general means of achieving a happy ambience and harmonious communication between writer and recipient. Its prevalent use among non-family seniors, intimate equals, and relatives demonstrates that these groups tend to focus more on the confirmation of familiarity and solidarity in relations, rather than on the status differences.

There are also differences between social roles in their choice of expressions of good wishes. The data shows that good wishes sent by teachers to students are restricted to a few topics, such as 'improvement' or 'smoothness,' while intimate equals or relatives send good wishes of various contents, such as 'good heath,' 'good luck,' 'happiness,' 'smoothness,' or 'improvement.' It seems that the choice of topic and the length of good wishes are related to the degree of intimacy or distance. Where a relationship is intimate either through family bonds or friendship ties, the choice of topics is wider and the pattern is intense. In contrast, in the case where a relationship is socially distant, the topics tend to be narrower and the pattern simpler.

6.4.1.3 Discussion This section has examined the influence of social roles on the choice of closing types, the selection of specific strategies for a single closing and the sequential patterning of closings. Findings show that between social roles there are differences in the performance of these three actions. With respect to the choice of closing, the obvious difference is that seniors in asymmetrical relationships (excluding senior relatives) use the single closing the most, while juniors in asymmetrical relationships (excluding junior relatives) use it the least, although both groups show an overall preference for the combined type. Such differences, as with openings, closely reflect the manifestation of authority by seniors and filial piety and deference by juniors in Chinese verbal interaction.

With regard to the choice of strategies within the single closing type, the findings reveal that the 'expression of good wishes' is related to cultural values of good health,

239 happiness, improvement and so forth and is prevalent among most social roles. The formal sign-off phrase stressing deference and formality is more appropriate for juniors, especially in the non-family setting. The overwhelming use of good wishes as a single closing by the majority of writers in various social roles reflects a wider acceptance of luck talk in the closing section of letters as a significant event to "create balance and harmonize communication behaviour" (Fong 2000: 233). For these Chinese writers, good wishes are not only symbolic in meaning, but the use of these luck expressions in closings may increase the recipients' good fortune in life. Therefore, most writers try to use this strategy when using the single closing. The use of the formal ritual (the sign off phrase) which focuses on deference by juniors can be interpreted by the fact that juniors in Chinese society should show deference and formality to seniors, in order to "preserve good human achievements and traditions" (Yang 1989: 125).

The possible motivations for the favouring of the personally oriented single closing strategy by intimate seniors and equals might be related to the attendance to 'expressive ties' in Chinese society. According to Hwang's model (1983), interpersonal relationships in Chinese society can be classified into three main categories on the basis of their expressive ties and instrumental components. The most important expressive ties are the relationships between members of a family and between friends. In order to maintain and enhance these relationships, filial piety to seniors, warm-heartedness towards the young, kindness to brothers and sisters, and demonstration of friendship to friends are greatly encouraged. Given this Chinese traditional background, it is quite understandable that the intimate seniors and equals employ personal and emotional acts in the single closing.

As for the differentiation between social roles in the choice of the combination pattern, the findings show that the personally oriented combination type is appropriate for family seniors, while the highly conventionalised pattern is appropriate for intimate and social equals, intimate and distant juniors, distant seniors, and relatives. The occurrence of these phenomena appears to be associated with two key factors: the Confucian ethics of parental behaviours and the attitude of individuals towards a particular highly conventionalised routine.

240 In their speech styles Chinese parents' strategies and sequential patterns are characterised by the interaction of authority and affection. This is a reflection of the influence of the Confucian principles of the human relations, in which the behavioural rules for parenthood are centred on "supervising and loving the junior" (Ebrey 1991: 46). Considering the Chinese cultural values regarding the father-son relationship, it is understandable that parents prefer more personalised patterns in closings.

As for the overwhelming use of the highly conventionalized pattern of closings by various social roles, Hymes has pointed out that the conventionalised routine "includes identification of idiomatic units, not only greeting formulas and the like, but the full range of utterances which acquire conventional significance, for an individual, group or whole culture" (1962: 38). According to Tannen and Oztek, combinations of speech acts become associated in people's minds and are often repeated in sequence (1981: 39). Coulmas suggests that through using commonly accepted routines, an individual is able to affirm his membership of a socio-cultural group (1981: 10). These views are relevant in explaining the overwhelming preference of various groups of Chinese letter writers for the highly conventionalised sequential pattern of closings. As a closing pattern, the structure 'pre-closings + expressions of good wishes/sign-off phrases' is characterized by its standardised accomplishment of 'proper initiation' and 'proper termination.' It thus seems to have a strong potential for 'socialisation' (normal participation in social interaction) and 'standardisation' (being suitable for most members of society) (Coulmas 1981: 7). Therefore, this sequential structure is capable of handing most situations, and it is natural that various social role groups in the Chinese context can share this pattern. The adoption of this pattern by Chinese writers can be seen as the "long persistence of a formula in a community" (Ferguson 1981: 32).

There are also differences between social roles in their preferences for different linguistic realisations of specific closing strategies within the same combined pattern. The findings indicate that for the pre-closing statement, the unmitigated announcement of a closing is appropriate for senior relatives. The mitigated declaration, preceded by a discourse marker or a reason, is common for intimate and social equals, and distant seniors. For the terminal expressions, family and distant juniors' choice of patterns suggest a tendency towards deference, while both family and social equals and distant

241 seniors' choice of patterns suggest an orientation towards a general sense of familiarity or solidarity. The different linguistic realisations of a specific strategy within one type of pattern demonstrate the influence on the language of status and intimacy or distance. Similarly, these features result from the influence of Confucian ethics on people's verbal interaction.

6.4.2 Closings and age/generation This section examines the impact of the factor of age/generation on the closing behaviour. It aims to find out the extent to which the factor of age/generation affects the choice of closing types, and especially the sequential way in which the closings are organised.

6.4.2.1 Closing types and age/generation

Table 6.1 0. Frequency and distribution of closing types across age/generation (N=259) Types Age/Generation Total Family Non-family DG: 55 SG: 53 DG: 36 SG:115 259 0-Y Y-0 0-Y Y-0 0-Y Y-0 0-Y Y-0 SA (34) (21) (26) (27) (10) (26) (61) (11) (43) (259) I. Single 17 4 3 4 5 3 12 2 9 59 closings (50) (14.3) (11.5) (14.8) (50) (11.5) (19.7) (18.1) (20.9) (22.8)

2. Combined 17 17 23 23 5 23 49 9 34 200 closings (50) (85.7) (88.5) (85.2) (50) (88.5) (80.3) (81.8) (79.1) (77.2)

Total 34 21 26 27 10 26 61 11 43 259 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: drfferent generation; SG: same generation; 0-Y: older to younger; Y-0: younger to older; SA: same age

Table 6.10 shows the relation between age/generation and the choice of closing types. It seems that the age factor does not affect older writers' choice of closing types in both family and non-family letters (50% vs. 50% and 50% vs. 50%). However, the situation for writers of other ages presents another picture. The younger writers within different generations and writers of different ages in the same generation all have a great preference for the use of the combined type of closing.

242 6.4.2.2 Combined closings and age/generation

Table 6.11. Frequency and distribution of combination patterns of closings across age/generation (N=200) Combination Age/Generation Total patterns Family Non-family DG: 34 SG: 46 DG: 28 SG: 92 0-Y Y-0 0-Y Y-0 0-Y Y-0 0-Y Y-0 SA (17) (17) (23) (23) (5) (23) (49) (9) (34) 200 1. Pre-closings 7 10 12 11 5 13 19 3 20 100 + expressions (41.2) (58.8) (52.2) (47.8) (100) (56.5) (38.8) (33.3) (58.8) (50) of good wishes/sign- offphrase 2. Pre-closings 3 0 1 2 0 0 8 3 3 20 +other (17.6) (0) (4.3) (8.7) (0) (0) (16.3) (33.3) (8.8) (10) expressive act 3. Pre-closings 2 3 5 9 0 2 15 3 11 50 +other (12.5) (16.7) (21.7) (39) (0) (8.7) (30.6) (22.2) (32.3) (25) expressive act+ expressions of good wishes/sign- off phrases (+other expressive act) 4.0ther 5 4 5 1 0 8 7 0 0 30 expressive (31.2) (22.2) (21.7) (4.3) (0) (34.8) (14.3) (0) (0) (15) act+ expressions of good wishes/sign- off phrases Total 17 17 23 23 5 23 49 9 34 200 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) DG: different generation, SG: same generation; 0-Y: older to younger; Y-0: younger to older; SA: same age

As shown in Table 6.11, some differences emerge in the distribution of the combination patterns between writers of different ages/generations in both family and non-family settings. Of the different generations in the family setting, older writers exhibit a stronger inclination to use Combination 4 than younger writers (31.2% vs. 22.2%), while younger writers show greater preference for the use of Combination 1 than older writers, although the difference is not very significant (58.8% vs. 41.2%). However, there is the converse occurrence in letters between different generations in the non-

243 family setting: older writers are more likely to use Combination 1 than younger ones (100% vs. 56.5%), while Combination 4 is appropriate for younger writers (34.8%).

Age differences between persons of the same generation seem to have a more complicated influence on the choice of sequencing patterns. In the family setting, the tendency for the younger to use Combination 3 is slightly stronger than it is for the elder (39% vs. 21.7%), while the elder are more likely to use Combination 4 than the younger (21.7% vs. 4.3%). In the non-family context, writers of the same age are more likely to use Combination 1 than the elder and the younger (58.8% vs. 38.8% and 33.3%); the younger use Combination 2 more than the elder and the same-aged (33.3% vs. 16.3% and 8.8%); and the use of Combination 4 is limited to elder writers (14.3%).

6.4.2.3 Discussion In the light of the findings described above, it is clear that age/generation is a relevant determinant in the closing performance, in regard to the choice of closing types and the organisation of the sequencing. For the closing types, the findings demonstrate that in both family and non-family circles, the elder are much more likely than the younger to adopt the single type of closing. Given that this type of closing reflects a minimal attendance to the requirements of mitigation and consolidation, its prevailing employment by elders reflect the fact that the closings of letters to the young are less social and less face work-oriented. This can be interpreted as the older writers' way of preserving their seniority and authority in verbal interaction. As discussed in chapter four and five, it is generally accepted that age is a symbol of seniority and power (authority) in the Chinese cultural context (Shih 1986: 37). Since seniority and authority are a warrant for the maintenance of one's face and one's powerful image, it is expected that the elder are more likely than the younger to perform a single type closing.

The impact of age/generation on the choice of sequential patterns in the combined type closing is much more complex. With regard to the different generations of the family setting, Combination 4 (the personally oriented pattern) is used more by the old than the young, while Combination 1 (the highly conventionalised pattern) is used more by the young than the old. However, these preferences are reversed in the non-family setting. The answer for this lies partly in the way in which age is perceived differently in these

244 two settings. The respective preferences of the young and old in the family context are due to the different orientations of the two sequential patterns. The old tend to be more personal in letter closings, so as to emphasise their affection and their somewhat senior position. The young tend to write in a more standardised way so as to preserve good conventions.

It is interesting that the opposite is true in the non-family setting where the preference for the personally oriented pattern is stronger among the young than among the old. This is an indication of the relation between the choice of language and settings. As analysed above, a young person (e.g. student) usually initiates a closing with an expression of thanks or apologies and then terminates it with good wishes of a sign-off phrase. By doing so, the young writer intends to emphasise a specific personal feeling which encodes deference. It is clear that such a sequential structure emphasises respect towards the older recipient, aiming to affirm or reaffirm the interpersonal relationship. This fact can be taken as further evidence of the relevance of age in the choice of closing sequential structures.

In regard to writers of the same generation, the influence of age also seems to be different in family and non-family letters. It is found that in family letters elders prefer to use the personally oriented pattern while the young prefer to use the most intense combination, although both groups demonstrate a preference for the highly conventionalised pattern. To explain this result, it may be hypothesised that the intense combination pattern is more politeness-oriented because it shares more common ground and feelings. There is a need for the young to send a strong social message and strong feelings to the elder, thus emphasising their solidarity and respect. On the other hand, the preference for the personally oriented pattern by the elder may show their attention to emphasising personal feeling in the initial position of a closing, stressing closeness and some degree of seniority in age.

