Heritage Language for Chinese Australians: …eprints.qut.edu.au/63295/1/Guanglun_Mu_Thesis.pdfThis...

Heritage Language for Chinese Australians: Negotiating ‘Chineseness’ and

Capitalising on Resources in the Lived World

Dr. Guanglun (Michael) Mu 穆光伦

Bachelor of Science (Beijing Normal University)

Master of Educational Administration (The University of New South Wales)

Principal Supervisor: Associate Professor Dr. Karen Dooley

Associate Supervisor: Dr. Paul Shield

Associate Supervisor: Dr. Catherine Doherty

Faculty of Education

Queensland University of Technology

Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

2013

i

Keywords

Heritage Language, Chinese Heritage Language proficiency, Heritage Language Learner,

Chinese Australians, motivation, investment, capital, ethnic identity, Chineseness, habitus,

field, Bourdieu, mixed methods design, Structural Equation Modelling

ii

Abstract

The ethnic identity and commitment of Heritage Language Learners play salient roles in

Heritage Language learning process. The mutually constitutive effect amongst Heritage

Language Learner’s ethnic identity, commitment, and Heritage Language proficiency has

been well documented in social psychological and poststructuralist literatures. Both social

psychological and poststructural schools offer meaningful insights into particular contexts but

receive critiques from other contexts. In addition, the two schools largely oppose each other.

This study uses Bourdieu’s sociological triad of habitus, capital, and field to reconcile the two

schools through the examination of Chinese Heritage Language Learners in Australia, an

idiosyncratic social, cultural, and historical context for these learners. Specifically, this study

investigates how young Chinese Australian adults (18-35 in age) negotiate their ‘Chineseness’

and capitalise on resources through Chinese Heritage Language learning in the lived world.

The study adopts an explanatory mixed methods design to combine the quantitative

approach with the qualitative approach. The initial quantitative phase addresses the first

research question: Is Chinese Heritage Language proficiency of young Chinese Australian

adults influenced by their investment of capital, the strength of their habitus of ‘Chineseness’,

or both? The subsequent qualitative phase addresses the second research question: How do

young Chinese Australian adults understand their Chinese Heritage Language learning in

relation to (potential) profits produced by this linguistic capital in given fields?

The initial quantitative phase applies Structural Equation Modelling to analyse the data

from an online survey with 230 respondents. Findings indicate the statistically significant

positive contribution made by the habitus of ‘Chineseness’ and by investment of capital to

Chinese Heritage Language proficiency (r = .71 and r = .86 respectively). Subsequent

multiple regression analysis demonstrates that 62% of the variance of Chinese Heritage

iii

Language proficiency can be accounted for by the joint contribution of ‘Chineseness’ and

‘capital’. The qualitative phase of the study uses multiple interviews with five participants. It

reveals that Chinese Heritage Language offers meaningful benefits for participants in the

forms of capital production and habitus capture or recapture. Findings from the two phases

talk to each other in terms of the inherent entanglement amongst habitus of ‘Chineseness’,

investment of capital, and Chinese Heritage Language proficiency.

The study offers important contributions. Theoretically, by virtue of Bourdieu’s

signature concepts of habitus, capital, and field, the study provides answers to questions that

both social psychological and poststructuralist theories have long been struggling to answer.

Methodologically, the position of ‘pluralism’ talks back to Bourdieu’s theory and forwards to

the mixed methods design. Particularly, the study makes a methodological breakthrough: A

set of instruments was developed and validated to quantify Bourdieu’s key concepts of

capital and habitus within certain social fields. Practically, understanding Chinese Australians’

heterogeneity and the potential drivers behind Chinese Heritage Language learning

contributes to the growing interest in Chinese Australians’ contemporary life experiences and

helps to better accommodate linguistically diverse Chinese Heritage Language Learners in

Chinese language courses. In addition, this study is very timely. It resonates with the recently

released Australia in the Asian Century White Paper: Chinese Australians, with sound

knowledge of Chinese culture and language obtained through negotiating their ‘Chineseness’

and capitalising on diverse resources for learning, will help to serve Australia’s economic,

social, and political needs in unique ways.

iv

Table of Contents

Keywords ................................................................................................................................................................i

Abstract ................................................................................................................................................................. ii

Table of Contents .................................................................................................................................................iv

List of Figures ..................................................................................................................................................... vii

List of Tables ...................................................................................................................................................... viii

List of Abbreviations ............................................................................................................................................. x

Statement of Original Authorship .................................................................................................................... xii

Acknowledgement ............................................................................................................................................. xiii

CHAPTER 1: INTRODUCTION ...........................................................................................1

1.1 Definitions ........................................................................................................................................................ 5 1.1.1 Definitions of Heritage Language .............................................................................................................. 6 1.1.2 Heritage Language Learners ...................................................................................................................... 6 1.1.3 Heritage Language Learners and Non-Heritage Language Learners ......................................................... 9

1.2 Background ................................................................................................................................................... 11 1.2.1 Cultural and language policies in Australia ............................................................................................. 12 1.2.2 Chinese immigrants in Australia .............................................................................................................. 20 1.2.3 Chinese language in Australia .................................................................................................................. 26

1.3 Researcher’s subjectivity .............................................................................................................................. 28

1.4 Significance of the research .......................................................................................................................... 29

1.5 Thesis structure ............................................................................................................................................. 30

CHAPTER 2: LITERATURE REVIEW .............................................................................33

2.1 Commitment to CHL learning: ‘Motivation’ and ‘investment’ ................................................................ 33 2.1.1 Motivation: A social psychological perspective ....................................................................................... 34 2.1.2 Investment: A poststructuralist perspective .............................................................................................. 39

2.2 Identity construction in CHL learning ........................................................................................................ 44 2.2.1 Identity: A social psychological perspective ............................................................................................ 44 2.2.2 Identity: A poststructuralist perspective ................................................................................................... 48

2.3 Reframing through a Bourdieusian stance ................................................................................................. 54 2.3.1 ‘Motivation’, ‘investment’, and ‘capital’ ................................................................................................. 56 2.3.2 ‘Ethnic identity’ and ‘habitus’ .................................................................................................................. 60

2.4 Chapter summary ......................................................................................................................................... 62

CHAPTER 3: THEORETICAL FRAMEWORK ...............................................................65

3.1 Capital ............................................................................................................................................................ 66

v

3.1.1 Economic capital ..................................................................................................................................... 67 3.1.2 Cultural capital ........................................................................................................................................ 67 3.1.3 Social capital ........................................................................................................................................... 69 3.1.4 Symbolic capital ...................................................................................................................................... 70 3.1.5 Capitalising on language: Linguistic capital ........................................................................................... 72 3.1.6 Section summary ..................................................................................................................................... 75

3.2 Field ............................................................................................................................................................... 76

3.3 Habitus .......................................................................................................................................................... 81 3.3.1 Conceptualising habitus .......................................................................................................................... 82 3.3.2 Interpreting ‘Chineseness’ through habitus ............................................................................................. 89 3.3.3 Confucian dispositions ............................................................................................................................ 94

3.4 Chapter summary ...................................................................................................................................... 103

CHAPTER 4: METHODOLOGY ..................................................................................... 107

4.1 Philosophical position ................................................................................................................................ 107

4.2 Mixed methods research ............................................................................................................................. 110

4.3 Quantitative phase....................................................................................................................................... 114 4.3.1 Research questions and hypotheses ........................................................................................................ 115 4.3.2 Target population .................................................................................................................................... 116 4.3.3 Instrument design and development ....................................................................................................... 118 4.3.4 Pilot study ............................................................................................................................................. 126 4.3.5 Main study ............................................................................................................................................ 128

4.4 Qualitative phase ........................................................................................................................................ 139 4.4.1 Thematising ........................................................................................................................................... 141 4.4.2 Designing .............................................................................................................................................. 142 4.4.3 Interviewing .......................................................................................................................................... 149 4.4.4 Transcribing .......................................................................................................................................... 153 4.4.5 Analysing .............................................................................................................................................. 154 4.4.6 Reporting ............................................................................................................................................... 156

4.5 Ethical considerations ................................................................................................................................ 157

4.6 Chapter summary ...................................................................................................................................... 160

CHAPTER 5: QUANTITATIVE DATA ANALYSIS AND FINDINGS ......................... 161

5.1 Pilot study ................................................................................................................................................... 161

5.2 Main study .................................................................................................................................................. 168 5.2.1 Validity and reliability of measurement models: EFA and CFA ............................................................ 172 5.2.2 Hypothesis testing: SEM ....................................................................................................................... 223 5.2.3 Habitus and capital working together: Regression ................................................................................ 234

5.3 Chapter summary ...................................................................................................................................... 241

CHAPTER 6: QUALITATIVE DATA ANALYSIS AND FINDINGS ............................ 243

6.1 How were the five participants selected? ................................................................................................. 243

6.2 Who were the five participants? ................................................................................................................ 246

vi

Adam: “I just need to catch up!” ..................................................................................................................... 246 Bob: “Chinese is part of me.” ......................................................................................................................... 248 Crystal: Chinese entertainment stuff is “really funny”. .................................................................................. 249 Dianna: “I felt how important (it was) to speak my own home language!” .................................................... 249 En-ning: “Learning Chinese is definitely something I am going to pursue in the rest of my life.” ................ 250

6.3 What did CHL mean to the five participants? .......................................................................................... 251 6.3.1 Potential production of economic capital through CHL learning ........................................................... 252 6.3.2 Production of cultural capital through CHL learning ............................................................................. 253 6.3.3 Production of social capital through CHL learning ................................................................................ 257 6.3.4 Production of symbolic capital through CHL learning .......................................................................... 261 6.3.5 Making sense of the habitus of ‘Chineseness’ through CHL learning ................................................... 263 6.3.6 Capital in different fields ....................................................................................................................... 268 6.3.7 Catering to the field of forces ................................................................................................................ 273

6.4 Are the findings consistent? ........................................................................................................................ 276 6.4.1 Capital and CHL proficiency ................................................................................................................. 276 6.4.2 ‘Chineseness’ and CHL proficiency ....................................................................................................... 280 6.4.3 Quantitative coda ................................................................................................................................... 283

6.5 Chapter summary ....................................................................................................................................... 287

CHAPTER 7: DISCUSSION AND CONCLUSION .........................................................291

7.1 Discussions ................................................................................................................................................... 291 7.1.1 Habitus, capital, and practice in fields ................................................................................................... 292 7.1.2 Field of forces ........................................................................................................................................ 297

7.2 Contributions of the study .......................................................................................................................... 300 7.3.1 Originality and contributions to new knowledge ................................................................................... 301 7.3.2 Methodological contributions ................................................................................................................ 302 7.3.3 Practical contributions ........................................................................................................................... 303

7.3 Limitations of the study .............................................................................................................................. 306

7.4 Future directions ......................................................................................................................................... 308

7.5 Overall conclusions of the study ................................................................................................................ 309

Appendices ......................................................................................................................................................... 315 Appendix 1: Information sheet for participation in the online survey ............................................................ 315 Appendix 2: Information sheet for participation in the interview ................................................................... 316 Appendix 3: The consent form ....................................................................................................................... 317 Appendix 4: Interview schedule ..................................................................................................................... 318

Bibliography ...................................................................................................................................................... 321

vii

List of Figures

Figure 2.1. Summary of literature review .................................................................................................... 63 Figure 3.1. Forms of capital and language ................................................................................................... 75 Figure 3.2. Forms of capital and language in fields ..................................................................................... 80 Figure 3.3. Habitus and language in fields ................................................................................................... 89 Figure 3.4. Theoretical framework ............................................................................................................. 105 Figure 4.1. Theoretical framework guiding the quantitative phase ............................................................ 116 Figure 4.2. Structural relationship between ‘Chineseness’ and ‘CHL proficiency’ ................................... 138 Figure 4.3. Structural relationship between ‘capital’ and ‘CHL proficiency’ ............................................ 138 Figure 4.4. Theoretical framework guiding the qualitative phase .............................................................. 142 Figure 5.1. Age distribution of participants ................................................................................................ 169 Figure 5.2. Distribution of participants’ resident cities .............................................................................. 169 Figure 5.3. Composition of participants’ birthplace ................................................................................... 170 Figure 5.4. Composition of participants’ generation .................................................................................. 170 Figure 5.5. Composition of participant’s language usage at home ............................................................. 171 Figure 5.6. Composition of Chinese dialects used at home ....................................................................... 171 Figure 5.7. Scree plot of the ‘Chineseness’ scale ....................................................................................... 177 Figure 5.8. Measurement model for the construct ‘Chineseness’ .............................................................. 178 Figure 5.9. Modified measurement model for the construct ‘Chineseness’ ............................................... 181 Figure 5.10. Scree plot of the ‘economic capital’ scale .............................................................................. 187 Figure 5.11. Measurement model for the construct ‘economic capital’ ..................................................... 188 Figure 5.12. Scree plot of the ‘cultural capital’ scale ................................................................................. 193 Figure 5.13. Measurement model for the construct ‘cultural capital’ ........................................................ 194 Figure 5.14. Modified measurement model for the construct ‘cultural capital’ ......................................... 196 Figure 5.15. Scree plot of the ‘social capital’ scale .................................................................................... 201 Figure 5.16. Measurement model for the construct ‘social capital’ ........................................................... 202 Figure 5.17. Scree plot of the ‘symbolic capital’ scale ............................................................................... 206 Figure 5.18. Measurement model for the construct ‘symbolic capital’ ...................................................... 207 Figure 5.19. Modified measurement model for the construct ‘symbolic capital’ ...................................... 210 Figure 5.20. Scree plot of the ‘CHL proficiency’ scale .............................................................................. 215 Figure 5.21. Measurement model for the construct ‘CHL proficiency’ ..................................................... 216 Figure 5.22. Modified measurement model for the construct ‘CHL proficiency’ ...................................... 219 Figure 5.23. Structural model for the relationship between ‘Chineseness’ and ‘CHL proficiency’ ........... 224 Figure 5.24. Scree plot of the ‘capital’ scale .............................................................................................. 229 Figure 5.25. Measurement model for the construct ‘capital’ ..................................................................... 230 Figure 5.26. Structural model for the relationship between ‘capital’ and ‘CHL proficiency’ .................... 232 Figure 5.27. Histogram and P-P plots of the standardised residuals of the ‘CHL proficiency’.................. 239 Figure 5.28. Scatter plots of the standardised and the studentised residuals of ‘CHL proficiency’ against the

standardised predicted values of ‘CHL proficiency’ .................................................................................. 240 Figure 5.29. Partial plots of the residuals of ‘CHL proficiency’ against ‘capital’ and ‘Chineseness’ ........ 240