The complex tendencies ofwriters within the same generation in the non-family setting are also worth noting. The frequent occurrence of the personally oriented pattern in closings written by elders may reflect their concern with personal bonds. In this respect they are similar to the elders of the family setting. The favouring of the highly

245 conventionalised pattern by writers of the same age group may be interpreted as a way of encouraging mutuality and solidarity, and maintaining the integrity of one's group (Bond & Hwang 1989: 259). The preference of the younger group for Combination 2 (pre-closings + other expressive act) can be seen as the following of a middle-path, halfway between conventional patterns and personal patterns. However, given the small sample number for this group, this tendency may be only suggested. Based on the evidence, it can be suggested that writers' motivations for choosing the type of combination in this category appears to be related to their perception of what is necessary to emphasise their desire to maintain the relationship.

6.4.3 Closings and gender This section explores the impact of the social factor of gender on the closing behaviour of Chinese personal letters. This examination centres on how gender factors affect the choice of the closing types and the manifestation ofboth single and combined closings.

6.4'.3.1 Differences in the choice of closing types As seen in Table 6.12, a majority both of females and males prefer combined closings over single ones. However, a further examination indicates that the occurrence of each closing type differs between the genders. In both family and non-family contexts, females have a stronger preference than males for the combination type of closing (family: 81.1% vs. 67.3%; non-family: 85.9% vs.7 3.8%). Males are more likely than females to use single closings (32.7% vs. 18.9%; 26.3% vs. 14.1%) although the validity for the latter trend is limited by the sample number. Nevertheless, this trend is consistent with that of openings, where males are more likely than females to use the single type. It is also found that gender differences are interconnected with the factor of age/generation. For instance, it seems that the single type of closing is more suitable for senior males in both family (13 out of 18) and non-family letters (12 out of21).

246 Table 6.12. Frequency and distribution of closing types across gender (N=259) Types Gender Total Family Non-family F:53 M: 55 F: 71 M: 80 259 N % N % N % N % N % 1. Single closings 10 I8.9 18 32.7 10 I4.1 2I 26.3 59 22.8 2. Combined 43 81.1 37 67.3 61 85.9 59 73.8 200 77.2 closings Total 53 100 55 100 71 100 80 100 259 100

6.4.3.2 Differences in the choice of the strategy of single closings Table 6.13 presents information regarding the two genders' selection of single closing strategies. Overall, there is no significant difference between females and males in their selection of most of the single strategies. The only obvious differences are in the selection of two strategies: females are more likely to select the strategy of 'longing for future contact,' and males are more likely to select the sign-off phrase. These tendencies indicate different orientations in the conduct of the single opening, although the evidence is scanty. Given that the 'longing for future contact' emphasises the writer's desire for the continuation of the relationship, this finding may suggest that females tend to use a relationship-oriented strategy. In contrast, the use of the sign-off phrase reveals that the writer's focus is on the status of the recipient and himself. This suggests that males are inclined to use a status-oriented strategy.

Table 6.13 Frequency and distribution of the strategy of single closings across gender (N=59) Strate1zies Gender Total Family Non-family F: 10 M: I8 F: 10 M: 2I 59 N % N % N % N % N % I. Expressions of wishes 5 50 10 55.6 6 60 I4 66.7 35 59.3 2. Pre-closings 2 20 5 27.8 1 IO 2 9.5 9 15.3 3. Sign-off phrases 0 0 1 5.6 0 0 2 9.5 4 6.8 4. Longing for future contact 2 20 1 5.6 I 10 0 0 4 6.8 5. Expressions of concern I 10 I 5.6 0 0 0 0 2 3.4 6. Expressions of thanks 0 0 0 0 1 10 1 4.8 2 3.4 7. Apologies 0 0 0 0 1 10 I 4.8 2 3.4 8. Offering help 0 0 0 0 0 0 I 4.8 I 1.7 Total 10 100 18 100 10 100 21 100 59 100

6.4.3.3 Differences in the choice of combination patterns of closings As shown in Table 6.14, it is clear that no significant differences are found between the genders in their choice of the four sequential patterns of closings. This finding is very

247 interesting, as it is in contrast to the findings on the genders' selection of opening sequential patterns. In the case of openings, females are more likely to choose personally oriented sequential patterns, while males prefer status-oriented sequences. Based on these results, it is suggested that in performing linguistic behaviours females are not always more affective than males.

Table 6.14. Frequency and distribution of combination patterns of closings across gender (N=200) Combination patterns Gender Total Family Non-family F:43 M:37 F:61 M:59 200 N % N % N % N % N % 1. Pre-closings + 22 51.2 18 48.6 30 49.2 30 50.8 100 50 expressions of good wishes/sign-off phrases 2. Pre-closings + other 3 7 3 8.1 6 9.8 8 13.6 20 10 expressive act 3. Pre-closings+ other 11 25.6 8 21.6 16 26.2 15 25.4 50 25 expressive acts + expressions of good wishes/sign-off phrases (+other expressive act 4. Other expressive act 7 16.3 8 21.6 9 14.8 6 10.2 30 15 + expressions of good wishes/sign-off phrases Total 43 100 37 100 61 100 59 100 200 100

6.4.3.4 Discussion The investigation of gender-related closing behaviours reveals that the gender difference does to some extent affect Chinese writers' performance of closings in personal letters. First, as with openings, there are significant differences in the choice of closing types: Females are more likely to choose combined closings, while males are more likely than females to choose the single type. The different preferences of the genders reflect their different perceptions of what is appropriate for a closing in the Chinese context. The combined type of closing can be seen as socially and ritually sufficient because it meets the two basic requirements for a closing: mitigation and consolidation, and it also achieves the need to be polite. It can thus be argued that the more frequent use of this type by females reflects their perception of the need to convey more social messages and personal feelings, aiming to achieve appreciation and assurance as to the continuation of their relationship. In contrast, the single type of

248 closing contains fewer social messages and thus shows less attendance to the recipient's face needs (e.g. being interested or caring). Thus, it can be suggested that the extensive use of the single closing by males demonstrates the orientation of this group: caring more about their own status or social position. The overall preference between genders in this particular respect appears consistent with the widespread view that women are more concerned about others and more polite in social interactions (Gal 1995; Herbert 1990; Holmes 1995; Preisler 1986; Smith 1992; Weatherall2002).

The repeated occurrence ofthese differences in ways of interacting with others, such the contrast in the choice of closing types, can be interpreted as an expression of psychological differences between genders. Women tend to pay more attention to connections with others and to involvement in the interdependencies between people, while men are more likely to be oriented to autonomy and detachment and to hierarchical relationships (Boe 1987; Holmes 1995). The second explanation is linked to social and cultural factors, the different social roles and status of the two genders in traditional Chinese society. Women are required to be more polite in speech because of their lower social status, while the verbal behaviour of men is not expected to involve so many prestige elements because of their higher social status. Thus, as suggested by Weatherall, the phenomenon of women's speech differing from men's can be understood as an effect of the cultural beliefs surrounding gender differences and male dominance (2002: 81).

Second, it is suggested that gender differences also influence the choice of specific closing strategies, particularly in the manifestation of the single closing. Women tend to use the personal and emotional-oriented acts (e.g. longing for future contact), while men tend to use the status-oriented strategies (e.g. sign-off phrases), which is similar to the case of openings. The distinct preferences of the genders in this respect can be explained by the different features of the nature of the two genders and their different focuses in social interactions.

Third, it is important to note that gender differences might be obscured by the influence of communicative goals of constructing a sequence in a particular location. This fact arises when choosing combination patterns in performing a combined closing. Males

249 and females have very similar preferences for each sequential pattern. Actually it is difficult, even impossible, to interpret this phenomenon according to the psychological linguistic approach and the gender dominance approach. The tentative answer lies in the assumption that "there is no one-to-one gender-based mapping of linguistic form on to any communicative function" (Weatherall2002: 74). According to Weatherall (2002: 95), gender is not a stable and enduring feature of the individual that can be reliably and transparently reflected in every speech feature. Instead, many speech features have social meanings rather than gender differences. This view is suitable for interpreting the indistinct differences between the genders in their choice of combination patterns in closings. As was examined previously, the primary functions of closings, which are served by a range of phatic expressions, are the mitigation of the potential face risk and the consolidation of the relationship (Laver 1975: 236). A closing has these functions because of its special communicative context, that is, the end of a letter. From the sociological point of view, there appear to be strong social meanings underlying the closing. In this situation, Chinese women are likely to be similar to Chinese men in the proportion in which they use combination patterns. Thus, it can be suggested that the investigation of gender differences in speech should take into account other factors, such as the goals of the communication. The perception that women's speech styles are co-operative and emotional, and men's speech styles impersonal and competitive needs to be understood in a general sense that relates these gender differences to the specificities of particular linguistic behaviours. This does not mean that one should abandon the search for gender differences in language styles. Instead it implies that the study of gender-based variables should include the examination ofparticular linguistic behaviours, as well as communicative goals and styles.

6.5 Concluding remarks This chapter has examined the closings of Chinese personal letters. Based on the analysis on the empirical data, several concluding points can be made. First, in a similar way to their performance of openings, Chinese writers employ a series of expressive acts (with few exceptions) to accomplish closings. The purposes of using these speech acts are to mitigate potential feelings of rejection that may arise from initiating the end of written contact, and to reaffirm and consolidate the relationship between writer and recipient. Second, good wishes and pre-closings seem to be the two

250 core strategies of a closing, and this reflects Chinese writers' strong orientation towards attending to the harmony of their social relationships with recipients and to the recipient's face need in the ending section of letters. Third, the combined type of closings occupies the dominant position in the performance of closings, which shows a tendency towards intensified componential constitution. This structure allows for the greater fulfilment of the two major functions embedded in these combinations. Fourth, all the sequential patterns of closings contain two major organisational parts: the initiating part and the terminating part. Each part is realized through certain phatic linguistic measures. In addition, there can be some rather lengthy series of strategies used in the terminal statement, which is seen as a way of offering greater reassurance as to the continuation of the relationship. Finally, various social factors including social roles, age/generation, and gender influence the closing performance in their different ways. This confirms the argument that closing an interaction is a matter not only of stopping the conversational machinery but is also, and more importantly, a matter of dealing with a social encounter which is culturally specific and related to social roles, age, and gender (Placencia 1997; Pavlidou 2000).

251 CHAPTER SEVEN CONCLUSION

7.0 Introduction This study has set out to examine the forms and functions of address, openings and closings in Chinese personal letters and to examine to what extent social roles, age, gender and cultural contexts affect writers' language choices. The aim has been to provide a better understanding of the conceptualisations and applications of certain written speech behaviours and intercultural communication in the Chinese context. Both qualitative and some simple quantitative techniques have been used to examine the language and the social meanings of the three behaviours of the authentic discourse. Some answers have been given to the research questions. There are some questions that can be refined in appropriate ways for future research. This chapter attempts to draw conclusions by summing up major findings in relation to research questions and by pointing out the theoretical and practical implications of these findings. It then presents the limitations ofthe study and suggests directions for future research.

7.1 Major findings of the study 7.1.1 Address forms as multiple indicators Detailed analyses show that in personal letters Chinese writers use a range of forms, including name-alone strategy, compound name strategy, and no-name strategy, to address recipients. The addressing practices in this kind of written communication are characterized by the diversity and elaboration in the Chinese context. In addition, the choice and the distribution of address forms are very much related to the status (i.e. age, social rank) of the writer and the recipient, the state of the interpersonal relationship between them, gender differences and cultural constraints as well. This phenomenon further confirms that address forms are one of the perspectives which can best reflect the close relation between language and society (Shih 1986: 98). The analysis also shows that the written addressing practice, as with that of conversation, has its own rules, which may differ from those of other domains. One obvious feature is the use of kinship terms for non-kin, which is restricted to circles of friends in which one of the participants is either senior or junior. It is important to stress that this extension of kinship terms to refer to non-kin is a means of emphasizing closeness, sympathy or

252 deference. This is intended to establish and maintain 'family solidarity' between writer and recipient, which is considered as one of the typical characteristics of Chinese (Wu 1990: 62).

Another feature which is also specific to the social rules of written address in personal letters is the elaborate address behaviour. This is the case in Chinese where name, kinship term, title or a complex is often modified by a certain positive term (e.g. intimate, diminutive, deferential term). This way explicitly reflects a practice of indicating both social status and the relationship between writer and recipient. The new strategically mixed alternatives can be characterized as especially positive, since they allow for some increase in linguistic politeness.

As far as the relationship of social factors and linguistic variations of address forms is concerned, the results of this study reveal that social factors have a significant effect on how Chinese choose address forms. First, social role and age interact closely with the address usages. Statistical evidence shows that the address behaviour of seniors and juniors in hierarchical relationships is asymmetrical. Seniors receive the deferential form or a form mixing deference and intimacy. They give back informal or formal address forms according to personal circumstances. Interestingly, sometimes the Chinese written address system permits seniors to contract "the psychological distance" (Hirijida & Sohn 1983: 147). This is often done by attaching a kinship term or a relation term to the name toward juniors, emphasizing closeness and shortening of distance.

The matter of address usage within people of the same generation is more complicated than that between those of different generations. Specifically, the effect of age difference overrides that of the symmetrical relationships (e.g. brotherhood, friendship) and its influence varies with settings. Family members of the same generation, especially siblings, use both non-reciprocal and reciprocal address forms: Elder siblings often use an informal or a solidary term while younger siblings often use a deferential and a solidary term. Non-family members of the same generation (i.e. friends), in contrast, do not exhibit fundamental differences in the choice of address forms, but there are still differences due to age variations. Both elder and the same-age writers

253 prefer to use the formal name to express familiarity, while younger writers favour the use of an informal form, indicating intimacy. Thus, it is interesting to note that variations in the choices of the writers of the same generation are caused by the different role and age hierarchies of family and non-family relationships.

Second, gender differences also affect the writer's use of address forms. In general, females demonstrate a preference for an intimately modified address term, particularly downwards towards younger females in both family and non-family letters. Males tend to choose informal names for younger females or males in family letters, or a term mixing informality and brotherliness downwards for younger males in non family letters. Both sexes are sensitive to the age and sex of recipients and to the settings.

7.1.2 Openings as a means of opening up a channel and establishing the relationship Written openings in Chinese personal letters are achieved by a series of expressive acts in the sense that they open up the communicative channel (Lyons 1977: 53) and establish or maintain the relationship between writer and recipient. Greetings, inquiries and an acknowledgement of receipt of a previous letter are the three key instances of expressives used for phatic functions. These acts are examples of how Chinese accept social restraints and adapt their talk to a speaker's age, rank and interpersonal relationship, by expressing sympathetic concern for others and by preserving sincerity in linguistic behaviours. Structurally, many phatic speech acts are often shaped within the compound framework, which serves to increase the force of the focused message.

The majority of openings in letters comprise systematic phatic sequences. This can be understood as the manifestation of an initial and then of a further display of the status of the writer and the recipient and relationships between them. In addition, the phatic sequences have their own particular properties and these are expressed in an orderly manner in relation to the letter types. For instance, in initial letters most sequences are initiated by a greeting, followed by inquiries, or an expressive act, or a combination of inquiries with an expressive act. Few are started by inquiries followed by an expressive act, which emphasizes both authority and affection, especially in family

254 letters. In reply letters the greater number of sequences is initiated by a greeting, and furthered either by an acknowledgement of receipt of a previous letter, or a combination of this strategy with an expressive act. At times the writer may begin the opening with the acknowledgement then add an expressive act afterwards. This can be also interpreted as a way of emphasizing the identification of the relationship, intermingled with convey of higher status ofthe writer, especially by family members.

Two of Grice's maxims (1975) have been found to be relevant for the structures exhibited in these sequences: the Maxim of Quantity and the Maxim of Relevance. In adding a single act or a combination following the routine of greeting or the exceptions (inquiry or acknowledgement as initiators), writers make an effort to enter into the relevant and clear framework of the whole discourse. A writer's first speech act and then their second (possibly also third) represent a movement from an initial to a full display of status and relationship. The arrangements of certain speech acts or routines require the writers to perceive the extension as an action necessary for efficient communication. On the other hand, such phatic sequences can also be seen as a way of expressing greater politeness, as writers share more common ground and experiences with recipients (Brown & Levinson 1987).

The analysis of the openings also suggests that the way in which writers have different orientations to structural aspects is due to variations in social roles or age. The differences mainly rest on two aspects: first, differences on the strategy type level; and second, differences concerning the orientation towards the patterning of sequences. It is important to note that among the writers examined the differences between these two aspects are not balanced, and the variations weigh differently according to the differences of generation and settings. For the strategy type, seniors of a different generation use the single strategy for openings more often than their juniors. One way of explaining this differential distribution between the single and the combined strategies is to link it with the cultural orientation of formality and respect from juniors towards seniors. It is interesting to find that the choice of strategy type by writers of the same generation vary between settings. Both seniors and juniors of this category in the family circle do not show differences in the choice of strategy type for openings. In contrast, different orientations can be found within non-family members (i.e. younger

255 friends use the single strategy type less than the same-age and the elder ones although the evidence is scanty). The differences shown in the usage between the same generation people of family and non-family may be related to the degree of certainty /uncertainty (Wolfson 1988) writers perceive in their relationships with recipients. It is as though the younger persons of the non-family setting are more uncertain about each other than aged ones. Thus, they need to avoid using the single strategy, which seems to be less polite than the combined strategy. As for the matter of choosing a specific act as the single opening, the results reveal that there is no significant difference among the writers of different social roles and ages: They all prefer to choose the act of greeting. Thus, once again, greeting is assumed to be an important accessible device that ensures the writer's entrance into written interaction (Laver 1975: 221).

As for the patterning of sequences, things are more complicated: They involve subcultural or even individual group differences. In the hierarchical relationship, the direct-blood seniors show preference for the pattern emphasizing intimacy and authority, while juniors demonstrate a strong preference in both initial and reply letters for the most intense structure stressing respect intermingled with affection. However, both seniors and juniors in the indirect-blood relations of family (i.e. relatives) and the non-family setting have a similar preference in the choice of sequences. The differences between them are only centred on the selection of the additional expressives following greetings in initial letters or following greetings plus acknowledgement of the receipt of previous letter in reply letters. In addition, the differences can also be found in the manipulation of each strategy in the whole discourse. Respectively, in groupness (for relatives) or familiarity (for non-family people) is stressed for seniors while respect and respect mixed with in-groupness receives more emphasis amongst

JUlliOrS.

The preferences adopted by persons of different social roles suggest that there is a direct association between opening patterning and the degree of authority and affectional distance. Family seniors tend to be highly concerned with impulse control and affection; family juniors tend to emphasize intensity of respect and affection. The overall preferences of both parties in other roles (e.g. teachers, students, both older and younger relatives) in an unequal relationship can be interpreted as formal and

256 distancing, but as successful management of interpersonal communication on different levels. Such explicit patterning might be attributed to the Confucian virtue of propriety (li) which obliges members of a society to adapt their talk to a participant's status, age and intimacy or distance (Hopper & Chen 1996). This attention to relationship issues leads Chinese writers to avoid the potentially face-threatening consequences of expressing excessive intimacy or familiarity outside of the family.

As for the writers m an equal relationship, the patterning of opening sequences embodies another quality. Both elder and younger writers demonstrate a preference for the commonly accepted pattern for both initial and reply letters. This involves recognition of status and relationship, and is linguistically shaped by devices emphasizing solidarity and intimacy. It has been argued that these people's equal preferences for the pattern and the single strategy are due to their orientation towards public compliance concerning harmony maintenance. This reflects the group-oriented value embedded in collectivism, a typical characteristic of the Chinese peers (Bond & Hwang 1989).

There is also variation between the opening behaviours of females and males. Overall, females avoid opening with a single act and make more use of the combination type than males in both family and non-family letters. When choosing the single strategy, females are restricted to a more limited act emphasizing affectional relationship (e.g. inquiry), whereas males use a greater variety of strategies. Such differences shown in this aspect can be explained by the fact that women are likely to focus on affective matters, whilst men are more topic-oriented. As for patterning opening sequences, females show a preference for the relationship-oriented pattern while males are more likely to use the status-oriented pattern, especially in initial letters. Such gender-based variations result from differences in terms of psychological and interactional socialisation between sexes. Similarly, the age factor interplays with gender differences in organizing openings, but its role weighs differently across sexes: the sequencing pattern preferred by females appears to be most appropriate for older people while that for males can be used by both older and younger writers. More markedly, gender differences in structuring openings appear to be restricted to intimates (e.g.

257 family members, friends), which suggests that the relationship between gender and speech style should be investigated within a particular community.

7.1.3 Closings as a means of ending the contact but continuing the relationship As with openings, closings in Chinese personal letters are mostly realized through speech acts of phatic functions, which serve to end the contact and consolidate the relationship so as to continue contact between writer and recipient. In addition, the two features of Chinese written closings, good wishes and pre-closings, represent key phrases in the closing process. This reflects Chinese people's strong perception of the need for offering 'warrants' for smoothly undertaking the end of contact (Schegloff & Sacks 1973) and of emphasizing harmony by luck talk (Fong 2000) at the end of written communication. Moreover, written closings include some speech acts (e.g. apologies, thanks, expression of concern) which can be used in either openings or closings, or in both. Such parallel features seem to have the effect of either making up for not using those strategies in the openings, or of emphasizing a focused meaning (e.g. duplicating wishes). For the latter, it would seem that the use of these constituents, along with their repetitions in the closing section, helps writers to emphasize the reassurance of the relationships, which can be seen as a means to "serve positive face" (Tannen 1989: 52). Furthermore, most speech acts in closings are characterized by being less elaborate in linguistic realization than those in openings. This reduction of elaboration can be explained by the special structuring way in which the routines are organized in the ending location.

Also, closings in Chinese personal letters are mostly achieved through sequences passing through pre-closings and terminal phrases respectively as are those of oral conversation (Schegloff & Sacks 1973). This is intended to mitigate ill feelings of rejection and to enhance the relationship between interactants (Laver 1975). In addition, there are ordered sets of routines, in which writers achieve the above two basic functions. For instance, the majority of closings are initiated by the pre-closing statement and terminated by the expression of good wishes or by a sign-off phrase, sometimes by an expressive act, or a combination of an expressive act with either good wishes or a sign-off phrase. Occasionally, closings are started by an expressive act and ended by expressions of good wishes (or a sign-off phrase). It is interesting to note

258 that both the initiation and termination of closings can be conducted with different strategies. It is also important to stress that there can be some rather lengthy sequences of acts in the production of the terminal expressions. This can be seen as the "dramatization ofrelationship features" (Knapp 1978: 107). However, it is interesting to note that certain elaborations between the pre-closings and the terminal expression seem to be independent on the surface level. There is no structural logical relationship between them and the pre-closings and the terminal statement, which is quite different from those embedded within openings (making something relevant and clear). But even with the unrelated structure, the indication of consolidating the relationship seems to develop from the initial stage to the further one (i.e. to gradually confirm the writer's positive feelings and attitudes towards the recipient).

It is worth noting that, unlike the openings, there is no relationship between structuring sequences and types of letters. The only variations are in the occurrence of certain sequences: The most elaborate pattern is appropriate in initial letters whereas the more personally-oriented structure (other expressive act + expressions of good wishes) is common in reply letters. The explanation for these differences is related to the frequency with which writer and recipient interact, thus linking structure to the degree of intimacy or distance.

The findings also show that writers produce closings differently according to role and age situations. As with openings, the differences are mainly centred on the strategy type and the patterns of sequences. First, for hierarchical relationships, in both family and non-family letters seniors use the single closing more than their juniors. However, for the equal relationship, there is a minor deviation in the choice of the single or the combined strategies between seniors and juniors in age. It could be argued that age does not affect the equal-relation writers' choice of the strategy type very much. This phenomenon could be explained by their similar emphasis on claiming solidarity and equality in particular linguistic behaviours. Another difference has to do with the choice of the specific single strategy. The results show that 'expressions of good wishes' emphasizing the harmony of the relationship are prevalent among most social roles. The formal sign-off phrase stressing deference and formality, however, is preferred by juniors, especially in the non-family setting. Thus, it can be concluded

259 that solidarity, authority and formality are determinants for the degree of deviation from the normative use of language.