viii

List of Tables

Table 5.1 Item sets and their corresponding constructs in the pilot study .................................................. 164 Table 5.2 Summary results of internal consistency reliability test ............................................................. 165 Table 5.3 Item sets and their corresponding constructs in the main study ................................................. 166 Table 5.4 Descriptive statistics of indicators .............................................................................................. 172 Table 5.5 Correlation matrix for the indicators within the construct ‘Chineseness’ ................................... 174 Table 5.6 KMO and Bartlett's test for the indicators within the construct ‘Chineseness’ .......................... 175 Table 5.7 Communalities for the indicators within the construct ‘Chineseness’ ........................................ 176 Table 5.8 Total variance of the construct ‘Chineseness’ explained by its indicators .................................. 176 Table 5.9 Assessment of normality for the indicators within the construct ‘Chineseness’ ......................... 177 Table 5.10 Model fit indices for the construct ‘Chineseness’ ..................................................................... 178 Table 5.11 Standardised residual covariances for the indicators within the construct ‘Chineseness’ ........ 179 Table 5.12 Modification indices for the construct ‘Chineseness’ ............................................................... 180 Table 5.13 Assessment of normality for the indicators within the construct ‘Chineseness’ (modified) ..... 180 Table 5.14 Model fit indices for the construct ‘Chineseness’ (modified) ................................................... 181 Table 5.15 Proportional factor score weights for the indicators within the construct ‘Chineseness’ ......... 182 Table 5.16 Correlation matrix for the indicators within the construct ‘economic capital’ ......................... 185 Table 5.17 KMO and Bartlett's test for the indicators within the construct ‘economic capital’ ................. 185 Table 5.18 Communalities for the indicators within the construct ‘economic capital’ .............................. 186 Table 5.19 Total variance of the construct ‘economic capital’ explained by its indicators ......................... 186 Table 5.20 Assessment of normality for the indicators within the construct ‘economic capital’ ............... 187 Table 5.21 Model fit indices for the construct ‘economic capital’ ............................................................. 189 Table 5.22 Proportional factor score weights for the indicators within the construct ‘economic capital’ .. 189 Table 5.23 Correlation matrix for the indicators within the construct ‘cultural capital’............................. 191 Table 5.24 KMO and Bartlett's test for the indicators within the construct ‘cultural capital’ .................... 191 Table 5.25 Communalities for the indicators within the construct ‘cultural capital’ .................................. 192 Table 5.26 Total variance of the construct ‘cultural capital’ explained by its indicators ............................ 192 Table 5.27 Assessment of normality for the indicators within the construct ‘cultural capital’ ................... 193 Table 5.28 Model fit indices for the construct ‘cultural capital’ ................................................................. 194 Table 5.29 Standardised residual covariances for the indicators within the construct ‘cultural capital’ .... 195 Table 5.30 Modification indices for the construct ‘cultural capital’ ........................................................... 195 Table 5.31 Assessment of normality for the indicators within the construct ‘cultural capital’ (modified) . 195 Table 5.32 Model fit indices for the construct ‘cultural capital’ (modified) ............................................... 197 Table 5.33 Proportional factor score weights for the indicators within the construct ‘cultural capital’ ..... 197 Table 5.34 Correlation matrix for the indicators within the construct ‘social capital’ ............................... 199 Table 5.35 KMO and Bartlett's test for the indicators within the construct ‘social capital’ ....................... 199 Table 5.36 Communalities for the indicators within the construct ‘social capital’ ..................................... 200 Table 5.37 Total variance of the construct ‘social capital’ explained by its indicators ............................... 200 Table 5.38 Assessment of normality for the indicators within the construct ‘social capital’...................... 201 Table 5.39 Model fit indices for the construct ‘social capital’ ................................................................... 203 Table 5.40 Proportional factor score weights for the indicators within the construct ‘social capital’ ........ 203 Table 5.41 Correlation matrix for the indicators within the construct ‘symbolic capital’ .......................... 205 Table 5.42 KMO and Bartlett's test for the indicators within the construct ‘symbolic capital’ .................. 205 Table 5.43 Communalities for the indicators within the construct ‘symbolic capital’ ............................... 206 Table 5.44 Total variance of the construct ‘symbolic capital’ explained by its indicators.......................... 206 Table 5.45 Assessment of normality for the indicators within the construct ‘symbolic capital’ ................ 207 Table 5.46 Model fit indices for the construct ‘symbolic capital’ .............................................................. 208 Table 5.47 Standardised residual covariances for the indicators within the construct ‘symbolic capital’ .. 209 Table 5.48 Modification indices for the construct ‘symbolic capital’ ........................................................ 209 Table 5.49 Assessment of normality for the indicators within the construct ‘symbolic capital’ (modified)

.................................................................................................................................................................... 209 Table 5.50 Model fit indices for the indicators within the construct ‘symbolic capital’ ............................ 211 Table 5.51 Proportional factor score weights for the indicators within the construct ‘symbolic capital’ ... 211 Table 5.52 Correlation matrix for the indicators within the construct ‘CHL proficiency’ ......................... 213 Table 5.53 KMO and Bartlett's test for the indicators within the construct ‘CHL proficiency’ ................. 214

ix

Table 5.54 Communalities for the indicators within the construct ‘CHL proficiency’ .............................. 214 Table 5.55 Total variance of the construct ‘CHL proficiency’ explained by its indicators ......................... 215 Table 5.56 Assessment of normality for the indicators within the construct ‘CHL proficiency’ ............... 216 Table 5.57 Model fit indices for the construct ‘CHL proficiency’ ............................................................. 217 Table 5.58 Standardised residual covariances for the indicators within the construct ‘CHL proficiency’ . 218 Table 5.59 Modification indices for the construct ‘CHL proficiency’ ....................................................... 218 Table 5.60 Assessment of normality for the indicators within the construct ‘CHL proficiency’ (modified)

.................................................................................................................................................................... 219 Table 5.61 Model fit indices for the construct ‘CHL proficiency’ (modified) ........................................... 220 Table 5.62 Proportional factor score weights for the indicators within the construct ‘CHL proficiency’ .. 220 Table 5.63 Item sets and their corresponding constructs in the (modified) measurement models ............. 222 Table 5.64 Model fit indices for the structural model: ‘Chineseness’ and ‘CHL proficiency’ ................... 225 Table 5.65 Formulae for computation of the four composite variables ...................................................... 226 Table 5.66 Correlation matrix for the indicators within the construct ‘capital’ ......................................... 227 Table 5.67 KMO and Bartlett's test for the indicators within the construct ‘capital’ ................................. 228 Table 5.68 Communalities for the indicators within the construct ‘capital’ ............................................... 228 Table 5.69 Total variance of the construct ‘capital’ explained by its indicators ......................................... 228 Table 5.70 Assessment of normality for the indicators within the construct ‘capital’................................ 229 Table 5.71 Model fit indices for the construct ‘capital’ ............................................................................. 230 Table 5.72 Proportional factor score weights for the indicators within the construct ‘capital’ .................. 231 Table 5.73 Model fit indices for the structural model: ‘Capital’ and ‘CHL proficiency’ ........................... 233 Table 5.74 Formulae for computation of the constructs ‘Chineseness’, ‘capital’, and ‘CHL proficiency’ 235 Table 5.75 Correlation matrix for the variables .......................................................................................... 235 Table 5.76 Model summary ........................................................................................................................ 237 Table 5.77 ANOVA results ......................................................................................................................... 237 Table 5.78 Model parameters ..................................................................................................................... 237 Table 5.79 Collinearity diagnostics ............................................................................................................ 238 Table 6.1 Case-wise diagnostics ................................................................................................................. 244 Table 6.2 Overview of participants ............................................................................................................ 246 Table 6.3 Correlation between forms of capital and social practices ......................................................... 284 Table 6.4 Interview data summary ............................................................................................................. 288 Table 6.5 Contributing factors to CHL proficiency .................................................................................... 289

x

List of Abbreviations

ABS Australian Bureau of Statistics

ACARA Australian Curriculum Assessment and Reporting Authority

ANOVA Analysis of Variance

CFA Confirmatory Factor Analysis

CFI Comparative Fit Index

CHL Chinese Heritage Language

CHLL(s) Chinese Heritage Language Learner(s)

COAG Council of Australian Government

DEET Department of Employment, Education, and Training

EFA Exploratory Factor Analysis

ESL English as a Second Language

H0 Null Hypothesis

HL Heritage Language

HLL(s) Heritage Language Learner(s)

IFI Incremental Fit Index

KMO Kaiser-Meyer-Olkin

NFI Normed Fit Index

NHLL(s) Non-Heritage Language Learner(s)

RFI Relative Fit Index

RMSEA Root Mean Square Error of Approximation

RQ (s) Research Question(s)

SEM Structural Equation Modelling

SMC Squared Multiple Correlations

xi

TL Target Language

TLI Tucker-Lewis Index

VIF Variance Inflation Factor

xii

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet requirements for

an award at this or any other higher education institution. To the best of my knowledge and

belief, the thesis contains no material previously published or written by another person

except where due reference is made.

Signature:

Date: 17 June 2013

xiii

Acknowledgement

My PhD project was completed under the sponsorship of Queensland University of

Technology (QUT). I would like to take this opportunity to acknowledge the substantial

financial assistance that I received from the QUT Postgraduate Research Award.

I have complex and rich feelings about this PhD journey. During the entire life of this

journey, there have been struggles and passions, frustrations and achievements, stress and

pleasure, suffering and benefits, loneliness and teamwork, as well as many other feelings.

However, I was not alone along this journey. I felt how lucky I was to have so many people

around me, supervising me, helping me, supporting me, and encouraging me. I am deeply

grateful to these people.

First and foremost, I would like to extend my utmost gratitude to my supervisory team,

Associate Prof. Karen Dooley, Dr. Catherine Doherty, and Dr. Paul Shield, who introduced

me to this body of scholarship, guided me through this project, and pushed me to be the best

that I can. They were so engaged in my project that they contributed significant time to

supervise me, without which I would have never ever come to the end of this PhD journey. I

will remember their inculcation all through my life. As the Chinese saying goes, “一日为师，

终身为师” (Be my teacher for a day, be my teacher for a lifetime). Special thanks are also

given to Prof. Allan Luke, who shines the lights to guide me down the academic road that I

am walking along.

Secondly, my participants volunteered their time to complete the online questionnaire.

Five of them spent extra time to participate in the subsequent interviews. They shared their

stories and life experiences with me and helped me produce and enrich my data. I have given

my thanks to the five participants before and after the interviews, but I have no way to thank

the rest of my participants in person because they did not leave their contact details. I

xiv

therefore owe thanks to these people. This feeling urged me to work even harder on my

project. I have tried my best to improve the quality of this thesis to respect the contribution of

their time.

Thirdly, I would like to express my sincere gratitude both to the internal reviewers at

the confirmation and the final stage, and to the external examiners of my thesis. Their

constructive comments and thought-provoking advice helped me bring my thesis to a higher

level. I am also thankful to the staff from the former Centre for Learning Innovation for their

‘scaffolding’ role along my PhD journey: Prof. Carmel Diezmann, Ms. Jeannean Botha, Ms.

Jennifer Yared, Dr. Briony Wainman, Dr. Mary Clowes, Prof. Jo Brownlee, Associate Prof.

Julie Davis, Dr. Jo Lampert, Ms. Carol Partridge, Ms. Helen Tanner, and Ms. Rowena

McGregor.

Next, I would like to extend my thanks, from the depth of my heart, to all my friends

who supported me during the entire three years of my PhD journey, both in academic and in

daily life: Dr. Yaxing Zhang, Dr. Yang Hu, Ms. Jianwei Liu, Mr. Xiangguo Wang, Dr. Jeanne

Yang, Ms. Xinxin Fan, Dr. Juncheng Dai, Ms. Qiuxiang Huang, Dr. Janet Zhong, Dr.

Katherine Hanna, Mr. Charles Allen, Mr. Ark Du, Mr. Rui Jiang, Ms. Tracy Liu, Mr. Saif

Alamri, Dr. Jill Fox, Dr. Yuming Guo, Ms. Shandy Li, Dr. Ingrid Wang, Ms. Pei Liu, Dr.

Dagang Wang, Dr. Janet Hou, Associate Prof. Sue Walker, Dr. Louise Mercer, Dr. Lyn May,

Prof. Kar-Tin Lee, Mr. Christopher Meakin, Emeritus Prof. Sandra Vianne McLean A.M., Dr.

Debby Syu, Mr. Leon Zhang, Dr. Xiaofang Ye, Mr. Juming Shen, Ms. Maggie Kim, Dr.

Susan Sim, Dr. Li Yuan, Associate Prof. Lisa Ehrich, and Associate Prof. Deborah Henderson.

Without the friendship of these people, my PhD journey would have been a lonely one.

Last but not least, I am indebted to my beloved family members: Ms. Weiming Li (李

伟明), Mr. Shuhuai Mu (穆书淮), and Mr. Shizhuo Gui (桂士卓), who accompanied me

xv

through the ups and downs along my PhD journey, helped me build up my confidence, and

taught me never to give up. I hope I have not failed their love and expectations. I would also

like to take this opportunity to thank our family friends: Ms. Marion Welburn, Mr. David

Welburn, and Dr. Cassie Welburn. I appreciate their time spent on proofreading my work.

Their contribution helped so much in terms of the presentation of this thesis.

I owe thanks to all these people, who witnessed me struggle along and cheered me on.

Without the help, support, encouragement, concern, trust, and love of these people, I would

not have been able to complete this thesis. This fulfilment never belongs to me alone. Instead,

it belongs to all these beloved people. By presentation of this thesis, I would like to show

them how this PhD journey has become a miracle in my life!

1

Chapter 1: Introduction

Language learners’ commitment and identity have gained scholarly attention in second

language learning research. Many of the available studies on learners’ commitment and their

identity are concerned with English, a language that is either the medium of instruction or a

mandatory subject of study in many international contexts within and beyond

English-speaking nations. In contrast, questions about learners’ commitment to, and their

identity in, Chinese language learning are newly important for policymakers, educators, and

researchers.

The absolute number of Chinese language learners remains relatively low in many

non-Chinese-speaking countries, especially in the West. However, Chinese language

programs have been growing steadily worldwide since the 1990s (Xing, 2006). Consequently,

Chinese has secured its presence among the priority taught languages in African, European,

Australasian, and wider American settings (Lo Bianco, 2011). The success of these initiatives

turns ultimately on the willing commitment of effort, time, and other resources from learners

over an extended period in which learners construct certain understandings of themselves,

others, and the world. There is widespread agreement that the nature of this commitment and

understanding varies with social contexts and learners’ characteristics. Differences between

learners of Chinese origin and those of other linguistic and cultural groups are considered by

some to be consequential for successful Chinese language programs (He, 2006; Tsung &

Cruickshank, 2011; Xing, 2006). Yet, there is little agreement about the nature of these

differences and their consequences, and there is little research to guide professional and

policy work. Moreover, there are limited studies about Chinese populations learning Chinese

2

in countries where Chinese is neither the medium of instruction nor a mandatory subject of

study (He, 2005; D. Li & Duff, 2008), such as Australia.

The research reported in this thesis investigates Chinese Australians’ commitment to

Chinese language learning and the negotiation of their ethnic identity through their Chinese

language learning. The research also explores the meaning of Chinese language for Chinese

Australians. A point of clarification is in order here. ‘Chinese’ is an umbrella term embracing

many dialects used mainly in China but also in Chinese diasporas as far as South East Asia,

North America, Australasia, and the UK. In this respect, Chinese-speaking communities use

varieties of Chinese language that is distinguished by its high level of internal diversity.

Essentially, these dialects can be categorised into seven major mutually unintelligible spoken

but orthographically similar varieties of Chinese language. W. Li (1994) summarises these

varieties and their distribution in China as:

1. Beifang (northern, the native language of more than 70% of the Chinese population):

The majority dialect, Mandarin Chinese, is a sub-variety of northern Chinese. It is

based on the speech of the capital city Beijing and has enjoyed political and cultural

significance for centuries (He, 2006). It is currently the official language of China.

2. Yue (Cantonese, 5.1% of the Chinese population): The majority of Cantonese

speakers are in Guangdong province and Hong Kong.