As far as the differences in patterning sequences are concerned, the analysis of the results suggests that direct-blood seniors in hierarchical relationships of family demonstrate a preference for the use of the personally-oriented pattern (other expressive act+ expressions of good wishes). By contrast, other people, including family juniors, both seniors and juniors of relatives, family and non-family equals, and non-family seniors and juniors, all exhibit a preference for the use the highly conventionalized closing sequence (pre-closings+ expressions of good wishes). The overall findings in the preferences of the highly conventional pattern by various groups suggest when a routine or a sequential routine has become highly socialized or highly fixed through the ages regardless of its linguistic variations, perhaps such a routine would be acquired by 'whole culture' (Hymes 1962). On the other hand, the preference for the personal oriented pattern by family seniors can be seen to be influenced by the special authority and affection for family juniors in Chinese culture.

It should be noted that although the majority of writers of various roles tend to use the highly conventionalized closing pattern, they differ in the use of language for achieving the pre-closings (e.g. bald vs. mitigated) and terminal expressions (e.g. deferential, familiar, intimate) across roles. The findings suggest that juniors in the hierarchical relationship in both family and non-family settings tend to use deferential-oriented phrases while the others (both senior and junior relatives, non-family seniors and both family and non-family equals) tend to use familiarity-oriented expressions, thus reflecting in-groupness (for both senior and junior relatives), familiarity (for non-family seniors) and solidarity (for equals). Therefore, it is important to stress that the manipulation of each strategy is constrained by the effect of roles and also by the interpersonal relationship. For example, the stylistic choices of relatives (i.e. unmitigated announcement of the end by seniors, or informal expression of good wishes by juniors) reflect their tendency of avoiding acknowledging overt intimacy or deference respectively. This implies a non-immediate family relationship between relatives.

260 The findings also reveal that gender factors influence writers' performance in closings, but the influence weighs differently in different aspects. First, as in openings, females use the combined strategy of closing more than males, and males use the single closing more than females. Thus, female writers are more adaptable to emotional and polite investment (more social messages are conveyed and more face work is done) than males. In addition, while choosing a specific strategy as the single opening, females use more informal and relationship-oriented expressions (longing for future contact) while males use more formal and status-oriented phrases (sign-off phrase). Thus, these stylistic differences reflect female and males' respective psychological and cultural orientations towards interaction. This explanation is favoured by a number of studies (Holmes 1995; Lee-Wong 2000; Malts and Borker 1982; Tannen 1990, 1994,).

The most important finding regarding the influence of gender on closings is that neither females nor males show obvious differences in the choice of the closing sequential patterns. The differences are only found in the choice of a specific expressive act within the relevant patterns. Females often prefer to explicitly emphasize their emotional bonds with recipients by asking for future contact, whereas males often prefer to stress their status by offering help or expressing apologies. One way of interpreting the similar distribution in organizing the sequences with certain strategies is related to the assumption that many speech behaviours have social meanings rather than gender differences. This is due to the special communicative features (a specific purpose in a specific location) of these speech behaviours. Thus, it has been argued that the investigation of gender differences should include examination of the broad communicative goals as well as the individual linguistic behaviours of gender.

7.2 Significance of this study The preceding data-based analysis of the linguistic realization of address forms, openings and closings in Chinese personal letters represents a contribution to the ongoing research on written interaction in the Chinese language. This empirical study is likely to be significant for sociolinguistic theory, language teaching, and cross cultural communication as outlined below.

261 7.2.1 Implications for sociolinguistic theory The findings of this study have considerable benefits for sociolinguistic research, specifically in the area of speech behaviours but also for research on written interaction in general. They provide detailed information about how names and other expressions are used in addressing, how letter writing is started and terminated linguistically and ritually, and how the ritual and formal patterns interact with social variables. This contributes to a better understanding of the operation of those behaviours in the written communication of Chinese society. This study will be a contribution to cross-cultural pragmatics.

As has been mentioned throughout, addressing, opening and closing behaviours have been discussed extensively in spoken discourse. However, people know little about how they actually work in written communication, particularly in personal letters. The empirical results from this study provide some explanation for these phenomena in authentic discourse in a systematic way.

The lexical and pragmatical analyses uncover the actual complexities of address forms in written interaction, which is a reflection of the diversity and elaboration of address form use in Chinese. The analyses also reveal the flexibility of address form use between peers in Chinese written communication: the use of kinship terms for non-kin. This can be seen as a strategic use of address forms, which tends to encourage greater solidarity. In addition, they methodologically confirm that the address usage emphasizes the differentiation in roles, age, gender and intimacy or distance. Typically, the effect of age is so firm that, for example, an adult, even if he is a lecturer at a university, or a general manager of a company, addresses his elder brother or sister with the deferential term, even within a same generation relationship. Moreover, seniority is not always seen as a rigid vertical construct. Seniors can strategically use informal forms intermingled with affection or solidarity downwards juniors, emphasizing intimacy or shortening distance. Such findings show the importance of this study for sociolinguistic theory in terms of lexical and pragmatic functions in address use, thus contributing to our understanding of the complex realities of the Chinese written addressing practices in the present era.

262 The syntactic and pragmatical approaches bring to light the structural properties of openings and closings in written discourse. As is known, openings and closings in oral conversation (i.e. face-to-face, or telephone) are organized within the two-part joint activity and they respectively serve to achieve certain social functions, such as identification and recognition (Schegloff 1979), and mitigation and consolidation (Laver 1975, 1981). However, openings and closings in personal written interaction are elaborate (with few exceptions) consisting of both 'formulae' and 'routine' speech acts used by writers to achieve certain social functions. For instance, in openings, sequences are based on how a writer indicates the status of the recipient and himself/herself and the relationship between them both initially and further. The basic form of initial-indication is the greeting (occasionally inquiry), followed by certain acts of further indication. In addition, the sequential acts of its organization have both structural and pragmatic relevance, a relevance based on Grice' maxims of Relevance and Clarity, which enable the writer to display an identification sufficiently and logically on different levels.

When examining closings, it is interesting that Schegloff and Sacks' overall structural framework and Laver's phatic functional approach can account to a considerable degree for written closings in letters, although the behaviour is accomplished only by one party, the writer. More specifically, in closings, sequential organization is deployed in accordance with the natural flow of pre-closings, and followed by the termination of written contact. The pre-closings are always achieved by the pre-closing phrase or expression. Some expressives (excluding good wishes and sign-off phrases) can be also used to pre-close as well as terminate the written communication. By using a particular sequence, the writer tends to mitigate the face threat which may result from the initiation of the end of contact and tends to reinforce the interpersonal relationship with the recipient. Structurally, the sequential acts are constructed as an independent organizational format on the surface level.

Taken together, these findings show first that the openings and closings in personal written communication are constructed by a sequencing of routine speech acts and second that there is a close link between sequencing and the status of the participants' relationship. Therefore, this research seems to make a substantial contribution to the

263 understanding of the properties and the sequence of speech events m written communication in general.

This study also highlights the need for studying speech acts, phatic communion and politeness within the domain of discourse. On the one hand, it shows that speech act theory can be extended to discourse analysis. As we can see, the speech acts which work as constitutive components of openings and closings in this study are not isolated moves. Rather, they appear in more global units of discourse, and carry phatic properties in an attempt to define and maintain relationships. For example, an act of greeting is a component of openings. In general, they function as an access ritual (Goffman 1971) and mark status and relationships (Firth 1972; Laver 1975, 1981). In any particular case, such a greeting can be read as a base form of openings, or the first 'move' of interaction. However, a greeting alone is not sufficient for displaying status and relationship, which is the function of the opening discourse. Therefore, the greeting is likely to generate the second act (and the third one) to accomplish this function. For example, the greeting is either followed by a phatic inquiry, or sometimes by an expressive act such as 'expressing missing' or 'expressing apologies.' It is through such a stretch of speech acts that the writers attempt to create an effect capable of achieving the phatic functions of discourse.

On the other hand, this study demonstrates that politeness may be conveyed through a series of moves (speech acts), rather than through a single expression. Looking at the evidence from this study, it appears that the writer's level of politeness is strategically realized through a range of speech acts. As examples, one can acknowledge the receipt of the previous letters initially, and then express some feelings or attitudes with respect to the action or result of receiving, such as expressing happiness or thanks. The first act implies that the writer is still sharing membership with the recipient, that is, positive politeness. The second one, expressing happiness, may reveal the writer's further confirmation of the in-group membership, which indicates that the writer is sharing more 'common ground' with the recipient, thus emphasizing positive politeness. Here the second act is not only a relevant move, but also an addition of politeness to the initial act. Thus, in this case we can claim that politeness plays a role in the process of

264 sequence in which all acts are necessary conditions for successful accomplishment of politeness.

Moreover, this study also illuminates the theoretical issue of the relationship between language use and social variables. Through an in-depth analysis of the way in which writers manipulate their language in the addressing, opening and closing behaviours (choosing a kind of type of strategy, sequencing speech acts, in particular situations), the study demonstrates what social factors contribute to the specific linguistic behaviour of a given setting, and uncovers the interrelationship between social constraints encoding socio-cultural values and normative language usage. Based on the results and discussion of this study, several culturally-specific proposals are made below:

1. Different hierarchical relationships are recognized in different relations,

In performing the linguistic behaviour which can best reflect the language of social relationship (i.e. addressing), different hierarchical relationships resulting from the role and age are acknowledged within both asymmetrical- and symmetrical-role contexts in both family and non-family settings.

2. Different hierarchical relationships are recognized in the asymmetrical role contexts, but not in the symmetrical role-relationships in both family and non-family settings.

The choice of the type of strategy and language use for every single speech act (e.g. single strategy vs. combined strategy; honorific vs. familiarity) distinguishes different status level. Role and age interact to produce hierarchy within the asymmetrical relationships, while the equal role overrides age differences to produce reciprocity in the symmetrical relationships in both family and non-family settings.

265 3. The degree of authority and intimacy or distance is the keys for sequencing certain speech events from writers.

In sequencing the acts of certain speech behaviours, a distinction between direct-blood relations and indirect-blood relations and non-family relations is made within the asymmetrical relationships based on the personal connections: Different hierarchical relationships are acknowledged in family unequals, but not in relatives or non-family unequals. Within the symmetrical relationships, the equal role overrides age differences to produce a symmetrical use of sequences.

4. Gender differences are recognized in many aspects oflanguage in relation to age and the degree ofintimacy or distance.

Gender differences are recognized lexically, in choosing a speech act, in choosing types of strategy and in patterning some ritual sequences in relation to age (mostly from the older/elder to the younger) and the degree of intimacy or distance (mostly between intimates).

5. Social meaning-weighed speech event is the key for the gender-based similarities.

When a speech behaviour (e.g. closings), particularly in sequences, has more social meanings than gender differences, there are usually small differences in the patterning of the sequences between women and men, although linguistic realizations of language patterns vary across sexes.

7.2.2 Implications for learning and teaching language and cross-cultural communication The empirical results of this study have implications for learning Chinese as a native language. As we have mentioned before, not all native speakers are simply competent in composing the addressing, opening and closings parts in personal letters, and in using the linguistic forms and structural patterns appropriately in a given setting. While taking into account native speakers' real language ability, Hymes (1972) argues that children can produce and understand all the grammatical sentences of a language, but

266 they could not have acquired the knowledge of how and when to use these sentences and to whom. This phenomenon claims that social verbal routines, such as addressing practices in written communication, and the socio-cultural norms embodied in these routines need to be acquired along with the formal properties of language.

The information obtained from this study can reveal the conventions, knowledge and cultural norms embedded in the addressing, opening and closing behaviours in personal letters. Being aware of the use of forms and patterns of these behaviours for various social functions, Chinese people can thoroughly understand their linguistic features and pragmatic meanings in this kind of written interaction. Thus, they can use these polite, phatic and formulaic expressions appropriately according to a particular social or personal circumstance while writing to the other. Hence, their competence in this authentic written language activity would be developed.

This study is also significant for learning/teaching Chinese as a second or foreign language. As has been mentioned previously, the nature of the written address, openings and closings and the rules governing the appropriate use of language in Chinese personal letters have not yet been fully clarified. Consequently, there are few Chinese textbooks used for foreign language teaching that specifically explains how Chinese perform these three speech behaviours in personal letters, that explicate the relationship between language and social factors in the larger socio-cultural context, and that provide explanations for possible causes of misunderstanding and miscommunication. Most materials studied in classroom settings rely on invented sentences used in isolation and they are often not substantiated by empirical results. As a result, foreign language learners often lack opportunities to acquire the linguistic and pragmatic rules of the addressing, opening and closing behaviours in Chinese personal letters. Understandably, this leaves them in a weak position for real interaction with native speakers of Chinese. In order to avoid miscommunication, an effective and appropriate use of language from pragmatic knowledge should be a component of second or foreign language learning.

Supported by data from authentic discourse, this study describes the genuine uses of written language in the three behaviours at the lexical, speech-act and discourse levels

267 by Chinese in different social contexts. In using the data from this study as authentic materials, teachers can present to beginner students the most frequently occurring lexical and syntactic structures and their pragmatic functions and social meanings which entail cultural beliefs and values. As students of Chinese become familiar with the general characteristics of the various forms and patterns and understand the underlying social meanings, teachers can introduce advanced students to the use of language at the discourse-level as well as variations across different contexts. This would further help learners to be aware of expression management and face work development such as knowing how to sequence speech acts in openings and closings. This can help learners solve difficulties in controlling discourse, which is very important for communicative competence.

Through knowing forms, functions and contexts, learners of Chinese would not only be linguistically competent but also communicatively competent, as they would have "the knowledge of linguistic and related communicative conventions that speakers must have to create and sustain conversational cooperation" (Gumperz 1982: 209). More importantly, this would aid in not only understanding pragmatic knowledge, but also in understanding the differences and similarities between certain speech events of peoples from different cultures. It would no doubt reduce misconceptions and misunderstandings, which is crucial for successful cross-cultural communication.

7.3 Limitations of this study

As in any research, there are several limitations to this study. First, the findings reported from this study are based on empirical data gathered from Xinjiang Autonomous Region in the far west of the country, the largest of China's provinces and autonomous regions. The ethnic and historical background and language use, and features of writers from here may be representative of the Han people of China in general. However, the geographical, economic and cultural features of Xinjiang, such as the remoteness in location and slow development in education and macro-economy, may result in linguistic behaviour of people in Xinjiang being different from other parts of China. Therefore, the generalization of the findings to all Han people in China must be

268 undertaken with caution. The analytical framework developed in this study needs to be further tested in other research settings.

The second limitation is the size of the sample population. Although the study uses both a quantitative and a qualitative analysis to compensate for the small number of subjects, some categories of subjects (teachers, relatives, children) are still too limited to test the statistical significance of the differences in language use between different categories of subjects (i.e. writers). It would be hard for these categories to be statistically replicated by other researchers. Therefore, due to the limited sampling of writers and their performances, the results should be understood as suggestive and not · definitive. Moreover, this study explores the addressing, opening and closings behaviours within specific contexts. Some relationships that need to be investigated are excluded (e.g. colleagues, acquaintances). As a result, the applicability of the analytical framework, developed from performances within specific contexts, may be of a restricted scope.

The third limitation of this study is that it is based on one type of data, that is, personal letters. There was a lack of data to supplement the authentic discourse, such as interviews and questionnaires. Therefore, the study does not show how recipients perceive the linguistic forms and functions embedded in the three events, nor reveal how they look at the relationship of language to social variables and to underlying cultural beliefs and values. The explanations and interpretations of the data are likely to contain the implicit assumptions and biases of the researcher. Thus, the subjectivity of the researcher possibly affected the explanations of the data. Considering this point, the subjects' attitudes towards the three behaviours in written communication should be studied in order to support the validity of the explanations given here.

7.4 Directions for future research

In light of the findings obtained from this investigation, there are a number of issues that could be explored in future studies. First, it is important to confirm and validate the linguistic realizations of the addressing, opening and closing events in wider research

269 settings. As this study is built on the data limited to certain social contexts, the conclusions should be thought of as tentative and suggestive. The analytical framework and naturally occurring forms and patterns need to be reexamined and refined with a larger sample and with regard to various social variables. Such an investigation will provide more valid information for examining and assessing the language and pragmatic features of Chinese personal communication.

Second, further research should study how these three behaviours are performed systematically in Chinese business letters. In general, business writing is another type of writing which has a special linguistic format and formula. It functions as an important communication tool, and it is closely linked to the professional image of . writers and business. Appropriate use of formulaic expressions in routine behaviours (e.g. openings and closings) is significant for "professional flavour and the personal touch" (Wei 1998: 1). Therefore, an in-depth comparison would uncover the properties of these speech behaviours in business correspondence and explicate differences and similarities in language uses between business and personal interaction. Thus, it would familiarize speakers and learners of Chinese with linguistic knowledge and business and personal etiquette for both writing settings.

The third issue that merits further investigation is the examination of socio-historical change in the use of addressing, opening and closing patterns, including the differences in perception of politeness in personal written interaction between Old China and New China, the periods demarcated by the liberation of 1949. Such an investigation would highlight the changing values of address forms and the sequencing of openings and closings in the context of political changes and social progress. It would also explore factors, such as social factors and politeness perceptions, which operate to influence these changing forms and patterns. Comparisons between the past and present would reflect the "correlation of shifting semantics and changing ideology" (Lee-Wong 1994: 300) in the contexts of feudalism and modernity. This would enable readers to acquire insight into the issues of social changes and changing linguistic behaviours in China.

A final direction for future research would be to conduct a contrastive study of these three behaviours between Chinese and English. This suggestion in particular has

270 implication for teaching English as a second or foreign language to Chinese speakers. English is regarded as a universal language in the world because of its wide use by a large proportion of the world's population. In China, learning English has become very popular as China increases its connection with English-speaking countries. Therefore, the findings to be obtained from such studies would allow language teachers and course designers to incorporate the pragmatic knowledge of written addressing, openings and closings in the English of daily life into texts and teaching. This is important for language planners and teachers towards a more effective approach in English teaching and learning.

Despite the limitations and suggestions for future research, it is hoped that this present work would have shed some light on the use of the three routine behaviours in Chinese personal letters. It is also hoped that this present work would have motivated further interest and exploration of these aspects by other researchers.

271 REFERENCES

Aijmer, K. (1996). Conversational routines in English: Convention and creativity. London: Longman. Akinasso, F. (1991). Literacy and individual consciousness. In E. Jennings & A. Purve (Eds.), Literacy systems and individual lives (pp.73-94). New York: State University of New York Press. Albert, S., & Kessler, S. (1976). Processes for ending social encounters: The conceptual archeology of a temporal place. Journal for the Theory of Social Behaviour, 6, 147-170. Albert, S., & Kessler, S. (1978). Ending social encounters. Journal of Experimental Social Psychology, 14, 541-553. Andersen, J. T. (1988). Relevance: Communication and cognition. Language in Society, 17, 604-608. Aries, E. (1996). Men and women in interaction: Reconsidering the differences. Oxford: Oxford University Press. Asianinfor. (2000). Education and literacy in China. Retrieved 17, June, 2003, from http://www.asianinfo.org/asianinfo.china/pro-education.htm. Austin, J. L. (1962). How to do things with words. Oxford: Clarendon Press. Bach, K. (1998). Speech acts. In E. Craig (Ed.), Routledge encyclopedia ofphilosophy (Vol. 9, pp. 81-87). London: Routledge. Bargiela-Chiappini, F., & Harris, S. J. (1996). Requests and status in business correspondence. Journal ofPragmatics, 26(5), 635-662. Barton, D., & Hall, N. (1999). Introduction. In D. Barton & N. Nigel (Eds.), Letter writing as a social practice. Amsterdam: John Benjamins. Bates, E., & Benigni, L. (1975). Rules of address in Italy: A sociological survey. Language in Society, 4, 271-288. Bauman, R., & Sherzer, J. (1974). Introduction. In R. Bauman & J. Sherzer (Eds.), Explorations in the ethnography of speaking (pp. 6-12). Cambridge: Cambridge University Press. Bayraktaroglu, A. (2001). Advice-giving in Turkish. In A. Bayraktaroglu & M. Sifianou (Eds.), Linguistic politeness across boundaries (pp. 177-208). Amsterdam: John Benjamins. Berger, C. R., & Bradac, J. S. (1982). Language and social knowledge: Uncertainty in

272 interpersonal relationships. London: Edward Arnold. Bernsen, N. 0., Dybkjer, H., & Dybkjier, L. (1996). Cooperative in human-machine and human-human spoken dialogue. Discourse Processes, 21,213-236. Besnier, N. (1989). Literacy and feelings: The encoding of affect in Nukulaelae letters. Text, 9(1), 69-91. Bhatia, V. K. (1994). Analyzing genre: Language use in professional settings. London: Longman. Bickmore, T. (1999). A computational model ofsmall talk. Retrieved May 16,2003, from

http://web.media.mit. edu/~bickmore/Mas962b/ Bitchener, J. W. (1999). The negotiation of meaning by advanced ESL learners: The effects of individual learner factors and task type. Unpublished Doctorial dissertation, Auckland University. Blum, S. ( 1997). Naming practices and the power of words in China. Language in Society, 26(3), 357-379. Blum-Kulka, S. (1990). You don't touch lettuce with your fingers: Parental politeness in family discourse. Journal ofPragmatics, 14(2), 259-288. Blum-Kulka, S., Danet, B., & Gerson, R. (1985). The language ofrequesting in Israeli society. In J. Forgas (Ed.), Language in social situations (pp. 113-139). New York: Springer-Verlag. Blum-Kulka, S., House, J., & Kasper, G. (1989). Cross-cultural pragmatics: Requests and apologies. Norwood, N.J: Ablex Pub. Boe, S. K. (1987). Language as an expression of caring in women. Anthropological Linguistics, 29(3), 271-285. Bond, M. H., & Hwang, K. K. (1989). The social psychology of Chinese people. In M. H. Bond (Ed.), The Psychology ofthe Chinese People (pp. 213-266). Oxford: Oxford University Press. Bouma, G. D. (2001). Research process. Oxford: Oxford University Press. Braun, F. (1988). Terms ofaddress: Problems ofpatterns and usage in various languages and cultures. Berlin: Mouton. Brown, P. (1998). How and why are women more polite: Some evidence from a Mayan community. In J. Coates (Ed.), Language and gender: A reader (pp. 81-99). Oxford: Blackwell. Brown, P., & Levinson, S. C. (1978). Universals m language usage: Politeness

273 phenomena. In E. N. Goody (Ed.), Questions and politeness: Strategies in social interaction (pp. 56-31 0). Cambridge: Cambridge University press. Brown, P., & Levinson, S.C. (1987). Politeness: Some universals oflanguage usage. Cambridge: Cambridge University Press. Brown, R. (1965). Social psychology. New York: Free Press. Brown, R. (1990). Politeness theory: Exemplar and exemplary. In I. Rock (Ed.), The legacy of Solomon Asch: Essays in cognition and social psychology (pp. 23-38). Hillsdale: Lawrence Erlbaum. Brown, R., & Ford, M. (1961). Address in American English. Journal ofAbnormal and Social Psychology, 62, 375-385. Brown, R., & Gilman, A. (1960). The pronouns of power and solidarity. InT. A. Sebeok (Ed.), Style in language (pp. 253-276). Cambridge, MA: MIT Press. Brown, R., & Gilman, A. (1989). Politeness theory and Shakespeare's four major tragedies. Language in Society, 18, 159-212. Bryman, A. (1988). Quantity and quality in social research. London: Unwin Hyman. Callan, H. (1970). Ethology and society. Oxford: Clarendon Press. Cameron, D. (1990). Feminism and linguistic theory. London: Macmillan. Canale, M., & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics, I, 1-47. Chambers, J. C. (1992). Linguistic correlates of gender and sex. English World-Wide, 13(2),173-218. Chao Yuenren. (1956). Chinese terms of address. Language, 32,217-241. Chao Yuenren. (1976). Aspects of Chinese sociolinguistics. CA: Stanford University Press. Chen Rong. (1993). Responding to compliments: A contrastive study of politeness strategies between American English and Chinese speakers. Journal of Pragmatics, 20, 49-75. Chen Song-cen. ( 1991 ). Social distribution and development of greeting expressions in China. International Journal ofthe Sociology ofLanguage, 92, 55-60. Chen Wuyun. (1999). Shuxin daquan [A selection ofletters]. Shanghai: Shanghai Education Press. Cherry, R. D. (1988). Politeness in written persuasion. Journal ofPragmatics, 12, 63-81. Chodorow, N. (1974). Family structure and feminine personality. In M. Z. Rosaldo & L.