3. Min (Hokkien, 4.1% of the Chinese population): Hokkien is spoken in Fujian

province, Taiwan island, and Hainan island.

4. Kejia (Hakka, 3.7% of the Chinese population): Hakka speakers come from small

agricultural areas and are now scattered throughout southeast China.

5. Wu (8.5% of the Chinese population): It is spoken in the lower Yangtze River

region, including urban, metropolitan centres such as Shanghai.

3

6. Xiang (4.8% of the Chinese population): It is mainly spoken in the south central

region.

7. Gan (2.4% of the Chinese population): It is spoken chiefly in the southeast inland

provinces.

For the purpose of this study, Chinese language includes all the above mentioned dialects.

‘Chinese Australians’ in this thesis refers to Chinese immigrants and their descendants who

have gained citizenship of Australia or permanent residency in Australia. In a ‘Chinese

learning Chinese’ context such as Australia, Chinese can be described as a ‘Heritage

Language (HL)’, a language other than English that is associated with an individual’s ethnic

or cultural background (Chinen & Tucker, 2005). When Chinese Australians learn Chinese as

a HL, they are considered ‘Chinese Heritage Language Learners (CHLLs)’. These learners

have Chinese ancestry, are raised in a home where Chinese may or may not be spoken, are

primarily educated in English, and may have a variable degree of Chinese language

proficiency (He, 2006). A deeper understanding of Heritage Language (HL) and Heritage

Language Learner (HLL) will be developed in Section 1.1.

According to the Australian Bureau of Statistics (ABS) 2011 Census, Chinese ancestry,

either alone or with another ancestry, was claimed by 866,200 permanent residents, while

there was a population of 574,200 who spoke Chinese at home. The statistics lead to some

questions. For those who speak Chinese at home, why might they be committed to learning or

speaking Chinese language? What does Chinese language mean in their lives? What role does

a sense of themselves play in the process of their learning and speaking Chinese? For the rest

who do not speak Chinese at home, what has happened? Have they shifted entirely to English

or are they struggling to maintain their CHL? How does CHL loss or CHL maintenance shape

their ethnic identity?

4

In Australia, CHLLs may be mistaken for native Chinese speakers in some situations,

largely because of looking Chinese; but they may also be criticised for their incomplete

mastery of Chinese in other situations (Ang, 2001). This results in a challenge, more or less,

for Chinese Australians to find a comfortable way of living within the overlapping boundaries

of Australian society and their heritage community. The core of the challenge is the internal

balance between nationally being Australian and ethnically being Chinese (Ang, 2001)

through the tensions between nationalism and multiculturalism in contemporary Australia.

This complex Australian context will be expounded in Section 1.2.

If Chinese Australians choose either to live a lifestyle of their Chinese heritage or move

completely into the Australian lifestyle, their choices appear mutually exclusive. Alternatively,

they may choose to lead lives that include elements of both languages and cultures. As a

reflection of their challenges, some fundamental but complex questions that Chinese

Australians keep asking themselves may be: Where am I from? Where do I belong? What

does Chinese heritage mean in my life? Why do I learn Chinese? How do my decisions to use

and learn Chinese and/or English affect my identity? These questions implicate ambiguity

and complexity in Chinese Australians’ identity and their commitment to CHL learning. This

ambiguity and complexity often reflects how Chinese Australians position themselves and

how they have been positioned, sometimes unfavourably and sometimes in contradictory

ways, by their family members, their peers, their communities, and their lived worlds.

The aim of this study is to examine the interrelationship amongst Chinese Australians’

‘Chineseness’, their commitment to CHL learning, and their CHL proficiency. To clarify, this

study understands ‘Chineseness’ as a set of propensities associated with people of Chinese

ancestry, embedded within their physical attributes, rooted in their Chinese cultural history

and heritage, and emergent from their family upbringing and social learning. This study is

particularly interested in young Chinese Australian adults. This reflects several considerations.

5

Firstly, the immigrant populations of young adults with Chinese ethnic backgrounds have not

received due attention in the literature (Levesque, 2007). Secondly, Australia is a complicated

lived social world for these young adults with Chinese ancestry. Australia’s peculiar historical

and cultural background will be discussed further in Section 1.2. For this reason, the current

study identifies young Chinese Australian adults as the subjects under investigation.

Accordingly, this study seeks to answer the following research questions:

Research Question One (RQ1): Is CHL proficiency of young Chinese Australian

adults influenced by their access to various resources, the strength of their

‘Chineseness’, or both?

Research Question Two (RQ2): What does CHL mean to young Chinese

Australian adults?

The aim of Chapter One is to introduce the thesis. Section 1.1 presents definitions of

HL and HLL. Heritage Language Learners (HLLs) are distinguished from Non-Heritage

Language Learners (NHLLs). Section 1.2 depicts the particular background of the study and

locates Chinese Australians in the context of learning CHL and being Chinese ethnically in

contemporary Australia. Section 1.3 poses the researcher’s subjectivity. Section 1.4 addresses

the significance of the study. At the end of this chapter, Section 1.5 outlines the structure of

this thesis.

1.1 Definitions

‘Heritage Language’ and ‘Heritage Language Learner’ are key terms used in the current study.

In this first section of the chapter, these terms are addressed in turn. These concepts are

6

defined from various perspectives in the literature. The definitions, with different foci,

emerge from, and function in, different contexts. The definitions that best fit the context of

this study are then developed.

1.1.1 Definitions of Heritage Language

The notion of ‘Heritage Language’ has existed for a long time along with various alternative

terms such as ‘home language’, ‘mother tongue’, or ‘community language’ (He, 2008). In the

United States, HL broadly refers to immigrant languages, indigenous languages, and colonial

languages (Fishman, 2001). An inclusive definition of HL used in the U.S. context views it as

a language of personal relevance other than English (Fishman, 1989). Fishman’s definition of

HL, although elegantly simple, overgeneralises to the extent that any language could be

claimed as a language of personal relevance by someone but this language may not

necessarily function as a HL. In English-speaking countries, HL denotes a language other

than English that is associated with one’s cultural background and may or may not be spoken

at home or formally learnt in schools (Chinen & Tucker, 2005; Cho, Cho, & Tse, 1997).

Given this definition, Chinese language, currently one of the major immigrant languages in

Australia, could be considered HL for Chinese Australians.

1.1.2 Heritage Language Learners

There are many definitions in the literature for the term ‘Heritage Language Learners’. These

definitions vary in terms of contexts and foci such as place, HL proficiency, heritage

membership, and identity. All definitions are valid and of value for particular contexts.

However, as HLLs cover a very heterogeneous population, there is no universal definition

capable of embracing all individuals under the heading of HLLs in all situations. The optimal

7

definition will function in a specific context where this definition emerges. Accordingly, the

following discussion develops a definition useful for this study.

HLLs in the United States are defined as individuals who are members of a community

with linguistic roots in a language other than English and who are learning the language of

that community (Cho, 2000). The main determinants of being a HLL in this definition are

association with a HL community and identification of membership in that community. By

this definition, proficiency in the HL is not taken into account. This definition may apply to

HLLs in a community that has a strong sense of HL roots. Where there are limited numbers

of HL speakers who are struggling to establish a community and are striving to reverse

language shift, this definition might be overly narrow.

Other definitions of HLLs extend to the ancestral language of an individual or group,

regardless of whether that language is still used at home (Fishman, 2001; Noels, 2005). HLLs

are individuals who have the desire to learn an ancestral language that was the language

spoken by previous generations of their families and that is not the language of the dominant

society (Cummins, 1998; Noels, 2005). Like the previous definition, this definition does not

take account of HL proficiency. However, it does highlight desire and intentional

commitment to HL learning.

Some definitions have emerged to take account of both heritage membership and HL

proficiency. One definition includes both learners who have either some degree of bilingual

proficiency or a cultural connection to the HL, and learners with a heritage motivation who

perceive a cultural connection that is more distant than that of first- or second-generation

immigrants (Van Deusen-Scholl, 2003). In an English-speaking country, HLLs are defined as

learners who a) are raised in and belong to a home and community where a language other

than English is spoken; b) speak or at least understand the HL; and c) are to some degree

bilingual in English and the HL (Valdés, 2001). These definitions indicate that both HL

8

proficiency and membership in a heritage community serve as criteria for determining HLL

status. This is a point of distinction from the definitions of Cho (2000), Cummins (1998),

Fishman (2001), and Noels (2005). However, the assumptions that HL is necessarily used at

home or in a heritage community and that HLLs have certain level of HL proficiency are

problematic.

Some definitions highlight the role of identity in the process of HL learning. HLLs are

those who have identity and/or linguistic needs with regard to HL learning that are related to

their heritage background (Carreira, 2004). In other words, their commitment to learning the

HL arises from a desire to connect more fully with their heritage or understand their identity

more deeply.

Considering the previous definitions, the term ‘Heritage Language Learner’ is used in

the current study to describe Chinese Australians who a) are Chinese learners with a cultural

and heritage connection to a Chinese language and its community; b) are educated primarily

in English; c) may or may not speak or understand a Chinese language; and d) may be

bilingual in a Chinese language and English language. This is an inclusive definition. It is

useful for the present study because it can describe the complexity of the Chinese Australian

population. Chinese Australians comprise a heterogeneous group ranging from those fluent in

Chinese language to those with no Chinese language at all. On one end of the spectrum, they

may be those who grew up with a Chinese home language other than English. On the other

end of the spectrum, they may be either third generation or further removed or inter-racial

adoptees with no exposure to CHL but feelings of cultural connection to CHL and Chinese

identity. The purpose of learning for this diverse population varies from CHL maintenance to

CHL development respectively. Nevertheless, CHLLs are very different from Non-Heritage

Language Learners.

9

1.1.3 Heritage Language Learners and Non-Heritage Language Learners

Heritage Language Learners are a particular class of language learners. Although they may

share certain characteristics and experiences with non-Heritage Language Learners, they

must nevertheless be conceptualised as language learners who possess distinct features from

NHLLs (Lo-Philip, 2010), in terms of their background knowledge of HL, familiarity with

the heritage culture, and forms of desire for HL learning (Campbell & Rosenthal, 2000).

Chinese and non-Chinese Australian learners of Chinese, for example, are likely to be

significantly different in at least some of these aspects. These aspects are discussed below.

Firstly, HLLs and NHLLs may differ in language proficiency. Due to limited exposure

to the HL and culture, HLLs may differ from First Language Learners of the same language

in that they often do not fulfil the linguistic competency in the HL at the same level of First

Language Learners (Carreira, 2004). However, there is a general agreement on the growing

evidence that even limited childhood language experience can help in the acquisition of that

language later in life (Au, Knightly, Jun, & Oh, 2002). In line with this perspective, HLLs are

more likely than Second Language Learners or Foreign Language Learners to have had some

prior exposure to the language, though often limited to early childhood, and therefore they

may become proficient in the language in ways that Second or Foreign Language Learners

are not and may never be. Prior linguistic and cultural knowledge of HL makes HLLs more

likely to achieve better language skills than Second or Foreign Language Learners,

particularly in speaking and listening (Ke, 1998; Shen, 2003; Xiao, 2006). On the other hand,

given differences in acquisition contexts, HLLs may have wide gaps in HL knowledge and

skills that NHLLs do not have (Carreira, 2004). For example, HLLs may need to connect oral

language skills developed through prior exposure to the language with the written form of the

language. In contrast, NHLLs often need to develop oral and written skills simultaneously at

the outset. Therefore, unlike NHLLs, HLLs may be quite comfortable with speaking and

10

listening but may often possess limited skills in reading and writing (A. Louie, 2004; Xiao,

2006). These gaps may prevent them from performing certain kinds of reading and writing

communicative tasks at a level commensurate with their listening and speaking skills (J. B.

Webb & Miller, 2000).

Secondly, most distinctions between HLLs and NHLLs are concerned with the way

that many HLLs deal with issues of identity (Peyton, Ranard, & McGinnis, 2001). HLLs

appear to have a multi-faceted identity as someone who is both similar to and different from

members of the heritage culture since they are socio-historically connected with the heritage

culture and yet experientially displaced from it (He, 2004). To a greater extent than NHLLs,

HLLs are likely to be driven by identification with the intrinsic cultural, affective, and

aesthetic values of the language (He, 2006), and by the goal of gaining greater access and

depth to their ethnic identity (Chinen & Tucker, 2005; He, 2006). Therefore, HLLs often

bring with themselves a set of ambiguities and complications in identity, which may be

absent in NHLLs (He, 2006).

Thirdly, HLLs and NHLLs also differ in terms of commitment to language learning.

HLLs often have different desires for studying the language than NHLLs for they are seeking

greater understanding of their heritage culture or stronger connection with their family and

community (Draper & Hicks, 2000). HLLs are committed to HL to communicate with family

members, to participate in the life of their HL community, to feel connected to their roots, to

contribute to the preservation of their beliefs and practices, or to overcome feelings of being

an outsider in their heritage community (Carreira, 2004). Of course, HLLs may also be

committed to HL learning for additional reasons such as to fulfil a language requirement as

part of a course or to pursue a career. Therefore, the commitment of HLLs to HL learning is

likely to be derived not only from pragmatic or utilitarian orientations, as tends to be the case

with NHLLs, but also from intrinsic cultural connections to the HL.

11

To help understand the nature and dynamics of Chinese Australians qua CHLLs, the

next section will introduce the complex cultural, social, and historical Australian context,

within which Chinese Australians negotiate tensions between nationally being Australian and

ethnically being Chinese.

1.2 Background

The compression of time and space by the innovation of information technology and the

intensification of transnational migrations of people are creating a more or less borderless

world. Mobility and connections are no longer as restricted to local regions as before. The

world is thus moving from territorially distant and mutually distinct spaces towards

multi-dimensional space of overlapping layers. There is a transition underway from a socially

distinct, culturally homogeneous, and politically sovereign world of nations to an

interconnected and intermingled one (Ang, 2001). People are intersecting and interacting in

such a world where they are on the move as they have never been before. This migration has

shaped the present time as an era of growing diversity. It has introduced new cultures and

languages to destination countries, and rapidly increased the cultural and linguistic diversity

of modern societies.

Against the background of regional and global migration as well as transnational

communications and interactions, the current section explores and understands the cultural

and linguistic diversity in Australia, a space of complicated entanglement. This section also

introduces the diverse cultural and linguistic backgrounds, needs, aspirations, and potentials

of Chinese Australians, who are faced with the challenges and opportunities in Australia, a

place of “togetherness-in-difference” (Ang, 2001, p. 17).

12

1.2.1 Cultural and language policies in Australia

Australia was already inhabited and owned by Aboriginal and Torres Strait Islander peoples

before the history of Australian immigration started. The early interactions between these

peoples and those from nearby islands were documented in archaeological work (Macknight,

1976): Fishermen from Makassar in the southern Celebes (Sulawesi) frequently visited the

coast between the Cobourg Peninsula and the Sir Edward Pellew Group in the early

eighteenth century. Given these early interactions, it is not surprising that some Aboriginal

languages contain a substantial number of Makassan words (Macknight, 1976).

The first cohort of British immigrants landed on this continent at Botany Bay on 26

January 1788. Since this historical moment, people with different cultures and languages

have immigrated to the continent from different parts of the world. The immigrant population

of Australia continues to grow rapidly. As shown in the 2011 Australian Census, three out of

ten Australian residents were born overseas, compared with one out of ten and two out of ten

in 1947 and 1971 respectively (Castles & Davidson, 2000). A wealth of linguistic and cultural

resources is available in Australia through its culturally and linguistically diverse population.