274 Lamphere (Eds.), Women, culture and society (pp. 39-59). Stanford: Stanford University Press. Chu Chauncey. (1998). Discourse grammar of Mandarin Chinese. New York: Peter Lang. Chu, L. (1979). The sensitivity of Chinese and American children to social influence. Journal ofSocial Psychology, 109, 175-186. Cicognani, A., & Maher, M. L. (1997). Design speech acts: "How to do things with words" in virtual communities. Retrieved 22 June, 2003, from http://www.arch.usyd.edu.au/anna.papers/caadf97.html Clark, H. H., & French, J. W. (1981). Telephone Goodbyes. Language in Society, 10, 1-19. Coates, J. (1988). Gossip revisited. In J. Coates & D. Cameron (Eds.), Women in their speech communities: New perspectives on language and sex (pp. 94-122). London: Longman. Cohen, A., & Olshtain, E. (1981). Developing a measure of sociocultural competence: The case of apology. Language Learning, 31 ( 1), 113-134. Cook-Gumperz, J. (1993). Dilemmas of identity: Oral and written literacies in the making of a basic writing student. Anthropology and Education Quarterly, 24, 336-356. Coulmas, F. (1981). Introduction. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 1-17). The Hague: Mouton. Coulmas, F. (1981). Poison to your soul. Thanks and apology contrastively viewed. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 69-91). The Hague: Mouton. Coupland, J. (2000). Introduction: Sociolinguistic perspectives on small talk. In J. Coupland (Ed.), Small talk (pp. 1-25). London: Longman. Coupland, J., Coupland, N., & Robinson, J. D. (1992). 'How are you?': Negotiating phatic communion. Language in Society, 21,207-230. Dale, I. R. H. (1978). Beyond intuition: The use of questionnaires m linguistic investigation. Anthropological Linguistics, 20, 158-166. De Klerk, V., & Bosch, B. (1999). Nicknames as evidence of verbal playfulness. Multilingua, 18(1), 1-16.

275 DeCapua, A., & Dunham, J. F. (1993). Strategies in the discourse of advice. Journal of Pragmatics, 20, 519-531. Dierks, K. (1999). The familiar letter and social refinement in America, 1750-1800. In D. Barton & N. Hall (Eds. ), Letter writing as a social practice (pp. 31-41 ). Amsterdam: John Benjamins. Dittmer, A. (1991). Letters: The personal touch in writing. English Journal, 80, 17-32. Duranti, A. (1985). Sociocultural dimensions of discourse. In T. A. V. Dijk (Ed.), Handbook of discourse analysis. (Vol. 1, pp. 193-230). London: Academic. Duranti, A. (1992). Language in context and language as context: The Samoan respect vocabulary. In A. Duranti & C. Goodwin (Eds.), Rethinking context: Language as an interactive phenomenon (pp. 77-100). Cambridge: Cambridge University Press. Ebrey, P. (1991). The Chinese family and the spread of Confucian values. In G. Rozman (Ed.), The East Asian region: Confucian heritage and its modern adaptation (pp. 45-83). Princeton, NJ: Princeton University Press. Edmondson, W. J. (1981). On saying you're sorry. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and

prepatterned speech (pp. 273-2~8). The Hague: Mouton. Belen, G. (2001). A critique ofpoliteness theories. Manchester, UK: St. Jerome. Ervin-Tripp, S. (1972). Sociolinguistic rules of address. In J. B. Pride & J. Holmes (Eds.), Sociolinguistics: Selected readings (pp. 225-240). Harmondsworth: Penguin. Ervin-Tripp, S. (1976). Is Sybil there?: The structure of American English directives. Language in Society, 2, 25-66. Faerch, C., & Kasper, G. (1984). Pragmatic knowledge: Rules and procedures. Applied Linguistics, 5(3), 214-225. Farr, M. (1993). Essayist literacy and other verbal performances. Written Communication, 10, 4-38. Ferguson, C. A. (1981). The structure and use of politeness formulas. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech. The Hague: Mouton. Firth, J. R. (1972). Verbal and bodily rituals of greeting and parting. In J. S. L. Fontaine (Ed.), The interpretation of ritual (pp. 1-38). London: Tavistock. Fitch, K. L., & Sanders, R. E. (1994). Culture, communication, and preferences for

276 directness in expression of directives. Communication Theory, 4, 219-245. Fong, M. (2000). 'Luck talk' in celebrating the Chinese New Year. Journal ofPragmatics, 32, 219-237. Forgas, J.P. (1999). On feeling good and being rude: Affective influences on language use and request formulations. Journal ofPersonality and Social Psychology, 76, 928-939. Fraser, B. (1981). On apology. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 259-271). The Hague: Mouton. Fraser, B. (1990). Perspectives on politeness. Journal ofPragmatics, 14, 219-236. Friedrich, P. (1966). Structural implications of Russian pronominal usage. In W. Bright (Ed.), Sociolinguistics (pp. 214-259). The Hague: Mouton. Gal, S. (1995). Language, gender, and power: An anthropological review. InK. Hall & M. Bucholtz (Eds.), Gender articulated: Language and the socially constructed self (pp. 169-182). New York: Routledge. Geis, M. L. (1998). Speech acts and conversational interaction. Cambridge: Cambridge University Press. Gilbert, M. A. (1999). Language, words and expressive speech acts. Paper presented at the Proceedings of the Fourth International Conference of the International Society for the Study of Argumentation. Glesne, C., & Peshkin, A. (1992). Becoming qualitative researchers: An introduction. London: Longman. Goffman, E. (1963). Behaviour in public places. New York: The Free Press. Goffman, E. (1967). Interaction ritual: Essays on face to face behaviour. Garden City, NY: Anchor Books. Goffman, E. (1971). Relations in public: Macrostudies of the public order. Harmondsworth: Penguin. Goffman, E. (1972). On face-work: An analysis of ritual elements in social interaction. In J. Laver & S. Hutcheson (Eds.), Communication inface-to-face interaction (pp. 319-346). Harmondsworth: Penguin. Goffman, E. (1976). Replies and responses. Language in Society, 5, 257-313. Goffman, E. (1981). Forms oftalk. Oxford: Blackwell. Goodwin, M. H. (1998). Cooperation and competition across girls' play activities. In J.

277 Coates (Ed.), Language and gender: A reader (pp. 121-146). Oxford: Blackwell Publishers. Goody, E. (1972). Greetings, beggings and the presentation of respect. In J. S. L. Fontaine (Ed.), The Interpretation ofritual (pp. 39-72). London: Tavistock. Gordon, E. (1997). Sex, speech and stereotypes: Why women use prestige speech forms more than men. Language in Society, 26, 47-63. Grice, H. P. (1957). Meaning. Philosophical Review, 66(3), 377-388. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semantics 3: Speech acts (pp. 41-58). New York: Academic Press. Grice, H. P. (1989). Studies in the way ofwords. Cambridge, MA: Harvard University Press. Gu Yueguo. (1990). Politeness phenomena in modem Chinese. Journal of Pragmatics, 14(2), 237-257. Gumperz, J. (1972). Introduction. In J. Gumperz & D. H. Hymes (Eds.), Discourse strategies (pp. 1-25). Cambridge: Cambridge University Press. Gumperz, J. (1982). Language and social identity. Cambridge: Cambridge University Press. Haberland, H., & Mey, G. L. (1977). Editorial: Linguistics and pragmatics. Journal of Pragmatics, 1(1), 1-12. Hammersley, M. (1998). Reading ethnographic research: A critical guide. London: Longman. Hank, W. F. (1996). Language form and communicative practices. In J. Gumperz. & S.C. Levinson (Eds.), Rethinking linguistic relativity (pp. 232-270). Cambridge: Cambridge University Press. Harberland, H. (1996). Communion or communication?: A historical note on one of the 'founding father's of pragmatics. In R. Sackman (Ed.), Theoretical linguistics and grammatical description: papers in honour ofHans-Heinrich Lieb (pp. 163-166). Amsterdam: John Benjamins. Hatch, E. M. (1992). Discourse and language education. Cambridge: Cambridge University Press. Held, G. (1989). On the role of maximization in verbal politeness. Multilingual, 8, 167-206. Herbert, R. K. (1990). Sex-based differences in compliment behaviours. Language in

278 Society, 19, 201-224. Hijirida, K., & Sohn, H. M. (1983). Commonality and relativity in address-reference term usage. Language Research, 19(2), 139-168. Holmes, J. (1984). Women's language: A functional approach. General Linguistics, 24(3), 149-178. Holmes, J. (1990). Apologies in New Zealand English. Language in Society, 19, 155-199. Holmes, J. (1992). An introduction to sociolinguistics. London: Longman. Holmes, J. (1995). Women, men, and politeness. London: Longman. Holmes, J. (2000). Doing collegiality and keeping control at work: Small talk in government departments. In J. Coupland (Ed.), Small talk (pp. 32-61). London: Longman. Holtgraves, T. M. (2002). Language as social action. London: Lawrence Erlbaum. Hong, B. (1985). Politeness in Chinese: Impersonal pronouns and personal greetings. Anthropological Linguistics, 27(2), 204-213. Hopper, R. (1992). Telephone conversation. Bloomington: Indiana University Press. Hopper, R., & Chen, C. H. (1996). Languages, cultures, relationships: Telephone openmgs in Taiwan. Research on Language and Social Interaction, 29(4), 291-313. Hopper, R., Doany, N., Johnson, M., & Drummond, K. (1990/1991). Universals and particulars in telephone openings. Research on Language and Social Interaction, 25, 367-385. Hopper, R., & Doany, N. K. (1989). Telephone openings and conversational universals: A study in three languages. In S. T. Toomey & F. Korzenny (Eds.), Language, communication and culture (pp. 157-179). Newbury Park: Sage. Hom, L. R. (1984). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In D. Schiffrin (Ed.), Meaning, form, and use in context: Linguistic applications (pp. 11-42). Washington, D.C.: Georgetown University Press. Houtkoop-Steenstra, H. (1991). Opening sequences in Dutch telephone conversation. In D. Boden & D. H. Zimmerman (Eds.), Talk and social structure (pp. 232-250). CA: University of California Press. Hudson, R. A. (1996). Sociolinguistics. Cambridge: Cambridge University Press. Hutchby, I. (1999). Frame attunement and footing in the organization of talk radio

279 openings. Journal of Sociolinguistics, 3(1 ), 41-63. Hwang Kwang-kuo. (1987). Face and favor: The Chinese power game. American Journal of Sociology, 92(4), 944-974. Hymes, D. (1972). On communicative competence. In J. B. Pride & J. Holmes (Eds.), Sociolinguistics (pp. 269-293). New York: Penguin. Hymes, D. H. (1962). The ethnography of speaking. InT. Gladwin & W. Sturtevant (Eds.), Anthropology and Human Behaviour (pp. 13-53). Washington, DC: Anthropological Society of Washington. Ide, S. (1989). Formal forms and discernment: Two neglected aspects of universals of linguistic politeness. Multilingual, 8, 223-248. Inchaurralde, C. (1997). What is behind a word: Cultural scripts. In M. Putz (Ed.), The cultural context in foreign language teaching (pp. 55-66). Berlin: Peter Lang. Irvine, J. T. (1974). Strategies of status manipulation in Wolofgreetings. In R. Bauman & J. Sherzer (Eds.), Explorations in the ethnography of speaking (pp. 167-191). Cambridge: Cambridge University Press. Irvine, J. T. (1979). Formality and informality in communicative events. American Anthropologist, 81, 773-790. Jakobson, R. (1960). Concluding statement: Linguistics and poetics. In T. A. Sebeok (Ed.), Style and language (pp. 350-377). Cambridge, MA: MIT Press. Jaworski, A. (2000). Silence and small talk. In J. Coupland (Ed.), Small talk (pp. 11 0-132). London: Longman. Jenkins, S., & Hinds, J. (1987). Business letter writing: English, French and Japanese. TESOL Quarterly, 21,327-345. Johnstone, B. (2000). Qualitative methods in sociolinguistics. Oxford: Oxford University Press. Kalman, J. (1996). Joint composition: The collaborative letter writing of a scribe and his client in Mexico. Written Communication, 13(2), 190-220. Kasper, G. (1989). Interactive procedures in interlanguage discourse. In W. Olesky (Ed.), Contrastive pragmatics (pp. 189-229). Amsterdam: John Benjamins. Kasper, G. (1990). Linguistic politeness: Current research issues. Journal ofPragmatics, 14(2), 193-218. Kataoka, K. (1997). Affect and letter-writing: Unconventional conventions in casual writing by young Japanese women. Language in Society, 26, 103-136.

280 Kendon, A., & Ferber, A. (1973). A description of some human greetings. In R. P. Michael & J. H. Cook (Eds.), Comparative ethology and behaviour ofprimates (pp. 591-668). New York: Academic Press. Kess, J., & Juricic, Z. (1978). Slovene pronominal address forms: Rural vs. urban sociolinguistic strategies. Anthropological Linguistics, 20(7), 297-311. Kirkpatrick, A. (1991 ). Information sequencing in Mandarin letters of request. Anthropological Linguistics, 2, 183-203. Knapp, M. (1978). Social intercourse: From greeting to goodbye. Boston: Allyn and Bacon. Knapp, M., Hart, R. P., Friedrich, G. W., & Schulman, G. M. (1973). The rhetoric of goodbye: Verbal and non-verbal correlates of human leaving taking. Speech Monographs, 40, 183-198. Koo, J. (1995). Politeness theory: Universality and specificity. Unpublished doctorial dissertation, Harvard University. Krivinos, P., & Knapp, M. (1975). What do you say when you say hello. Central States Speech Journal, XXVI(2), 115-125. Kroger, R. 0., & Wood, L.A. (1992). Are the rules of address universal?: Comparison of Chinese, Korean, Greek, and German usage. Journal of Cross-Cultural Psychology, 23(2), 148-162. Labov, W. (1972a). Sociolinguistic patterns. Philadelphia: University of Pennsylvania. Labov, W. (1972b). Language in the inner city: Studies in the black English vernacular. Philadelphia: University of Pennsylvania Press. Lakoff, R. (1973). The logic ofpoliteness: Or, minding your p 'sand q 's. Paper presented at the The Ninth Regional Meeting of the Chicago Linguistic Society, Chicago. Lakoff, R. (1975). Language and woman's place. New York: Harper & Row. Lakoff, R. (1982). Some of my favorite writers are literate: The meaning of oral and literate strategies in written communication. In D. Tannen (Ed.), Spoken and written language (pp. 239-260). Norwood, NJ: Ablex. Lamber, W., & Tucker, G. R. (1976). Tit, vous, usted: A social psychological study of address patterns. Rowley, MA: Newbury House. Lang, L. B. (1994). Letter writing. Glasgow: Harper Collins. Laver, J. (1975). Communicative functions ofphatic communion. In A. Kendon, R. M. Harris & M. R. Key (Eds.), The organization of behaviour in face-to-face

281 interaction (pp. 215-238). The Hague: Mouton. Laver, J. (1981 ). Linguistic routines and politeness in greeting and parting. In F. Coulmas. (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 289-304). The Hague: Mouton. Leech, G. N. (1983). Principles ofpragmatics. London: Longman. Lee-Wong, S.M. (2000). Politeness and face in Chinese culture. New York: Peter Lang. Lee-Wong Song-mei. (1994). Address forms in modem China: Changing ideologies and shifting semantics. Linguistics, 32, 299-324. Li, C. N., & Thompson, S. A. (1981). Mandarin Chinese: A functional reference grammar. CA: University of California Press. Li, D. C. S. (1997). Borrowed identity: Signaling involvement with a Western name. Journal ofPragmatics, 28(4), 489-513. Lindstrom, A. (1994). Identification and recognition in Swedish telephone conversation opening. Language in Society, 23,231-252. Luger, H. H. (1983). Some aspects of ritual communication. Journal of Pragmatics, 7, 695-711. Lyons, J. (1977). Semantics (Vol. 1). Cambridge: Cambridge University Press. Ma Hongji, & Chang Qingfeng. (1998). Chen wei yu [Forms of address]. Beijing: XinHua Press. Maier, P. (1992). Politeness strategies in business letters by native and non-native English speakers. English for Specific Purposes, 11, 189-205. Malinowski, B. (1922). Argonauts ofthe Western Pacific. London: Routledge. Malinowski, B. (1923). The problem of meaning in primitive languages. In C. K. Ogden & I. A. Richards (Eds.), The meaning of meaning (pp. 451-510). London: Routledge and Kegan Paul. Malinowski, B. (1972). Phatic communion. In J. Laver & S. Hutcheson (Eds.), Communication in face-to-face interaction (pp. 146-152). Harmondsworth: Penguin. Maltz, D., & Borker, R. (1982). A cultural approach to male-female miscommunication. In J. Gumperz (Ed.), Language and social identity (pp. 196-216). Cambridge: Cambridge University Press. Mao, L. M. R. (1994). Beyond politeness theory: 'Face' revisited and renewed. Journal of Pragmatics, 21(5), 451-486.

282 Martiny, T. (1996). Forms of address in French and Dutch: A sociopragmatic approach. Language Sciences, 18(3-4), 765-775. Matsumoto, Y. (1988). Re-examination of the universality of face: Politeness phenomena in Japanese. Journal ofPragmatics, 12, 403-426. Matsumoto, Y. (1989). Politeness and conversational universals: Observations from Japanese. Multilingual, 8, 207-221. Mayumi, U. (2002). Discourse politeness in Japanese conversation. Tokyo: Hituzi Syobo. McElhinny, B. (1995). Challenging hegemonic masculinities: Females and males police officers handing domestic violence. InK. Hall & M. Bucholtz (Eds.), Gender articulated: Language and the socially constructed self(pp. 215-243). New York: Routledge. McNaughton, W. (1974). The confucian vision. Ann Arbor: University of Michigan Press. Mey, J. L. (2001). Pragmatics: An introduction. Oxford: Blackwell Publishers. Miller, G. (1998). Building bridges: The possibility of analytic dialogue between ethnography, conversation analysis and Foucault. In D. Silverman (ed.), Qualitative research : Theory, method and practice (pp.24-44). London: SAGE Publications. Moerman, M. (1988). Talking culture: Ethnography and conversation analysis. Philadelphia: University of Pennsylvania Press. Moerman, M. (1991). Exploring talk and interaction. Research on Language and Social Interaction, 24, 173-187. Moeschler, J. (2002). Speech act theory and the analysis of conversations. In D. Vanderveken & S. Kubo (Eds.), Essays in speech act theory (pp. 239-261). Amsterdam: John Benjamins. Moles, J. (1974). Decisions and variability: The usage of address forms, pronouns and languages by Quechua-Spanish bilinguals in Peru. Anthropological Linguistics, 16(91), 442-463. Moulin, B., & Rousseau, D. (2002). An approach for modeling and simulating conversations. In D. Vanderveken & S. Kubo (Eds.), Essays in speech act theory (pp. 175-205). Amsterdam: John Benjamins. Nevalainen, T., & Raumolin-Brunberg, H. (1995). Constraints on politeness: The

283 pragmatics of address formulae in early English correspondence. In H. J. Andreas (Ed.), Historical pragmatics (pp.541-601). Amsterdam: John Benjamins. Nin Yongquan, & Liu Yongjun. (1993). Shuxin yingyu [How to write English letters]. Beijing: Foreign Language Teaching and Research Press. Norrick, N. R. (1978). Expressive illocutionary acts. Journal ofPragmatics, 2, 277-291. O'Driscoll, J. (1996). About face: A deference and elaboration of universal dualism. Journal ofPragmatics, 25, 1-32. Okamoto, N. (1990). Denwa niyorukaiwaskuketsu no kenkyu [A study of closing the conversation on the telephone]. Nihonggo Kyoiku, 72, 145-159. Okolo, B. (1995). Opening-up closings: Negotiating settlements in problematic discourse encounters. Text, 15(2), 191-208. Oliver, R. T. (1971). Communication and culture in ancient India and China. New York: Syracuse University Press. Olshtain, E., & Cohen, A. D. (1983). Apology: A speech act set. InN. Wolfson & E. Judd (Eds.), Sociolinguistics and language acquisition (pp. 18-35). Rowley, MA: Newbury. Omar, A. (1992). Opening and closing conversations in Kiswahili: A study of the performance ofnative speakers and learners. Unpublished Doctorial dissertation, Indiana University. Ostor, A. (1982). Terms of address and Hungarian society. Language Science, 4(1), 55-69. Oyetade, S. 0. (1995). A sociolinguistic analysis of address forms in Yoruba. Language in Society, 24(4), 515-535. Pavlidou, T. (1997). The last five turns: Preliminary remarks on closings in Greek and German telephone calls. International Journal of Society and Language, 126, 145-162. Pavlidou, T. (2000). Telephone conversations in Greek and German: Attending to the relationship aspect of communication. In H. Spencer-Oatey (Ed.), Culturally speaking: Managing rapport through talk across cultures (pp. 121-142). London: Continuum. Pilegaard, M. (1997). Politeness in written business discourse: A textlinguistic perspective on requests. Journal ofPragmatics, 28, 223-244. Placencia, M. E. (1997). Opening up closings - the Ecuadorian way. Text, 17(1 ), 53-81.

284 Preisler, B. (1986). Linguistic sex roles in conversation. Berlin: Mouton de Gruyter. Quina, K., Wingard, J. A., & Bates, H. G. (1987). Language style and gender stereotypes in person perception. Psychology of Women Quarterly, 11(1), 111-122. Rostila, I. (1995). The relationship between social worker and client in closing conversations. Text, 15(1), 69-102. Sacks, H. (1992). Lectures on conversation. Oxford: Blackwell. Sankoff, G. (197 4). A quantitative paradigm for the study of communicative competence. In R. Bauman & J. Sherzer (Eds.), Explorations in the ethnography of speaking (pp. 18-49). Cambridge: Cambridge University Press. Sarantakos, S. (1993). Social research. Melbourne: Macmillan Education. Schaef, A. W. (1985). Women reality: A emerging female system in a white male society. San Francisco: Harper & Row. Schegloff, E. A. (1968). Sequencing m conversational operungs. American Anthropologist, 70, 1075-1095. Schegloff, E. A. (1979). Identification and recognition in telephone conversation openings. In G. Psathas (Ed.), Everyday language: Studies in ethnomethodology (pp. 23-78). New York: Irvington. Schegloff, E. A. (1986). The routine as achievement. Human Studies, 9, 111-152. Schegloff, E. A. (1988). Presequences and indirection: Applying speech act theory to ordinary conversation. Journal ofPragmatics, 12(1 ), 55-62. Schegloff, E. A., & Sacks, H. (1973). Opening up closings. Semiotica, 8, 289-327. Schiffrin, D. (1977). Opening encounters. American Sociological Review, 4215, 679-691. Schmied, J. (1991). English in Africa: An introduction. London: Longman. Schneider, K. (1988). Small Talk: Analyzing phatic discourse. Marburg: Hitzeroth. Schottman, W. (1995). The daily ritual of greeting among the Baatombu of Benin. Anthropological Linguistics, 37(4), 487-523. Scollon, R., & Scollon, S. (1983). Face in interethnic communication. In J. C. Richards & R. W. Schmidt (Eds.), Language and communication (pp. 156-188). London: Longman. Scotton, C. M., & Jin., Z. W. (1983). Tongzhi in China: Language change and its conversational consequences. Language in Society, 12,477-494. Searle, J. R. (1969). Speech acts: An essay in the philosophy of language. Cambridge: Cambridge University Press.