This diversity will continue to be a characteristic of future Australia (Holmes, Hughes, &

Julian, 2007).

The wealth of cultural diversity in Australia has been valued and supported by the

nation through its implementation of multiculturalism as a political strategy to accommodate

and include ethnic minorities (Ang & Stratton, 2001). However, in Australia, tensions exist

between a ‘unified nation’ and a ‘multicultural one’ (Holmes et al., 2007). A narrative gloss of

these tensions reflected by Australia’s immigration policy over complex processes could

include (Holmes et al., 2007): 1901 to the early 1970s (White Australia policy); 1950s to the

early 1970s (dismantled White Australia policy with assimilation); and the early 1970s

onwards (multiculturalism policy).

13

One of the first pieces of legislation enacted by the new national parliament upon

Federation was the Immigration Restriction Act 1901. This act was meant to manage

competition and labour disputes in the goldfields and growing tensions around Australian

nationalism. The Act signified the commencement of the so-called ‘White Australia Policy’

and formed its legislative basis. The White Australia Policy, including the 1901 Act and the

subsequent acts to strengthen it, intentionally restricted non-white immigration to Australia

from 1901 to the early 1970s.

During World War II, the arrival of thousands of refugees in Australia with the

Japanese advance; the service of numerous Australian Aborigines, Torres Strait Islanders,

Papua New Guineans, and Timorese in the frontline to defend Australia; and Australia’s

vulnerability in the Pacific War because of its small population brought Australia’s racially

discriminatory immigration and political rights into focus. After the conclusion of World War

II, the White Australia Policy was dismantled in stages by successive governments with the

encouragement of the first non-British and later non-white immigrants. Wartime service also

gave many Indigenous Australians confidence in the claim of their equal rights upon return to

civilian life.

In the post-war period, assimilation became the dominant policy in Australia, where

assimilation refers to the process by which indigenous people and immigrants were expected

to relinquish their culture and become indistinguishable from the Anglo-Australian dominant

group (Holmes et al., 2007). The ideal or illusion of national-cultural homogeneity was that

all indigenous people and immigrants would be incorporated smoothly within the dominant

national culture, leaving their original cultures happily behind. The dominant culture was

considered the national culture on which national identity was to be constructed. Ethnic

cultures were defined by the dominant culture as ‘different’. The national identity was

homogeneous and it denied ethnic cultures that threatened it.

14

The ‘assimilation’ model was a political strategy to balance inclusion and exclusion.

On the one hand, the model incorporated ethnic cultures only insofar as they could contribute

to the ideological discourse of cultural diversity in Australia. On the other, this model, per se,

marked immigrants as ‘others’ to the dominant group. As such, assimilation can never be

fully successful. The acquired, rather than inherited or ascribed character of cultural traits

gained in the process of assimilation turns the assimilating subjects into less than ‘real’ and

still inferior Australians (Ang, 2001). This led to the introduction of multiculturalism, which

marked the dismantling of the White Australia Policy in the early 1970s.

In the 1970s, the Australian Ethnic Council adopted a formal statement against

assimilation and immigration policy moved from assimilation to multiculturalism (Holmes et

al., 2007). Since then, multiculturalism has become a key element of government cultural

policy and the description of Australia as a multicultural nation has become commonplace

(Ang & Stratton, 2001; Holmes et al., 2007). However, a distinction between multicultural

Australia and multiculturalism in Australia has been made in terms of the distinction between

practice and policy (Hodge & O’Carroll, 2006). ‘Multicultural Australia’ refers to the

cultural diversity in which interaction between different cultures is in process (Holmes et al.,

2007). It is a shifting and dynamic interweaving of cultures and diversities actually happening

and existing in Australia. In contrast, ‘multiculturalism’, as a centrepiece of national cultural

policy, refers to the management of cultural diversity (Ang & Stratton, 2001) but within

certain and well-demarcated limits without disturbing or threatening national unity (Ang,

2001). This policy expects that all members of society have equal rights, regardless of ethnic

background (Holmes et al., 2007). It is an ideology to prescribe what should be happening or

speculate what could be happening by the acceptance and promotion of multiple cultures in

the demographic context of a specific place, such as organisations, communities, cities or

nations.

15

Multiculturalism, as a national policy in Australia, has a complicated history. Firstly,

the adoption of multiculturalism functions to discard the racist past in Australia. The

exclusionary and homogeneous White Australia Policy was replaced by multiculturalism, a

discourse in favour of pluralism and heterogeneity and recognition of valuable cultural

diversity. Secondly, multiculturalism functions to convince Australians of the public fiction

that they live in a harmonious and inclusive society. Multiculturalism becomes an ideological

discourse addressed to Australians to promote unity within diversity. Thirdly, the ideological

discourse of multiculturalism has a political orientation. It functions to serve the policy of the

government. Politicians have announced with pride that Australia is one of the most

successful multicultural societies in the world, though support for the creation of

multicultural Australia has always been less than whole-hearted (Ang & Stratton, 2001).

Language policies in Australia largely reflect Australia’s complicated cultural history.

During the ‘White Australia’ era, English was assumed to be the national and only necessary

language. Immigrants were expected to learn English, leaving their native languages happily

behind. There was no concept of multiculturalism as there is today. In the 1950s and 1960s,

the White Australia Policy was gradually dismantled. Although the period saw little change

from the traditional and classical ‘English only’ approach, there were remarkable

improvements in teaching English as a Second Language (ESL), in order to enable

immigrants arriving in Australia from different parts of the world to assimilate more readily.

The Adult Migrant Education Program was very innovative in its approach to English

teaching and learning to support ‘assimilation’. However, White culture was still the

dominant core surrounded by multiple minority ethnic cultures.

From the dysfunction of the White Australia Policy in the early 1970s (Holmes et al.,

2007) to the mid 1990s, Australia probably led the English-speaking world in systematic

language policy making (Ingram, 2000). The 1970 review, Teaching Asian Languages and

16

Cultures in Australia, made the first attempt to direct language policy towards Asia

(Commonwealth Advisory Committee, 1970). The review reflected Australia’s growing

political realignment towards Asia, but paid little attention to the economic importance of

Asian languages to Australia. Following the review, there was a sudden realisation that

Australia, almost accidentally, imported numerous languages that would be wasted if no

language policy enabled them to survive and to be used; and that many children entered

schools speaking their HL and needed an opportunity to commence their education in their

HL. Accordingly, there were many unsystematic attempts to reform language education. The

Department of Education advocated that there were “strong educational and social reasons for

migrant children” entering schools with “inadequate” English proficiency to continue the

learning of their HL (1976, p. 35). Guided by this initiative, funds were allocated to schools

to establish HL programs and to community schools to support the teaching of HL.

Another important document during the multicultural period of the 1970s was the

Galbally Report released by the Australian Federal Government in 1978. It took the position

that “every person should be able to maintain his or her culture without prejudice or

disadvantage and should be encouraged to understand and embrace other cultures” (Galbally,

1978, p. 4) and further that “cultural and racial differences must be reflected in educational

programs designed to foster intercultural and interracial understanding” (pp. 104-105). These

fundamental initiatives, though unsystematic, were to encourage enrichment and diversity

within an adherence to certain core values in the society. These initiatives also provided the

foundation for the development of the first National Policy on Languages (Lo Bianco, 1987)

in the English-speaking world.

On its release, the National Policy on Languages became the standard against which

State and Territory policies could be compared. This policy renewed language policy in

Australia, stressing such aspects as the support for the maintenance of HLs, Aboriginal and

17

Torres Strait Islander peoples’ languages, and the extensions to programs for teaching ESL. It

made a wide range of recommendations, taking account of the breadth of the social context in

which the policy was placed (Ingram, 2000). Moreover, the policy focused on the labour

market and the ways in which tackling adult illiteracy levels, extending English proficiency,

and teaching ‘trade languages’ would benefit Australia’s economic performance (Lo Bianco,

1990). To a large extent, the National Policy on Languages culminated the phases of

pluralistically oriented language policy of the 1970s and 1980s (Lo Bianco, 2000). However,

the policy lacked rigorous frameworks to address teacher education and teacher supply.

Funding distributed to stimulate the teaching of other languages was mainly through projects.

Unfortunately, when such funding dried up, the long-term effect of the projects was

questionable. There was scant on-going monitoring and evaluation of the policy and its

programs to ensure that the policy was continually evolving in response to emerging needs.

These deficiencies of the National Policy on Languages contributed to the development of

the Australian Language and Literacy Policy released by the Department of Employment,

Education, and Training (DEET) in 1991 (Ingram, 2000).

The Australian Language and Literacy Policy explicitly claimed to be derived from

and closely influenced by the National Policy on Languages (DEET, 1991). The 1991 policy

had the basic position that “Australian English” was the “national language”, but the national

priority languages, such as Asian languages, helped to “enrich the intellectual and cultural

vitality” of Australia and secure the “future economic well-being” of Australia (DEET, 1991,

pp. iii-iv). Accordingly, it foregrounded English and Asian languages and tied the skills of

these languages to education, employment, and trade. Despite the value attached to

multiculturalism and the maintenance and development of Aboriginal and Torres Strait

Islander peoples’ languages and HLs, the policy reflected the prevailing economic rationalism

that had dominated Australian language policy thinking and making since the late 1980s

18

(Ingram, 2000). In other words, it placed more emphasis on economic reasons for language

education than cultural or multicultural importance. As such, it focused more on language

skills and problems than language rights and resources.

As a complement to the Australian Language and Literacy Policy, another policy,

Asian Languages and Australia’s Economic Future, released by the Council of Australian

Governments (COAG) in 1994, continued to highlight the importance of teaching and

learning of Asian languages, in particular the four priority languages of Japanese, Chinese,

Indonesian, and Korean. Similar to the previous policy, the 1994 report emphasised the

economic value of Asian language skills as tools to facilitate competitive international trade.

The report set proficiency targets to be achieved in schools, encouraged early commencement

of learning Asian languages in Year 3, and recommended that a nationally agreed minimum

skill level be specified for Asian language teachers (COAG, 1994). However, the report was

criticised for being impractical in its targets and assumptions (Lo Bianco, 2000). An

under-supply of qualified Asian language teachers stood against its implementation. A further

criticism concerned the over-emphasis on economic reasons for language learning and the

neglect of cultural and intellectual values (Ingram, 2000).

Building on previous experiences, the Australian Curriculum, Assessment, and

Reporting Authority (ACARA) released the Shape of Australian Curriculum: Languages in

2011 to guide the development of language curriculum in Australia. The language curriculum

recognised the importance of a capability in languages in economics, diplomacy, trade,

cultural exchange, and national security. The language-specific curricula will be developed

for different languages and different groups of language learners that include First, Second,

and Heritage Language Learners. Within each of these groups there are differences in

proficiency in using the target language, with the span of language experiences of HLLs

particularly wide and the affiliations with their HLs particularly diverse (ACARA, 2011).

19

Being aware of these differences, the language curriculum designed the aims, pathways,

programs, hours of study, and achievement standards for language learning. Therefore, the

language curriculum is expected to improve language education in Australia, increase

investment in research and in public services dealing with languages, support ESL programs

for immigrants and the intergenerational maintenance of immigrants’ HLs, ensure recognition

of Aboriginal and Torres Strait Islander peoples’ languages, and facilitate language revival

initiatives for indigenous languages at risk of extinction (ACARA, 2011).

Being conscious of the “unstoppable” transformation of the Asian region into the

economic powerhouse of the world and the “gathering pace” of this transformation, the

Australian Federal Government released the Australia in the Asian Century White Paper in

October 2012, aiming to help “seize the economic opportunities” and “manage the strategic

challenges” of this development (Australian Government, 2012, p. ii). To broaden and deepen

the people-to-people links between Australia and the Asian region, particularly to strengthen

the nation’s principal relationships with China, India, Indonesia, and Japan, the White Paper

urged Australians to become more “Asia literate” and “Asia capable” (Australian

Government, 2012, p. iii). Detailed strategies for studies of Asia will become a core part of

school education and all schools will engage with at least one school in Asia to support the

teaching of a priority Asian language. These languages are Chinese (Mandarin), Hindi,

Indonesian, and Japanese. All Australian students will be encouraged to study and will have

access to at least one priority Asian language. Measures to track how Australian students are

increasing their knowledge of Asia will be developed. Industries and communities will be

encouraged to increase their understanding of the benefits of learning priority Asian

languages so that the demand for Asian language studies will be boosted. Australian

universities will be supported to increase the number of students who undertake Asian studies

and Asian languages as part of their university education, and will be encouraged to establish

20

an exchange arrangement involving transferable credits with at least one major Asian

university. Funding will be available, through the Australian Research Council and other

mechanisms, to strengthen research and teaching links between Australian and Asian

institutions. In brief, support and encouragement of Asian languages through school

education, university teaching and research, and industrial and community engagement are

key elements of the White Paper. The White Paper not only addresses the economic demand

for Asian languages but also makes the cultural diversity available in Australia through its

support and encouragement of language pluralism (Australian Government, 2012).

To sum up, the current model of multiculturalism and language policy in Australia

conceptualises ethnicity, national identity, and globalisation. This is a model of ethnicity as

flux and flow, national identity as a heterogeneous construction, and globalisation as a set of

ongoing processes that challenge the boundaries and sovereignty of the nation-state (Holmes

et al., 2007). This model reconfigures Australian nationalism, shifting from a racially

exclusionary form to an inclusive and open-ended form. This model offers a view from the

margins of Australian society (Holmes et al., 2007). It challenges the dominant group’s views

on ethnicity and nationalism. It is a vision of negotiated social cohesion predicated on the

acceptance of cultural difference as an integral part of contemporary Australia. It is within

these complex cultural and language contexts that Chinese Australians negotiate their

identities and learn CHL, willingly or unwillingly.

1.2.2 Chinese immigrants in Australia

As immigrants from every corner of the world continue to seek entry into Australia,

Australian cultural and language policies have to deal with the question of how to manage

proliferating differences and diversities. These policies have a remarkable impact on

21

immigrants in Australia. As one of the earliest non-White settlers, Chinese immigrants and

their descendents have formulated various forms of life politics in response to the shifting

cultural and language policies in Australia. The following discussion focuses particularly on

Chinese immigration in Australia, looking at both the impact of cultural and language policies

of Australia and that of the historic situations of China.

The Chinese people have a long history of emigration overseas. One of the earliest

emigrations can be dated back to the Ming Dynasty when He Zheng (1371-1435) became the

envoy of the Ming Dynasty. He sent people to explore and trade in the South China Sea and

in the Indian Ocean. The great Chinese diaspora began in the 19th century, the boom age of

European colonialism. Due to the lack of labourers in many colonies and the poverty brought

by population pressure, natural disasters, heavy taxation, destructive rebellions, and

widespread opium smoking in southeast China, many Chinese emigrated from southeast

China to work in Southeast Asia, where they had established earlier links since the Ming

Dynasty. Many other Chinese migrated to countries in North America and Australasia, where

there was a great demand for labour in gold mining and railway construction. On the one

hand, widespread famine and poverty in southeast China impelled many Chinese to work in

these countries so that they could earn more money to improve the living conditions of their

family members and relatives. On the other hand, the demand for labour in these countries

was addressed. Since the early 19th

century, Chinese emigration has been directed primarily

to North America and Australasia.