285 Searle, J. R. (1975). Indirect speech acts. In P. Cole & J. Morgan (Eds.), Syntax and Semantics: Vol. 3 Speech acts (pp. 59-82). New York: Academic Press. Searle, J. R. (1976). A classification ofillocutionary acts. Language in Society, 5, 1-23. Searle, J. R. (1979). Expression and meaning: Studies in the theory of speech acts. Cambridge: Cambridge University Press. Searle, J. R., Kiefer, F., & Bierwisch, M. (Eds.). (1980). Speech act theory and pragmatics. Boston: Reidel. Shethar, A. (1993). Literacy and "empowerment"?: A case study ofliteracy behind bars. Anthropology and Education Quarterly, 24, 357-372. Shih, Y. H. (1986). Conversational politeness and foreign language teaching. Taipei: Crane. Sifianou, M. (1989). On the telephone again!: Differences in behaviour: England versus Greece. Language in Society, 18, 527-544. Simkins-Bullock, J. A., & Wildman, B. G. (1991). An investigation into the relationships between gender and language. Sex Roles, 24(3/4), 149-160. Simpson, J. A., & Weiner, E. S.C. (Eds.). (1989). Oxford English dictionary (2nd ed.). Oxford: Clarendon Press. Slobin, D. L., Miller, S. H., & Porter, L. W. (1968). Forms of address and social relations in a business organization. Journal of Personality and Social Psychology, 8, 289-293. So'o, A., & Liddicoat, A. J. (2000). Telephone openings in Samoan. Australian Review of Applied Linguistics, 23(1), 95-107. Spencer-Oatey, H. (1996). Reconsidering power and distance. Journal ofPragmatics, 26, 1-24. Sperber, D., & Wilson, D. (1986). Relevance: Communication and cognition. Oxford: Blackwell. Stekeler-Weithofer, P., & Kannetzky, F. (2000). Speech act and Interpretation. Retrieved

24, June, 2003, from http://www.uni-leipzig.de/~fkv/projekte/p3e.html. Sunland. (n.d.). Xinjiang introduction. Retrieved 19, July, 2003, from www.xinjiang-tourism.com/page10.htm Suzman, S.M. (1994). Names as pointers: Zulu personal naming practices. Language in Society, 23(2), 253-272. Takami, T. (2002). A study on closing sections of Japanese telephone conversations.

286 Working Papers in Educational Linguistics, 18(1), 67-85. Tannen, D. (1984). Conversational style: Analyzing talk among friends. Norwood, NJ: Ablex. Tannen, D. (1986). That's not what I meant!: How conversational style makes or breaks your relations with others. London: Virago Press. Tannen, D. (1989). Talking voices: Repetition, dialogue, and imagery in conversational discourse. Cambridge: Cambridge University Press. Tannen, D. (1990). Youjust don't understand: Women and men in conversation. New York: Ballantine. Tannen, D. (1992). Rethinking power and solidarity in gender and dominance. In C. Kramsch & S. McConnell-Ginet (Eds.), Text and context: Cross-disciplinary perspectives on language study (pp. 135-147). Lexington, MA: Heath. Tannen, D. (1994). Gender and discourse. Oxford: Oxford University Press. Tannen, D., & Oztek, P. C. (1981). Health to our mouths: Formulaic expressions in Turkish and Greek. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 37-54). The Hague: Mouton. Thomas, J. (1995). Meaning in interaction: An introduction to pragmatics. London: Longman. Tian Huigang. (1998). Zhong xi renji chenwei xitong [Addressing systems between the Western and Chinese interaction]. Beijing: Foreign Language Teaching and Research Press. Too, L. (1996). Feng Shui. New York: Barnes and Noble. Tracy, K. (1990). The many faces of facework. In H. Giles & W. P. Robinson (Eds.), Handbook of language and social psychology (pp. 209-226). New York: John Wiley. Tracy, K., & Naughton, J. M. (2000). Institutional identity-work: A better lens. In J. Coupland (Ed.), Small talk (pp. 62-83). London: Longman. Trudgill, P. (1972). Sex, convert prestige and linguistic change in the urban British English ofNorwich. Language in Society, 1, 179-195. U.S. Diplomatic-Representation. (1997). China: Background. Retrieved 17, June, 2003, from http://clinton3 .nara.gov/WH/New/China/china.html Vanderveken, D. (2002). Universal grammar and speech act theory. In D. Vanderveken &

287 S. Kubo (Eds.), Essays in speech act theory (pp. 25-62). Amsterdam: John Benjamins. Vanderveken, D., & Kubo, S. (2002). Introduction. In D. Vanderveken & S. Kubo (Eds.), Essays in speech act theory (pp. 1-21). Amsterdam: John Benjamins. Verschueren, J. (1981). The semantics of forgotten routines. In F. Coulmas (Ed.), Conversational routine: Explorations in standardized communication situations and prepatterned speech (pp. 133-153). The Hague: Monton. Walters, J. (1980). Grammar, meaning, and sociocultural appropriateness in second language acquisition. Canadian Journal ofPsychology, 34(4), 337-345. Watts, R. J., Ide, S., & Ehlich, K. (1992). Introduction. In R. J. Watts, S. Ide & K. Ehlich (Eds.), Politeness in language: Studies in its history, theory and practice (pp. 1-17). Berlin: Mouton. Weatherall, A. (2002). Gender, language and discourse. Berlin: Routledge. Wei Hong. (1996). An empirical study of Chinese request strategies. International Journal ofthe Sociology ofLanguage, 122, 127-138. Wei Hong. (1997). Gender differences in Chinese request patterns. Journal of Chinese Linguistics, 25, 193-210. Wei Hong. (1998). An empirical study of Chinese business writing. Journal of the Chinese Language Teachers Association, 33(3), 1-12. West, C., & Zimmerman, D. (1983). Small insults: A study of interruptions in cross-sex conversations between unacquainted persons. In B. Thome, C. Kramarae & N. Henley (Eds.), Language, gender and society (pp. 102-117). Rowley, MA: Newbury House. Wetzel, P. J. (1993). The language of vertical relationships and linguistic analysis. Multilingua, 12(4), 387-406. Whalen, M. R., & Zimmerman, D. H. (1987). Sequential and institutional contexts in calls for help. Social Psychology Quarterly, 50, 172-185. Wierzbicka, A. (1991). Cross-cultural pragmatics: The semantics of human interaction. Berlin: Mouton de Gruyter. Winograd, T., & Flores, F. (1986). Understanding computer and cognition: A new foundation for design. Norwood: NJ: Ablex. Wolfson, N. (1976). Speech events and natural speech: Some implications for sociolinguistic methodology. Language in Society, 5, 189-209.

288 Wolfson, N. (1983). Rules of speaking. In J. C. Richards & R. W. Schmidt (Eds.), Language and communication (pp. 61-87). London: Longman. Wolfson, N. (1988). The bulge: A theory of speech behaviour and social distance. In J. Fine (Ed.), Second language discourse: A textbook of current research (pp. 21-38). Norwood. NJ: Ablex. Wolfson, N. (1989). Perspectives: Sociolinguistics and TESOL. Cambridge, UK: Newbury House Publishers. Wolfson, N. (1992). Intercultural communication and the analysis of conversation. In H. R (Ed.), language and society in Africa. Johannesburg: University of the Witwatersrand Press. Wolfson, N., & Manes, J. (1980a). The compliment as a social strategy. Paper in . Linguistics: International Journal ofHuman Communication, 13(3), 391-410. Wolfson, N., & Manes, J. (1980b). Don't dear me! InS. McConnell-Ginet, S. Borker & N. Furman (Eds.), Women and language in literature and society (pp. 79-92). New York: Praeger. Wood, L. A, & Kroger, R. 0. (1994). Politeness and forms of address. Language and Social Psychology, 143-168. Wu Yongyi. (1990). The usage of kinship address forms amongst non-kin in Mandarin Chinese. Australian Journal ofLinguistics, 10, 61-88. Xu Jianpin. (2000). A practice course of English-Chinese and Chinese-English translation. Beijing: Qinghua University Press. Xu Yiqing, & Zhang Hexian. (1987). Xingming Qutan [Interesting names and naming]. Shanghai: Shanghai Literary Publisher. Yahya-Othman, S. (1995). Aren't you going to greet me?: Impoliteness in Swahili greetings. Text, 15(2), 209-227. Yang Kuoshu. (1989). Chinese personality and its change. In M. H. Bond (Ed.), The Psychology of the Chinese People (pp. 160-170). Oxford: Oxford University Press. Yeung, L. N. T. (1997). Polite requests in English and Chinese business correspondence in Hong Kong. Journal ofPragmatics, 27, 505-522. Zhan Kaidi. (1992). The strategies of politeness in the Chinese language. CA: The Institute of East Asian Study. Zhao Yingling. (1997). Yingyu chenhu yu de shehui gongneng [The socio-pragmatic

289 functions of English address forms]. Journal of Foreign Language Learning, 1, 14-19. Zhu Weifang. (1997). Shuxin yupian limao xianxiang diaocha [An investigation of politeness phenomena in the discourse of letters]. Journal of Foreign Language Teaching and Research, 1, 15-21. Zhu Yunxia. (1999). Business communication in China. New York: Nova Science Publishers. Zimin, S. (1981). Sex and politeness: Factors in first-and second-language use. International Journal ofSociology ofLanguage, 27, 35-38. Zimmerman, D. H. (1992). The interactional organization of calls for emergency assistance. In P. Drew & J. Heritage (Eds.), Talk at work: Interaction in institutional settings (pp. 418-496). Cambridge: Cambridge University Press.

290 LIST OF TABLES

Table 3.1 Composition of the corpus of letters 71 Table 3.2 Information about social roles of writers 74 Table 3.3 Biographical information about writers 74 Table 3.4 Education background of writers 75 Table 3.5 Descriptive framework for address forms in the addressing section 80 Table 3.6 Descriptive framework for components in the opening and closing section 86 Table 3. 7 Descriptive framework for sequences in the opening and closing section 87 Table 4.1 Categories and distribution of address forms in Chinese personal letters 96 Table 4.2 Frequency and distribution of address forms across social roles 110 Table 4.3 Frequency and distribution of address forms across age/generation 117 Table 4.4 Frequency and distribution of address forms across gender 121 Table 5.1 Categories and distribution of opening strategies in Chinese personalletters 129 Table 5.2 Frequency and distribution of opening types in Chinese personal letters 153 Table 5.3 Frequency and distribution of the strategy of single openings 153 Table 5.4 Frequency and distribution of single openings in initial and reply letters 154 Table 5.5 Combination patterns of openings and their distribution in initial letters 155 Table 5.6 Combination patterns of openings and their distribution in reply letters 162 Table 5.7 Frequency and distribution of opening types across social roles 174 Table 5.8 Frequency and distribution ofthe strategy of single openings across social roles 175 Table 5.9 Frequency and distribution of combination patterns of openings across social roles in initial letters 176 Table 5.10 Frequency and distribution of combination patterns of openings across social roles in reply letters 179 Table 5.11 Frequency and distribution of opening types across age/ generation 183 Table 5.12 Frequency and distribution of combination patterns of openings across age/ generation in initial letters 184 Table 5.13 Frequency and distribution of combination patterns of openings across age/generation in reply letters 185

291 Table 5.14 Frequency and distribution of opening types across gender 188 Table 5.15 Frequency and distribution of the strategy of single openings across gender 188 Table 5.16 Frequency and distribution of combination patterns of openings across gender in initial letters 189 Table 5.17 Frequency and distribution of combination patterns of openings across gender in reply letters 190 Table 6.1 Categories and distribution of closings strategies in Chinese personalletters 196 Table 6.2 Frequency and distribution of closing types in Chinese personalletters 212 Table 6.3 Frequency and distribution ofthe strategy in single closings 212 Table 6.4 Frequency and distribution of single closings in initial and reply letters 213 Table 6.5 Combination patterns of closings and their distribution in Chinese personal letters 215 Table 6.6 Frequency and distribution of combination patterns of closings across letter types 233 Table 6. 7 Frequency and distribution of closing types across social roles 235 Table 6.8 Frequency and distribution of the strategy of single closings across social roles 236 Table 6.9 Frequency and distribution of combination patterns of closings across social roles 237 Table 6.10 Frequency and distribution of closing types across age/generation 242 Table 6.11 Frequency and distribution of combination patterns of closings across age/generation 243 Table 6.12 Frequency and distribution of closing types across gender 247 Table 6.13 Frequency and distribution of the strategy of single closings across gender 24 7 Table 6.14 Frequency and distribution of combination patterns of closings across gender in Chinese personal letters 248

292 LIST OF FIGURES

Figure 3.1 Text-sequential realization of a personal letter 78 Figure 4.1 Summary diagram of social roles that constrain the choice of address forms in family letters 112 Figure 4.2 Summary diagram of social roles that constrain the choice of address forms in non-family letters 113 Figure 5.1 Summary diagram of the sequential combination of strategies of openings in initial letters 171 Figure 5.2 Summary diagram ofthe sequential combination of strategies of openings in reply letters 172 Figure 6.1 Summary diagram of the sequential combination of strategies of closings in Chinese personal letters 231

293