From the very beginning of Australian immigration history, links with China were

established when several ships dropped off their convict load in Australia then sailed for

southeast China to pick up goods for their return to England (Cushman, 1984). The earliest

documented Chinese immigration to Australia dates back almost two hundred years, with

Mak Sai Ying being the first recorded Chinese settler in 1818 and with the first large group of

http://en.wikipedia.org/wiki/South_China_Sea

http://en.wikipedia.org/wiki/Indian_Ocean

http://en.wikipedia.org/wiki/Chinese_Migration

http://en.wikipedia.org/wiki/Colonialism

http://en.wikipedia.org/wiki/Ming_Dynasty

http://en.wikipedia.org/wiki/Gold_mining

http://en.wikipedia.org/wiki/Railway

22

immigrants arriving from southeast China in October 1848 (Cushman, 1984). From this time

onwards till the late nineteenth century, Chinese immigration was seen as part of a solution to

labour shortages in Australia (Cushman, 1984).

In the mid 1850s, many more Chinese gold-seekers arrived at the various diggings.

Large numbers of early Chinese immigrants worked on goldfields. Others started to open

stores and became merchants and hawkers. Over the 1860s and 1870s, fishing and fish curing

industry were commonly operated by Chinese immigrants in Sydney, supplying dried fish to

Chinese people throughout New South Wales as well as Victoria. Stores and dormitories run

by Chinese people soon developed to support the miners on the fields as well as those on

their way to the diggings or back to China. The attraction of gold and competition in gold

mining aroused the resentment of European diggers towards Chinese diggers due to cheaper

Chinese labour. At the same time, attempts to maintain Chinese as indentured labourers were

difficult because some Chinese diggers often deserted their employers for more lucrative

individual gold mining. These problems regarding the Chinese population drew the

government’s attention and resulted in restrictive anti-Chinese legislation in the late 1850s

and the early 1860s. The objection at this time to Chinese immigration was economic

competition and cultural differences between the white and Chinese ways of life, rather than

feelings of racial superiority (Choi, 1975).

These restrictive measures effectively reduced the inflow of Chinese. But with a labour

shortage in the gold-mining industry and the dramatic decrease of Chinese immigrants, the

government repealed the restrictive acts in the 1860s (Choi, 1975). The Chinese population

continued to decrease for a short while after the repeal but then rapidly increased. Again, the

government became concerned with the growing number of Chinese in the mining areas and

feared that conflict between Chinese and European diggers and breaches of law and order

might occur. This resulted in restrictive Acts in the 1880s.

23

Despite the shifting immigration laws, Chinese immigrants managed to remain and

often prosper (Choi, 1975). While the major body of Chinese immigrants was digging on the

goldfields, many others endeavoured to try other ways of earning a living very soon after

their arrival. By the 1890s, Chinese people were represented in a wide variety of occupations

as scrub cutters, cooks, tobacco farmers, market gardeners, cabinet-makers, storekeepers,

laundry workers, and drapers, though by this time the fishing industry seemed to have

disappeared. They ran stores, import trade, and several Chinese language newspapers. They

were also part of an international community involved in political events in China such as

sending delegates to the Peking Parliament and making donations at times of natural disaster.

By the 1890s, the colonial parliaments had placed a series of restrictions on the

migration of ‘coloureds’ in general and Chinese in particular (Jones, 2005). Continuing

competition and labour disputes in the goldfields, as well as Australian nationalism, created

an environment of racial antagonism during the second half of the 19th

century. This led to the

new Federation’s Immigration Restriction Act 1901, the so-called White Australia Policy. It

was this policy that severely curtailed the development of the Chinese communities from the

late 19th

century onwards and led to the steady decline of Chinese population in Australia

(Jones, 2005). Continued discrimination, both legal and social, reduced the occupational

range of Chinese people. Gardening became one of the major occupations. Chinese

immigrants, throughout the first 30 years or so of the 20th

century, relied on successful

Chinese merchants to assist them to negotiate with the Immigration Restriction Act

bureaucracy. It was the rise of a new generation of Australian-born Chinese people, combined

with new immigrants that the merchants and others sponsored, that prevented the Chinese

population from dramatically declining (Choi, 1975).

At the beginning of World War II, Australian-born Chinese began to outnumber

China-born Chinese for the first time. However, with the Japanese invasion in China and

24

other Asia-Pacific regions, large numbers of refugees boosted the numbers of arrivals from

China again (Jones, 2004). Some were Chinese crewmembers who refused to return to

Japanese-held areas. Others were residents of the many Islands of the South China Sea

evacuated in the face of the Japanese advance. Still others were of Australian birth and were

able to leave Hong Kong for Australia on the approach of the Japanese (Jones, 2004).

After the conclusion of World War II, the White Australia Policy was gradually

dismantled. At the same time, restaurants began to replace gardening as the major source of

employment and avenue for bringing in new Chinese immigrants. These changes brought

about the end of the dominance of south China in the link between China and Australia that

had existed for over 100 years. For the first time numbers of Chinese immigrants from

non-Cantonese speaking parts of China significantly increased. During the decades from the

1950s to the 1970s, Australia’s restrictive migration policy was gradually relaxed before its

formal abolition. However, following the establishment of the People’s Republic of China in

1949, movement from China was most difficult, with a small number of Chinese arriving

each year. There was almost an absence of Chinese emigration from the 1950s to the 1970s.

During this period, China placed severe restrictions on the movement of its citizens.

From 1973, the White Australia Policy was for all practical purposes defunct. In 1975,

the Australian government passed the Racial Discrimination Act, which signified the end of

the White Australia Policy. Multiculturalism started as official policy in Australia. Only a few

years later, China opened its door again to the world. The end of the White Australia Policy

and the commencement of the Chinese Opening-up Policy saw new arrivals from China in

Australia. Since then, immigrants from the Chinese Mainland arrived in increasing numbers.

From the 1970s onwards until the end of the twentieth century, there were two significant

booms of emigration from China, in the late 1970s and the early 1990s respectively. In the

late 1970s, emigration restrictions in China were eased as a result in part of the Opening-up

25

Policy. More liberalised emigration policies were enacted to facilitate the legal departure of

increasing numbers of Chinese who joined their overseas Chinese relatives and friends. At the

same time, the ‘Four Modernisations Program’ that required Chinese students and scholars,

particularly scientists, to attend foreign education and research institutions brought about

increased contact with the outside world. Anyone who had the necessary economic resources

could apply for permission to study abroad. Other political events in China also promoted

immigration. In 1984, Britain agreed to transfer sovereignty of Hong Kong to China, and in

1989, a student movement broke out at Tiananmen Square. These two historical events

triggered a lack of confidence in the Chinese government and accelerated another wave of

migration to the United Kingdom, Australia, Canada, USA, and other parts of the world in the

early 1990s. The wave calmed after Hong Kong’s transfer of sovereignty in 1997.

Australian immigration policy has been transformed from ‘White Australia’ to

‘multiculturalism’. Almost at the same time, Chinese diplomatic policy has changed from

‘closing-down’ to ‘opening-up’. These policy changes saw new arrivals from China to

Australia. New institutions were established for these arrivals and old ones, such as the

Chinese Chamber of Commerce, revived. Chinese language newspapers were once again

published. Shortwave and longwave radio channels began to broadcast in Cantonese and

Mandarin. The Australian public broadcaster SBS also provided television and radio

programs in both languages. The Chinese language, mainly Mandarin and Cantonese, became

available as a subject in some secondary schools. Community language schools started to

operate on weekends. Several Chinese Australians have received the Order of Australia award

and there are current Chinese Australian representatives in both State and Federal parliaments

(Jones, 2005).

The population with Chinese ancestry keeps increasing in Australia. Today, Chinese are

the third largest group among all immigrants in Australia, just behind people from the United

http://en.wikipedia.org/wiki/Hong_Kong

http://en.wikipedia.org/wiki/Shortwave_radio

http://en.wikipedia.org/wiki/Standard_Cantonese

http://en.wikipedia.org/wiki/Standard_Mandarin

http://en.wikipedia.org/wiki/Television

http://en.wikipedia.org/wiki/Radio

http://en.wikipedia.org/wiki/Secondary_school

http://en.wikipedia.org/wiki/Order_of_Australia

http://en.wikipedia.org/wiki/Australian_Parliament

http://en.wikipedia.org/wiki/Australian_Parliament

26

Kingdom and New Zealand (ABS 2011 Census). With Chinese immigration, Chinese

language was brought to Australia. It is now the most widely spoken language other than

English at home (ABS 2011 Census). The following section provides some background

knowledge about Chinese language in Australia.

1.2.3 Chinese language in Australia

Chinese immigrants have brought and continue to bring Chinese language to Australia. With

the increase of Chinese immigrants in Australia, the population that speaks Chinese at home

has also increased. Where it was 401,357 in 2001 (ABS 2001 Census), it was 574,200 in 2011

(ABS 2011 Census). This makes Chinese the most common of all languages other than

English brought to Australia. Three percent (3.0%) of the national population speaks Chinese

at home (ABS 2011 Census).

On the international stage, China continues to speak to the world about its sheer

vastness, its huge population, its fascinating history and culture, its idiosyncratic Chinese

socialism, and its rapid economic growth. There has been growing interest in China as an

emerging international power and a potential cooperative business power (Zhang &

Slaughter-Defoe, 2009). For Australia, China is now a regional neighbour and trade partner. It

has been recognised that the development of the relationship with China and mutual benefits

of such require a solid pool of Australians in a range of sectors who can understand China

deeply and speak Chinese well (Orton, 2008). This strategic viewpoint was stressed in the

Australia in the Asian Century White Paper released by the Australian Government in

October 2012.

The teaching and learning of Chinese began two decades ago to produce

Chinese-speaking graduates from Australia’s schools to serve the country’s economic,

27

cultural, and political interests (Sturak & Naughten, 2010). However, the share of Australian

students learning Chinese is relatively small and has fallen in recent times. In recognising this

decline as well as the need to build a sound knowledge of Chinese in schools, the Australia in

the Asian Century White Paper advocates that the languages component of the Australian

Curriculum will enable all students to learn a language other than English, with a curriculum

for Chinese being one of the first in development. Compared to the previous situation,

ambitious targets have been set, especially in curriculum design, assessment mechanisms,

textbook development, and in the diversity of Chinese programs offered.

There has been increasing attention to teaching and learning Chinese as a Foreign

Language in Australia, not only due to the effort of the Australian Government but also

because of the recent sponsorship of China’s National Office for Teaching Chinese as a

Foreign Language (abbreviated Chinese name: Hanban). China has recognised the

importance of promoting Chinese language and culture to the international community, and

began to establish non-profit institutions in 2004. These institutions were given the name

‘Confucius Institutes’ or ‘Confucius Classrooms’. They aim to promote Chinese language and

culture in foreign countries. Since 2005, nine Confucius Institutes and 11 Confucius

Classrooms have been established in Australia, offering diverse Chinese programs at various

levels outside the regular education systems.

Chinese is now one of the major languages taught, both in the education system and in

other Australian settings. Interestingly, the majority of Chinese learners in Australia are

Chinese Australians. At senior secondary level, an overwhelming proportion (90%) of

Chinese learners are Chinese Australian students (Orton, 2008). In higher education, the

positive gains in Chinese language learners’ enrolments largely reflect the growing number of

young Chinese Australian adults (McLaren, 2011). In these contexts, Chinese language is

taught and learnt as a HL rather than a Foreign Language. This raises the question as to how

28

well Chinese Foreign Language programs match the needs of CHLLs. Teaching and learning

Chinese as a Foreign Language is becoming a topic of interest in Australia. However,

teaching and learning Chinese as a HL has not been effectively addressed in relation to its

unique CHLL population (McGinnis, 2008). Though Chinese language is the most common

HL with a growing number of CHLLs in Australia, CHL has received little attention in terms

of theory-building thus far (Wang, 2007). Very little is known about the nature and the

dynamics of CHL learning as well as the rate and route of CHL maintenance and

development (He, 2006). The study reported in this thesis seeks to fill the gaps in the

literature and to address the associated practical issues.

1.3 Researcher’s subjectivity

Taking account of the weight of the quantitative component in this mixed methods study, ‘the

researcher’, instead of ‘I’, was used throughout the thesis in line with the convention of

reporting quantitative research. However, equally mindful of qualitative research conventions,

it is also worth addressing the researcher’s subjectivity at this point. As a Chinese citizen and

a native Mandarin speaker, the researcher came to Australia to pursue his doctoral degree.

Recognising the transformation of China into the economic powerhouse of the global

economy and the accelerating pace of this transformation, the researcher was interested in the

cultural and social impact of this transformation on Australia, a country that has identified

China as a strategic partner in the Asia-Pacific region. He is convinced of the role that

Chinese language can play in strengthening this partnership to mutual benefit. There is an

additional consideration with regard to researching the Chinese Australian population.

Because the researcher is culturally closer and physically more similar to Chinese Australians

than any white scholar would be, he may achieve better entrée to and rapport with the

29

Chinese community in Australia and access to data that may not be available to white

scholars. However, this relatively easy access and affinity with the research participants do

not necessarily lead to biased interpretation of the data or taking the data for granted. As an

international student from China, the researcher was relatively confined to the university life

and thus removed from the more complex lived worlds of Chinese Australians. Such social

difference offered the researcher some distance from the participants’ experiences as an

outsider. This combination of ‘insider’ and ‘outsider’ status will be revisited in Chapter Four.

Beyond these contextual and personal reasons, the researcher investigates how Chinese

Australians negotiate their identities and capitalise on various resources through their CHL

learning.

1.4 Significance of the research

This study has theoretical, methodological, and practical significance. Theoretically, HLLs’

ethnic identity and their commitment to HL learning have received attention from different

perspectives, including but not limited to, classical social psychological scholarship and

poststructuralist scholarship. Both schools offer meaningful perspectives in certain contexts,

but receive critiques from other contexts, as will be discussed in Chapter Two. Relevant

issues associated with CHLLs seemed to have been overlooked in the literature. These issues

are particularly salient for Australia, a complex social place with an idiosyncratic historical

and cultural background. The current study makes a theoretical breakthrough to examine

Chinese Australians’ commitment to CHL learning and their ethnic identity in CHL learning

through Bourdieu’s sociological theory to reconcile the oppositions between classical social

psychological and poststructuralist perspectives. This will be discussed in detail in Chapters

Two and Three. Methodologically, empirical studies using Bourdieu’s sociological theory to

30

address language learning-related problems have often adopted qualitative research methods.

This mixed methods study combines a quantitative approach and a qualitative approach to

add depth and scope, which resonates with Bourdieu’s own methodological thinking.

Particularly, the study makes a methodological breakthrough: A set of instruments was

developed and validated to quantify Bourdieu’s key concepts of capital and habitus within

certain social fields, as will be discussed in Chapter Four. Practically, the benefits of HL, for

both individuals and society, have been well documented in recent studies (C. Baker, 2003;

Cho, 2000; J. S. Lee, 2002; McGinnis, 2005; Shin, 2010; Tse, 2001) and current language

policies. The current study attempts to understand Chinese Australians’ personal, social, and

cultural engagement in CHL learning, which may make Chinese Australians better

understand themselves within the social world that they live in and make them better

understood by the Australian society; and further promote enhanced intercultural relations,

adaptable thinking, and mutual understanding within multicultural Australia. As such, this

study resonates with the political discourse in the current language and cultural policy in

Australia.

1.5 Thesis structure

Chapter One has introduced the topic of the thesis and the research questions of the study.

Definitions of HL and HLLs have been provided. Background information related to Chinese

immigrants and Chinese language within multicultural Australia has been introduced. This

chapter also poses the researcher’s subjectivity and highlights the significance of the study.

Chapter Two provides a review of relevant literature and identifies the gaps in the literature.

Based upon the literature review, a theoretical framework is developed in Chapter Three.

Bourdieu’s signature notions of habitus, capital, and field are examined in turn to help

31

construct the theoretical framework. The methodology for this research project is presented in

Chapter Four. The combination of quantitative and qualitative approaches within the study is

explicated. The rationale for the research design is explained in detail. Chapters Five and Six

provide data analysis of the initial quantitative phase and the subsequent qualitative phase

respectively. Main findings addressing the research questions are reported. The thesis

concludes in Chapter Seven by discussing important issues emerging from the data analysis,

outlining the unique contributions of the study, addressing limitations of the research,

highlighting future directions for research, and providing an overall conclusion to the thesis.

32

33

Chapter 2: Literature Review

As articulated in Chapter One, this study addresses two research questions:

RQ1. Is CHL proficiency of young Chinese Australian adults influenced by their

access to various resources, the strength of their ‘Chineseness’, or both?

RQ2. What does CHL mean to young Chinese Australian adults?

In relation to these two questions, the literature review provided in the present chapter

straddles different bodies of scholarship on commitment to CHL learning and identity

construction in CHL learning. Section 2.1 unpacks the notions of ‘motivation’ and

‘investment’ from a social psychological perspective and a poststructuralist perspective

respectively. This section also evaluates how empirical studies make sense of these two

notions in CHL learning contexts. Section 2.2 reviews social psychological and

poststructuralist understandings of identity, with particular regard to CHLLs’ ethnic identity

construction through their CHL learning contexts. Section 2.3 explicates a Bourdieusian

stance and argues the salience of Bourdieu’s notions of ‘capital’ and ‘habitus’ in the current

study. Section 2.4 summarises the literature review, identifies the gaps in the previous

research, and establishes a conceptual direction for this study.

2.1 Commitment to CHL learning: ‘Motivation’ and ‘investment’

The benefits of learning and retaining HLs, both for individuals and for society, have been

well documented (C. Baker, 2003; Cho, 2000; McGinnis, 2005; Tse, 2001), but in

English-dominant countries, loss of HL occurs with the shift from HL to English (C. Baker,

2003; Fishman, 1991; Tse, 2001). In these contexts, CHLLs’ commitment to CHL learning is

34

apt to vary considerably (D. Li & Duff, 2008). CHLLs may study Chinese because their

parents push them to; or maybe it is easier for them to learn Chinese than other foreign

languages; or perhaps they want to communicate within the Chinese-speaking community

(Xing, 2006). In the past two decades, the rapid economic growth in China has extended

these commitments to include personal benefit and future opportunities of employment

(Tsung & Cruickshank, 2011). The following sections present an extensive review of

empirical research, addressing various reasons behind CHLLs’ different levels and forms of

commitment to CHL learning in different situations.

2.1.1 Motivation: A social psychological perspective

Motivation is the activation of goal-oriented commitment promoted by individuals’ specific

reasons and orientations for doing things. Motivation deals with explanations of why people

do the things they do, how long they are willing to sustain the activity, and how hard they

intend to pursue it (Dörnyei & Skehan, 2003). In language learning contexts, motivation is

defined as the directed exertion by which the learner works or strives to learn the Target

Language (TL) due to a desire to do so and the feeling of accomplishment and satisfaction

experienced in the activity (Gardner, 1985). This definition indicates that motivation is the

combination of desire and effort to learn the language and favourable attitudes towards

learning the language.

Social psychological approaches to motivation seek to explain the individual

characteristics that affect language learning, and sometimes how social context influences

these characteristics. The pioneering studies on motivation with classical social psychological

perspectives in Second Language learning by Gardner and Lambert (1959, 1972) identify two

major types of motivation, namely integrative motivation and instrumental motivation. By

35

their account, integrative motivation is characterised by the learner’s positive attitudes

towards the TL group and the desire to learn the TL that helps the learners to integrate into

the TL community. Integrative motivation includes the desire to learn a language in order to

communicate with people who speak that language, the desire to identify closely with the TL

group, and the desire for cultural understanding of the TL community. The concept of

integrative motivation implies that successful language learning depends on a willingness or

desire to be like valued members of the TL (Gardner, 1968). The access of language learners

to the TL community is a function of learners’ integrative motivation (Norton, 1995). In

contrast, instrumental motivation, by Gardner and Lambert’s (1959, 1972) account, refers to

the desire to fulfil certain pragmatic and utilitarian goals to gain some social or economic

rewards through the TL achievement. It refers to a more functional reason for language

learning, such as getting a job or passing an examination. Some cross-sectional studies on

CHLLs’ motivation have yielded mixed results about learners’ reasons for affiliating with

CHL. They are reported as follows.

Guided by Gardner and Lambert’s (1959, 1972) motivation theory, Wen (1997)

investigated the motivation for learning Chinese by US university students from Asian and

Asian-American backgrounds. Seventy-seven students at the beginning and intermediate

levels of Chinese proficiency participated in the study. Ninety-two percent of the participants

had some level of Chinese due to their Chinese heritage. These participants were regarded as

CHLLs. Data were collected by a questionnaire.

The results of multiple regression analysis showed that integrative interest in Chinese

culture and the desire to understand one’s own cultural heritage were the initial motivation

for students to start learning Chinese. This suggested that integrative motivation played an

important role in the preliminary stage of Chinese learning. Students’ presumption that

36

Chinese courses were less demanding than other courses also motivated them to choose

Chinese. This suggested that beginning students were also instrumentally motivated in

learning Chinese to fulfil course requirements. In addition, the results demonstrated that

instrumental motivation contributed to efforts by intermediate learners. Those who were

motivated to exert effort in learning Chinese for certain practical purposes tended to think

that learning outcomes and performance would lead to certain meaningful results or valued

instrumentality.

Similarly, Yang (2003) followed Gardner and Lambert’s (1972) social psychological

motivation model and surveyed 341 college students enrolled in East Asian languages classes,

looking at their language learning motivational orientations. These students were Chinese,

Japanese, and Korean language learners from both heritage and non-heritage backgrounds.

Although some motivational orientations of East Asian language learners were common to

broader Second Language Acquisition or Foreign Language learning situations, integrative

motivation was shown to be more important than instrumental motivation across ethnic

backgrounds. Heritage was the most important variable affecting their integrative

motivational orientations, especially for Korean and Chinese learners. Of the Chinese

learners, 41% claimed Chinese ethnic background. This study implied that Chinese heritage

might function as an integrative motivation in CHL learning.

Yang’s (2003) findings seem inconsistent with Lu and Li’s (2008) study, which

contended that the 120 CHLLs examined were more influenced by instrumental motivation to

pursue Chinese than NHLLs. Interestingly, Wen’s (2011) investigation involving 176 CHLLs

and 141 NHLLs from three US universities observed no significant group differences in

terms of instrumental motivation. The study indicated that all learners valued the usefulness

37

of Chinese language proficiency and studied the language for future opportunities. Although

all learners endorsed instrumentality, instrumental motivation was a significant predictor for

continuing Chinese studies with CHLLs only, suggesting that the decision to continue their

CHL was closely related to perceived usefulness of the language career-wise, and the

perceived importance of the language in the current global economy.

Li (2005) conducted a survey of 695 students at a western Canadian university.

Ninety-two percent of the students were CHLLs. The author did follow-up interviews with 20

CHLLs, five each from Chinese Mainland, Taiwan, and Hong Kong, as well as five other

ethnic Chinese from other locations. The study revealed that most CHLLs were driven by

both integrative and instrumental orientations. No one studied Chinese for purely academic

reasons. They were motivated to learn more about themselves and their ethnic cultures. At the

same time, they hoped to increase their future career opportunities in relation to the growing

Chinese economy. The study also found that CHLLs’ commitment to CHL learning was

shifting. Unlike their childhood CHL learning experiences, which were reportedly

“annoying”, their university CHL learning experiences were reportedly “interesting”,

“rewarding”, and “important”.

Although Gardner’s integrative versus instrumental orientation approach is prevalent in

motivation studies, it is not the only social psychological approach for motivation research.

Comanaru and Noels (2009) examined the motivation of CHLLs in light of

self-determination theory, which contends that learning commitment results from intrinsic

motivation and some forms of extrinsic motivation. One hundred and forty-five participants

were university-level students enrolled in credit-bearing Chinese courses. They differed in

their degrees of exposure to Chinese. They were divided into three groups: (1) Chinese as

38

First Language; (2) English as First Language with at least one Chinese-speaking parent; and

(3) English as First Language with no Chinese-speaking parent. The first two groups were

treated as CHLLs.

Through a survey, the study examined whether CHLLs with different degrees of

exposure to CHL differed in terms of their motivation. The study also investigated the

implications of intrinsic and extrinsic orientations for commitment to CHL learning. The

more CHLLs felt that learning Chinese was personally meaningful and fun, the more they

engaged in the learning process. There were no motivational differences between the first two

groups of CHLLs but they were more motivated than the third group because they felt much

more strongly that Chinese was a central part of themselves, they felt more pressure from

others to learn Chinese, or they had a stronger self-imposed feeling that they ought to learn

the language.

An open-ended question asked participants to describe their motivation for learning

Chinese. Some participants only claimed intrinsic motivation but most participants cited

extrinsic motivation. Several participants indicated multiple considerations for learning

Chinese, showing a mix of intrinsic and extrinsic orientations.

In summary, traditional dichotomies of motivation become blurred and blended among

CHLLs (D. Li & Duff, 2008). Empirical studies have acknowledged the challenges of

distinguishing types of motivation in CHL contexts. Empirical studies have also revealed that

CHLLs’ motivation in CHL learning is not fixed but shifts over time. These findings pointed

out the limitation of Gardner’s and other analogous motivation theories in CHL contexts. The

theories used in the studies attempted to investigate the link between motivation and language

learning within cognitive, individualistic, and reductionist frameworks. This assumes that

motivation is no different than a static physical entity that individual learners either have or

39

do not have, or either have or do not have enough of; and that the more disposed and

motivated learners are to master a language, the more successful they will be in doing so.

These views put the full blame for ineffective language learning outcomes on the learners, as

they are believed to have failed to sustain a necessary level of commitment and drive in the

language learning process. These theories overlook the fact that motivation and social context

are indivisible and that motivation is anchored in individual learners’ reciprocal relationships

and collective practices (Celik, 2007).

Unlike the classical social psychological approaches to motivation, poststructuralist

theorists view commitment to language learning as a co-construction with discursive and

social structures that cannot be simply and easily compartmentalised into one type or another

originating uniquely from or residing within the individual learner (D. Li & Duff, 2008). The

following section will review studies regarding commitment to CHL learning from a

poststructuralist perspective.

2.1.2 Investment: A poststructuralist perspective

‘Poststructuralism’ refers to an intellectual movement that entails new explorations of

differences between cultures, societies, institutions, and individuals (Holmes et al., 2007). It

is an approach to research that embraces seeming contradictions and questions fixed

categories or structures, oppositional binaries, closed systems, and stable truth (Pavlenko,

2002). Block (2007) briefly explains ‘structuralism’ versus ‘poststructuralism’. By his

account, structuralism is defined as the search for universal and invariant laws of humanity

that are operative at all level of human life; in contrast, poststructuralism, the surpassing of

structuralism, is about moving beyond the search for universal and invariant laws of

humanity to more nuanced, multileveled, and ultimately, complicated framings of the world

40

around us. It investigates how social relations between individuals, groups, and communities

impact on the life chances of individuals at a given time and place (Norton, 2000).

In light of feminist poststructuralism, Norton (1995) argues that the classical social

psychological concept of motivation does not pay attention to the complex and dynamic

relationship between the learner and the social world. She criticises this conceptualisation for

viewing learners as static individuals who are inadequately labelled as a constellation of

binaries, including ‘motivated’ versus ‘unmotivated’. In an attempt to critically conceptualise

the relationship between the language learner and the social world and to succinctly identify

the complex identities of language learners, Norton (1995) challenges the static

conceptualisation of motivation in Gardner and Lambert’s (1959, 1972) model and proposes

the concept of ‘investment’ to recast the concept of ‘motivation’.

‘Investment’, it should be noted, was used earlier by Bourdieu (1986, 1989a) to denote

a form of commitment that will produce returns in the form of material and symbolic profits.

Later, Norton (2000) used ‘investment’ to denote learners’ commitment to learning a

language in order to achieve some imagined future self. By her account, language learners

will expect some return on that investment, which is commensurate with the effort expended

on language learning. The metaphor of investment is economic in origin, but differs from the

economic understanding of investment. The economic approach to investment is underpinned

by the rationality of individuals, implying that investment is based on individual decision

making (Coyle, 2002). In contrast, Norton’s feminist poststructuralist approach to investment

emphasises the complexity of individuals. As a result, language learners’ “investment in the

target language may be complex, contradictory, and in a state of flux” (Norton, 2000, p. 11).

The concept of investment in this context was expanded to include material and symbolic

resources rather than being restricted to economic resources. For this reason, learning a

language is investing with the hope of gaining access to a wider range of symbolic resources,

41

such as education, occupation, and friendship, and material resources, such as real estate and

money (Norton, 2000).

Investment is also viewed as closely connected to the ongoing production of language

learners’ multiple and contradictory social identities (Ellis, 1997; Norton, 1995). Investment

is socially constructed in relations of power, changing over time and space, and possibly

coexisting in contradictory ways within an individual language learner. The concept

presupposes that when language learners speak, they are not only exchanging information but

also constantly restructuring their social identities and negotiating the tensions between these

identities and the social world (Norton, 1995).

A distinction should be made between ‘instrumental motivation’ and ‘investment’.

Instrumental motivation is derived from the desire to acquire a tangible reward for learning

the TL. The notion of instrumental motivation presupposes a unitary and fixed language

learner who desires access to material resources that are the privilege of the TL speaker

(Norton, 1994). In contrast, the desire to acquire returns on the investment is more

complicated than a simple tangible reward. The concept of investment necessarily

complicates the ways in which instrumental motivation has been traditionally understood as a

fixed personality trait. The concept conceives of language learners as having a complex

identity, which is multiple, contradictory, and changeable. The notion recognises that learners’

imagined future uses of TL affect their choices of engagement in the language learning

process. The notion also recognises that the relationship between the learner and other

speakers and the relationship between the learner and the social world play significant roles

in the language learning process.

As with poststructuralist perspectives, recasting ‘motivation’ as ‘investment’ offers a

theoretical instrument to examine how individual commitments to TL learning are shaped and

reshaped in and by particular social contexts and how individual learners are rewarded

42

through their commitment to TL learning at a given time and place (Pavlenko, 2002). Some

existing studies illuminate how CHLLs view their investment in CHL learning as access to a

wider range of resources.

Weger-Guntharp (2006) conducted a study in the Chinese language classroom at a

private American university in a major east coast city. Of the 25 undergraduate participants,

eight students were identified as CHLLs. Their ages ranged from 18 to 22. Drawing insights

from Norton’s (1995) concept of investment, this study investigated CHLLs’ commitment to

CHL learning in their first semester of Chinese classes. Biographical profiles of the

participants, an online questionnaire, and an interview consisting of open-ended questions

were used to collect data.

The results of the online questionnaire demonstrated that CHLLs’ commitment to CHL

learning varied, including the intended use of Chinese for future work, attaining advanced

levels of Chinese, and understanding the cultural heritage, indicating that investment in CHL

helped them gain access to a wider range of (potential) resources. The interview results

indicated that investment was changing and shifting across time. In many cases, participants

felt learning Chinese was not “cool” (Weger-Guntharp, 2006, p. 37) because their parents

made them do so when they were in middle school. However, when they were learning

Chinese at the university, they were fond of learning it instead of being pushed to do so by

their parents.

Wong and Xiao (2010) interviewed 64 CHLLs (30 males and 34 females) in US

universities. Thirty-seven were born in the US and 27 migrated with their parents at a young

age. Forty-five spoke a Chinese dialect or dialects as their HL, and 19 spoke Mandarin as

their HL. Norton’s (1995) concept of investment informed the data analysis. Many of the

43

students admitted that learning Chinese was once an unpleasant activity forced on them by

their parents. Later, however, they considered Chinese learning a wise and worthwhile

investment. They considered Chinese to be a prominent currency in the world economy and

contended that Chinese proficiency would enhance their job prospects and favourably

position them in global markets. All the students reported that Chinese learning would give

them a competitive edge in their careers and help them realise their goals. In addition, most

students viewed Chinese learning as a means to foster their connection to Chinese-speaking

networks. Their investment was expected to pay off financially and socially. The findings

confirmed Norton’s (2000) contention that language learning enables learners to access a

wider range of resources as a return on their investment in language learning.

The findings of the reviewed studies indicate that CHLLs’ commitment to CHL

learning can be considered as an investment. Learners were expecting some added value from

their investment in CHL learning. The findings supported Norton’s (1995) argument that

investment in language learning is a means of acquiring greater access to symbolic and

material resources. That is to say, CHLLs’ investment in language made more resources

available and accessible for them. The findings of some empirical studies (Francis, Archer, &

Mau, 2009; A. Hancock, 2006; Lao, 2004; Weger-Guntharp, 2006; J. Zhang, 2009) also

indicated that an investment in CHL learning was an investment in CHLLs’ identities, ethnic

identity in particular. At the same time, learners’ commitments to language learning were

shaped by the range of identities available to the language learners (Pavlenko, 2002). The

mutual relationship between CHL learning and CHLLs’ identity construction will be

discussed in more detail in the next section.

44

2.2 Identity construction in CHL learning

Biologically, individuals are distinguished by their genetic or physical characteristics,

including but not limited to skin colour, gender, facial features, and body size. Socially,

individuals are produced in conditions that determine individual membership in categories

based on class, religion, culture, education, family, and peer group amongst other things.

These biological and social areas are what ‘identity’ is popularly understood to include.

Historically, the identity-language link has attracted a great deal of attention. Johann

Gottfried von Herder (1744-1803), the German philosopher, observed that even the smallest

nations cherish the great deeds of their forefathers in and through their languages, and

concluded that language is the collective treasure of group feeling (Barnard, 1969). Similarly,

Wilhelm von Humboldt (1767-1835), the Prussian philosopher, insisted that language is the

“spiritual exhalation” of a nation (Cowan, 1963, p. 277).

Following these historical assumptions of a necessary language-identity link, the

mutually constitutive relationship between language and identity construction has been

widely recognised in recent literature. Specifically, studies of CHLLs (Chao, 1997; He, 2008)

have demonstrated that learners may study Chinese to search for their ethnic identities. CHL

has an innate relationship with CHLLs’ ethnic identity (Lin, 2004). This immanent

relationship has been examined in CHL contexts from different theoretical perspectives.

2.2.1 Identity: A social psychological perspective

The notion of ‘identity’ has been the object of extensive scholarly treatment in the social

psychological literature in recent decades. Some definitions of identity are based on the

commonalities belonging to the in-group as well as the uniqueness that distinguishes them

from the out-group. For example, identity is defined as a subjective feeling of sameness and

45

continuity that provides individuals with a sense of self and serves as a guide to choices in

key areas of one’s life (Erikson, 1968). Identity is also defined as part of an individual’s

self-concept that derives from knowledge about the individual’s membership of a social

group (or groups) together with the value and emotional significance attached to that

membership (Tajfel, 1981). These definitions imply that identity is a distinctive character

belonging to a given individual, or shared by all members of a particular group (Rummens,

2003). The character shared by the in-group and distinguished from the out-group determines

the recognition of, and affiliation to, a membership. In these ways, identity marks the ways in

which individuals are the same as others or in which the in-group are different from the

out-group. These understandings are attuned to the psychological notion of ‘self-identity’ that

concerns individuals with the state of being a unique person distinct from all others as

reflexively understood by that individual through time (Rummens, 2003). Individuals are

shaped and formed, in a relatively fixed worldview, by sharing common or similar beliefs and

opinions, common or similar emotional attitudes, as well as common or similar behavioural

dispositions.

The social psychological understanding of ethnic identity makes sense when

investigating ethnic groups’ HL loss, maintenance, or development. Ethnic groups must

collectively decide whether to maintain their HL or to let it erode (Giles & Johnson, 1987). In

other words, HL erosion, maintenance, or development depends, to a certain extent, on

decisions being made and strategies being enacted. Following this traditional approach, some

studies investigated the HL and ethnic identity link from a social psychological perspective. A

few empirical studies addressed Chinese immigrants and their descendants in terms of the

relationship between their ethnic identity and their CHL learning. These studies are reported

below.

46

The primary goal of Kiang’s (2008) study was to examine individual differences in

ethnic label choice among young American adults from Chinese backgrounds. Self-reported

ethnic labels were examined among 242 young American adults with Chinese ancestry (age

range = 18-32 years, M = 23.97; 73% female, 27% male). Approximately 30% were

first-generation immigrants, 64% were second-generation or born in the US, 5% were

third-generation, and 1% were fourth-generation.

Associations between ethnic labels and generation, the ethnic diversity of one’s

community and peer group, ethnic identity, and language proficiency were considered by

using a quantitative approach. Internet technology was used to recruit participants and to

administer self-report online questionnaires. Ethnic labels fell under broad categories

whereby 22% reported heritage national labels (e.g., Chinese), 35% added ‘American’ to

their heritage national label (e.g., Chinese American), and 42% reported panethnic-American

labels (e.g., Asian American). Findings indicated that ethnic identity was linked to greater

CHL proficiency and more positive relationships with parents and Asian peers.

Oh and Fuligni (2010) conducted a cross-ethnicity study. This study focused on the

influence of HL proficiency and use on the social development of adolescents from

immigrant backgrounds. The influence of HL proficiency on adolescents’ ethnic identity was

also examined. The analyses described in the study focused on the 414 participants with

either Latin American or Asian backgrounds (mean age = 14.9 years; 49% male, 50% female,

1% did not report gender) who were either immigrants themselves or who had immigrant

parents. Of the 229 Asian American participants, the majority (67%) were of Chinese

descent.

Correlation analysis revealed significant positive associations between the ethnic

identity subscales and HL proficiency. Levels of ethnic identity also varied with language use.

47

HL proficiency seemed to be more important than language use patterns for family

relationships and ethnic identity. It was HL proficiency, not language use patterns, that was

associated with the quality of parent-adolescent relationships and the strength of ethnic

identity. In the case of family relationships, HL proficiency was found to be associated with

the quality of adolescents’ relationships with their parents, but these relationships did not

vary with language use patterns. As for ethnic identity, though HL proficiency and language

use patterns seemed separately associated with the strength of adolescents’ identification with

their ethnic group, when taken together, it appeared that HL proficiency was the stronger

predictor of ethnic identity.

Empirical findings of a few other studies also indicated the role played by ethnic

identity in the process of CHL learning. The sense of relatedness to Chinese family and

community was found to be the most consistent predictor of a self-determined orientation to

CHL learning (Comanaru & Noels, 2009). Examination of CHLLs’ preferences for simplified

or traditional scripts and their attitudes toward different dialects demonstrated connections

between CHLLs’ evolving identities and reasons for the preference of script and dialect (D.

Li, 2005). It indicated the relationship between learning a CHL dialect and sub-cultural

Chinese identity.

In summary, ethnic identity has been commonly understood in social psychological

treatments as one’s alignment or affiliation with, or membership in a particular ethnic group.

It is also a sense of emotional ties that a person has with the group and the meanings of the

ties to the person. The existing literature about CHLLs’ identity focuses heavily on ethnic

identity. Key areas of investigation include CHLLs’ self-labelling, self-identification, and

sense of attachment and belonging (Kiang, 2008; Rosenthal & Feldman, 1992). Some studies

48

have addressed the positive relationship between ethnic identity and CHL proficiency or

usage (Comanaru & Noels, 2009; Feuerverger, 1991; Kiang, 2008; D. Li, 2005; Oh & Fuligni,

2010; Rosenthal & Feldman, 1992). As such, the emphasis is clearly on how individuals

identify themselves, as well as on how related factors such as place of birth, generation,

gender, age, socialisation, and CHL proficiency, help to inform such self-identifications.

Much of this work falls into the social psychological realm (Rummens, 2003).

2.2.2 Identity: A poststructuralist perspective

In contrast to the social psychological perspective, poststructuralism conceptualises identity

as the conscious and unconscious thoughts and emotions of individuals, their sense of

themselves and others, and their ways of understanding their relation to the world (Norton,

1995, 2000). Identity is the unpredictable outcome of a combination of diverse systems in

which people come to terms with who they are in relation to others around them, and is

strongly influenced by the way they view the past, present, and future (Norton, 2000). These

definitions of identity take account of the relationships amongst individuals, others, and the

social world. Firstly, identity explains how individuals perform, interpret, and project their

sense of themselves by the body they look like, by the language they speak, and by the social

behaviour they adopt (Block, 2007). Secondly, identity offers an idea of how individuals

understand others around them and how they relate to others (Woodward, 1997). Thirdly,

identity examines how a person fits in or belongs to the social world and links our social

positions with social situations (Woodward, 2003). Consequently, identity gives individuals a

location in the world and a link to the society that they belong to (Weerakkody, 2010). It also

makes sense of individuals’ relations to the social, cultural, economic, and political changes

taking place in the social world (Woodward, 2003). These are poststructuralist understandings

49

of identity in that they are concerned with how individuals are formed as subjects, how

individuals adopt their subject position, and how individuals experience social variations

across time and space (Holmes et al., 2007). The marking of commonality and difference is

achieved not only symbolically through representational systems but also socially through

inclusion and exclusion of certain groups of people (Woodward, 1997). In what follows, three

defining characteristics of identity of the poststructuralist perspective are discussed.

Firstly, identity is multiple (Norton, 1995). It is produced by, and at the same time

produces, the meaning-making practices of a variety of social sites, such as home, workplace,

school, and community, in which individuals may take up different ‘subject positions’. Some

positions may be different from others. Although a person may be positioned in a particular

way within a given site, the person might not be considered a passive subject to take the

given or marginalised position. Rather, the person might take up new and powerful positions

that do not displace the previous ones (Lo-Philip, 2010). Therefore, individuals have multiple

identities.

Secondly, identity is a site of struggle (Norton, 1995). This conception of identity as a

site of struggle is an extension of the position that identity is multiple. Poststructuralism

insists that the individual is always at the site of conflict and disunity, embracing quite

contradictory modes of identity at different moments (Weedon, 1997). Given identity

construction in a particular system of meanings and values, individuals may find themselves

resisting alternatives. However, when they move out of familiar circles, through education or

politics, they may be exposed to alternative ways of constituting the meaning of their

experience. These situations may well be contradictory, leading to struggle. For example,

individuals sometimes live across social, geographical, and linguistic border crossings.

Living at the intersections of different cultures with different discourses and expectations,

they must negotiate group boundaries when travelling between, or dealing with, different

50

communities (Root, 1992). These are the boundaries to mark inclusion or exclusion,

integration or isolation, alliance or alienation. Crossing the borders, they find that their

identity is destabilised and that they enter a period of struggle to reach equilibrium (Block,

2002). The outcome of the struggle is not to add something new to the old. Nor is it a

half-and-half situation where individuals become half of what they were and half of what

they have been. Identity is constructed through the contradiction of differences. It is about

negotiation at the crossroads of the past, present, and the future. The entire process is

conflictive and ambivalent as opposed to harmonious. In this frame, identity is understood as

a site of struggle.

Thirdly, poststructuralism also argues that identity is socially constructed and produced

through language (Norton, 1995). Identity is produced in a whole range of discursive

practices in which individuals “are constantly organising and reorganising a sense of who

they are and how they relate to the social world” (Norton, 2000, p. 11). Language is the place

where actual forms of social organisation are defined, where social meanings are constituted

and produced, where individual signs acquire meanings, and where identity is constructed in

ways that are socially specific (Weedon, 1997). In contrast to social psychological

understandings of identity, these arguments imply that identity is neither innate nor

genetically determined, but is socially emerging, modified, transformed, and differentiated

through the use of language. From this perspective, identity is open to continuous redefinition

and constant slippage. As individuals acquire language, they learn to give voice and meaning

to their experiences and to understand language according to particular ways of thinking. This

language learning process constitutes and structures consciousness and sense of self, that is,

identity. As Pennycook (2003, p. 528) states, “It is not that people use language varieties

because of who they are, but rather that we perform who we are (among other things) using

varieties of language.” That is to say, it is through the use of language that identity is

51

expressed, enacted, and symbolised. In line with these poststructuralist perspectives, some

empirical studies have examined CHLLs’ identity construction in CHL learning contexts.

These studies are reported below.

Chao (1997) reported the challenges that Chinese Americans had during their CHL

learning. This study is not about Chinese Americans’ identity constructed in their CHL

learning per se. However, the most noteworthy finding emerging from the data was the

change in cultural identification. When younger, the participants did not appreciate the

importance of their Chinese culture. During their teenage years, they reportedly had a strong

desire to be independent, to be integrated into American culture, and to be accepted as part of

American society. Thus, their priority was to seek identity in the English-speaking

community. In contrast, their Chinese identity was suppressed and CHL, in turn, became an

unnecessary burden that required extra effort to learn. After many years of blending into

American society, the participants gradually realised that they could no longer deny a part of

them that was embedded in CHL and Chinese heritage culture. They then looked at the

opportunity to study Chinese on their own terms. The study reflected the multiple,

contradictory, and shifting identities of Chinese Americans. Their “ethno-racial identity” can

be the impetus towards learning their “native tongue” (Chao, 1997, p. 8).

Shin’s (2010) study examined HL experience and identity of adults with mixed

heritages through in-depth and semi-structured autobiographical interviews. Each participant

had an English-speaking white American parent and an HL (Chinese, Japanese, Korean,

Spanish or Vietnamese)-speaking immigrant parent. The findings suggested that HL

proficiency varied widely. Participants reportedly had different degrees of commitment to HL

learning in order to connect with part of their heritage and to define their identity. Shin’s

52

findings indicated that HL was a considerable facet of mixed-heritage identity. Resonating

with a poststructuralist perspective, the study concluded that mixed-heritage identity was

constructed in dynamic, diversified, and open-ended ways, continuously shaped by changing

social contexts.

Shin’s (2010) findings also addressed some participants with Chinese heritage. One

participant said she looked neither Chinese nor American and she did not really know about

herself. This indicated an ambivalent identity. Though parental support for CHL ranged from

“very active support” to “no support”, participants reported a positive attitude towards CHL

learning. They all pointed out the positive relation between CHL learning and Chinese ethnic

identity.

Wong and Xiao (2010) interviewed 64 CHLLs in US universities. All the students

interviewed reported a contingent side of their ethnic selves. They could be Chinese,

Chinese-American, or somewhere in between, depending on the time, location, and context.

Their findings indicated that CHLLs’ identities are flexible formations, variously possessed,

produced, and practiced through CHL learning process. These findings are consonant with

poststructuralist understandings of identity as changeable with shifts of social space and

passage of social time.

Ang (2001) engages with identity studies in an age of globalisation and diaspora. She

focuses on a transnational condition that spatially and temporally sprawls socio-cultural

formations of people and creates imagined communities with blurred and fluctuating

boundaries sustained by real or symbolic ties to the original homeland. The starting point for

her discussion was the experience of her first trip to China. Though of Chinese descent, she

was born in Indonesia and grew up in the Netherlands, before relocating to Australia as an

53

adult. In China, she was different because she could not speak Chinese. In the West, she was

also different because she looked Chinese. A sense of alienation took hold of her. Due to ‘not

speaking Chinese’, she was positioned as a ‘fake’ Chinese by ‘real’ Chinese. That is, she

failed to legitimise her ‘Chineseness’ because ‘not speaking Chinese’ did not give her a

recognised identity as a ‘real’ Chinese. This was the beginning of her engagement with the

predicaments of ‘Chineseness’ in diaspora. She had the difficulty of constructing the

indeterminacy of ‘Chineseness’ as a signifier for identity.

However, ‘not speaking Chinese’ can cease being a problem for overseas Chinese in

diasporic contexts (Ang, 2001). Being transnational; linking the local and the global; the here

and the there; the past, present, and future, diasporic Chinese have the potential to challenge

static, essentialist, and universal conceptions of ‘Chineseness’. Their complex and flexible

positioning between host countries and China keeps a creative tension between current lived

social place and historical cultural history. This tension fills the space in the bipolar

dichotomy of the present and the past with new forms of culture in the collision of the two

(Ang, 2001), and constructs a syncretic identity of ‘Chineseness’ suitable for the current lived

social world. China, the mythic homeland, need not be the absolute norm for ‘Chineseness’

against which all other Chinese diasporic cultures are measured.

In this mixed-up, interdependent, mobile, and volatile world, clinging to a traditional

notion of ethnic identity is ultimately self-defeating (Ang, 2001). ‘Chineseness’ in a

poststructuralist notion is not fixed and pre-given, but constantly renegotiated and

rearticulated. Diasporic ‘Chineseness’ cannot be envisioned in any unified or homogeneous

way. Rather, it is a diverse, heterogeneous, and ultimately precarious hybridity.

To sum up, poststructuralist approaches acknowledge the multiplicity and

open-endedness of identity processes. They see all individuals as users of multiple linguistic

54

resources and as members of complex layers in multiple communities and societies. Transient

language users are able to move between different and complex contexts. With a

poststructuralist perspective, identity is understood as a cover term for “a range of social

personae, including social statuses, roles, positions, relationships, and institutional and other

relevant community identities one may attempt to claim or assign in the course of social life”

(Ochs, 1993, p. 288). Identity consists of multidimensional, contradictory, and ever-changing

images, descriptions, and evaluations of oneself as active and changing subjects in the eyes of

oneself, others, and the society. Identity is socially constructed rather than born, invented

rather than given, always in a process of change rather than at a standstill. In brief, language

learners’ identities are multiple, dynamic, and contradictory (Norton, 1995; Pavlenko, 2002);

socially produced rather than innate or genetically determined (Weedon, 1997).

2.3 Reframing through a Bourdieusian stance

Both social psychological and poststructuralist schools have made important contributions to

the understanding of language learners’ commitment in certain contexts. Motivation, with its

fixed categories, is understood as an individual entity by classical social psychological

scholarship and is identified as one of the core predictors to language learning outcome. As

such, this perspective puts the full responsibility for unsuccessful learning outcomes on the

learners themselves, without taking enough account of social impacts. In contrast, the

poststructuralist notion of investment accounts for language learners’ commitment in relation

to their lived social world. As a return on this investment, language learners expect to gain

access to a wider range of material and symbolic resources. However, the notion of

investment did not conceptualise the nature of these resources or detail the full learning

trajectories of language learners. In this respect, Bourdieu’s notion of capital can complement

55

the notion of investment because investment demands capital and capital has value or adds

value to realise the return on investment. By definition, ‘capital’ here refers to accumulated

resources and has the potential to produce profits and to reproduce itself in an identical or

expanded form (Bourdieu, 1986).

In relation to understanding identity, social psychological scholarship contends that

ethnic identity is an internal trait of HLLs, while poststructuralist scholarship considers it

socially constructed. On the one hand, some dimensions that are biologically embedded in

ethnic identity cannot be made over and some ethnic perceptions are lasting over time and

across space. On the other hand, dispositions and propensities within a particular ethnic group

may shift over time, being shaped by their lived social world. Bourdieu’s notion of habitus

attempts to think through these two opposing truths. ‘Habitus’ here refers to durable

cognitive structures and a dispositional sense of action that direct people to appropriate

responses to given situations (Bourdieu, 1998). It helps to dissolve a plethora of oppositions

between social psychological and poststructuralist schools. Drawing insights from Bourdieu’s

notion of habitus, identity can be understood as a (un)conscious and dynamic sense of self.

Identity is about how individual dispositions are defined, contained, and enabled by the

bodily features that individuals possess, by the actions that they take, and by the languages

that they speak; and it concerns how individuals internalise various forms of external

elements and produce their sense of self in response to the social structures.

In brief, the social psychological approach focuses on micro-level individual

preferences as the sole explanation of individual language behaviours, while the

poststructuralist approach tries to create macro-level socially constructed factors that shape

individual language choices. This section attempts to reconcile the social psychological

inside-out approach and the poststructuralist outside-in approach. The section develops a

Bourdieusian stance to understand CHLLs’ commitment to CHL learning and their ethnic

56

identity construction through CHL learning. By virtue of Bourdieu’s notions of capital and

habitus, the limitations of both the social psychological and poststructuralist approaches can

be addressed.

2.3.1 ‘Motivation’, ‘investment’, and ‘capital’

In the Second Language context, the concepts of ‘motivation’ and ‘investment’ were

proposed by Gardner and Lambert (1959) and Norton (1995) respectively. The two concepts

of motivation and investment are constructs rooted in significantly different research

frameworks of language learning, and therefore function in different contexts and serve

different research purposes. Gardner and Lambert engaged in quantitative investigations of

motivation as an individual trait from a classical social psychological perspective, whereas

Norton looked qualitatively at language learners’ complex social identities and multiple

desires within the social world and was prompted to reconsider and extend the concept

‘motivation’ from a feminist poststructuralist perspective. As such, Norton’s feminist

poststructuralist alternative, ‘investment’, gives purchase on empirical phenomena that

‘motivation’ cannot.

However, the concept of investment has not gone without criticism over the past

decade. Menard-Warwick (2005) suggested that the construct does not adequately consider

both the structural and agentive forces that shape a learner’s language development.

Moreover, Menard-Warwick faulted the notion of investment for not offering an adequate

investigation into learners’ trajectories. Brief touching on learners’ histories is not enough to

examine the sources of learners’ investment (Menard-Warwick, 2005). In addition, language

learners may face social constraints beyond their control that to some extent restrict what they

could invest, despite the resources and agency that learners bring to their language learning

57

(Menard-Warwick, 2005). To complement the notion of investment, the concept of capital

within certain social spaces has found wide purchase in language education scholarship. With

the complement of the concept ‘capital’, ‘investment’ in language learning fits well with the

economic view of investment as “the process of using resources to produce new capital”

(Case & Fair, 1996, p. 33). In other words, these resources increase the value of capital that

the learners invest in a given social context (Celik, 2007; Norton, 2000). In addition,

Bourdieu’s concept of capital conceptualises various resources into different forms of capital

that language learners can invest and access. According to Bourdieu (1996), language

functions as capital in relation to particular markets, where language learners invest in the

language learning actions to construct social reality and negotiate the very conditions under

which their communicative exchanges take place. The concept ‘capital’ and its various forms

will be developed further in Chapter Three. Although empirical studies did not explicitly use

Bourdieu’s concept of capital in CHL contexts, the capital metaphor could make sense in

these studies. Some cross-sectional studies are reviewed below.

A study identified and analysed factors that may contribute to or impede Chinese

American children’s CHL maintenance (J. Zhang, 2009). Interviews with six children

explored how these CHLLs put symbolic, material, and other resources into the enterprise of

language learning based on a cost-benefit assessment. Only three interviews were reported in

the paper. The three CHLLs were either born in the US or moved to live in the US before the

age of 13.

Zhang’s first case was of a 10-year-old boy Frank, who was eager to speak Chinese, a

language that he was proud of. This boy spent time regularly on CHL and Chinese cultural

activities. His Chinese painting was envied and praised by his classmates. His social studies

teacher often praised him for his bilingual skills. He was excited when recalling these

58

experiences. As a return on this investment in CHL and Chinese cultural activities, he had a

reputation as a Chinese painter and speaker that made him stand out in a good way among his

peers. This reputation could be interpreted as ‘symbolic capital’ in Bourdieu’s term.

Zhang’s second case was of a 14-year-old girl Linda. This girl was confident in

speaking but not reading and writing Mandarin. She was a frequent winner in speech contests

at the Chinese community school. Her sister published an article in the local Chinese

newspaper. She was keen to make faster progress in CHL so that someday she could have an

article published too. She enjoyed attending Chinese community school, as she had friends

there whom she had known for nearly 10 years. The returns on Linda’s investment in CHL

included prizes awarded in Chinese language contests and access to her friends’ network. In

Bourdieu’s terms, these profits of CHL learning could be construed as ‘symbolic capital’,

‘institutionalised cultural capital’, and ‘social capital’.

Zhang’s third case was of a 17-year-old boy George. This boy understood Chinese but

usually spoke English. In contrast to Frank and Linda’s CHL proficiency, his Chinese was

“stuttering and disconnected” (J. Zhang, 2009, p. 203). His parents sent him to China every

other year, each time for approximately two months. His oral Mandarin improved with each

visit. Though he did not enjoy learning in the Chinese community school, he did attend

mathematics classes offered by the school. This helped him in his regular school where he

was labelled a “math genius” (J. Zhang, 2009, p. 204) and assigned the task of tutoring his

classmates. George also liked to associate with his Chinese friends because “it’s easier to get

closer to them and it’s easier to establish friendship” (J. Zhang, 2009, p. 204) than it is with

non-Chinese peers. As a return on his parents’ investment of various resources, George won

the reputation of being a “math genius” and gained more knowledge in math. Moreover, his

CHL also established friendships for him. In Bourdieu’s term, these benefits could be

understood as ‘symbolic capital’, ‘embodied cultural capital’, and ‘social capital’.

59

Francis, Archer, and Mau (2009) conducted an interview study to explore 60

British-Chinese pupils’ discursive constructions of the purposes and benefits of CHL learning

in Chinese community schools. The necessity of learning CHL to facilitate communication

with family members at home and relatives in the motherland was a key theme in the data. In

Bourdieu’s term, CHL proficiency helped tighten family ties as ‘social capital’. Many pupils

also described not being able to speak Chinese as a “disgrace”, “embarrassing”, and a reason

for being “ashamed” (Francis et al., 2009, p. 529). In Bourdieu’s sense, lack of CHL

proficiency could be associated with lack of ‘symbolic capital’. Additionally, some pupils

reported that Chinese learning would facilitate future job opportunities in China. These job

opportunities are convertible into ‘economic capital’ in Bourdieu’s term.

Another study conducted in the US investigated parents’ attitudes toward

Chinese-English bilingual education and CHL usage at home (Lao, 2004). Eighty-six parents

were surveyed. The major reasons for parents’ investment in their children’s CHL included

better career opportunities and effective communication with Chinese-speaking family

members and community. In Bourdieu’s term, these are ‘economic capital’ and ‘social capital’

respectively. The study noted a lack of Chinese books held at home by all the surveyed

families. This is consistent with Xiao’s (2008) study that discovered inadequate Chinese

reading materials in Chinese immigrant homes in the US. The print-poor environment at

home was understood to impede further development of children’s CHL. This indicated that

parents’ investment in ‘objectified cultural capital’ might contribute to children’s CHL

learning.

Luo and Wiseman (2000) examined the familial and peer influences on 250 immigrant

Chinese American children’s CHL maintenance. The results indicated that Chinese-speaking

60

peer influence was the most important factor in Chinese American children’s CHL retention.

Parent-child and grandparent-grandchild cohesion also influenced the children’s CHL

retention. These findings indicated that friendship networks and family environments have

impacts on children’s CHL retention. In Bourdieu’s term, ‘social capital’ can contribute to

CHL proficiency.

The studies reported above did not explicitly use Bourdieu’s concept of capital.

However, Bourdieu’s capital metaphor and typology could shed light on these empirical

studies and give answers to these questions: Why do motivated CHLLs nevertheless fail to

acquire CHL commensurate with their goals? Why does their commitment to CHL learning

differ and vary in that they are sometimes motivated, extroverted, and confident, but

sometimes not? Do they change with different reasons or goals for CHL learning at different

ages, and with different degrees of intensity and effort from one context to the next? What are

the returns on their CHL learning?

2.3.2 ‘Ethnic identity’ and ‘habitus’

In HL literature, HL proficiency, ethnic identity, and the relationship between HL and ethnic

identity have been key foci of investigation (Comanaru & Noels, 2009; Feuerverger, 1989,

1991; Kiang, 2008; D. Li, 2005; Oh & Fuligni, 2010; Rosenthal & Feldman, 1992). The

burgeoning of fruitful cross-disciplinary research examining the relationship between HL and

ethnic identity started in the 1980s (Giles & Johnson, 1987). In the literature, HL is

frequently cited as one of the most important and powerful contributors to ethnic identity

construction. HL is considered to have a key influence on ethnic identity for language

minority individuals (Fishman, 1977; Hurtado & Gurin, 1995; Smolicz, 1981). It has been

61

shown to serve as a symbol of ethnic identity and cultural solidarity to remind language

minority groups about their cultural heritage and transmit group feelings (Giles, Bourhis, &

Taylor, 1977). Other studies (Cho, 2000; Cho et al., 1997) reported that HL development

could be an important part of, and contributor to, identity formation, and could help HLLs

retain a strong sense of identity to their own ethnic group. On the other hand, some studies

have documented that a lack of HL proficiency contributes to descending ethnic identity

because language barriers result in intergenerational conflict as children become frustrated

when they are unable to communicate effectively with family members. When family

relationships weaken and parental influence fades, the older generation is reportedly

hampered in its efforts to transmit ethnic values (Hinton, 1999; Wong-Fillmore, 1991).

Concomitantly, ethnic identity fades away. In short, a consequential link between HL and

ethnic identity has been identified in a variety of studies (Bankston & Zhou, 1995; Fishman,

1989; Imbens-Bailey, 1996).

As argued earlier, many examinations of HL, ethnic identity, and their relationship falls

into the social psychological realm, while a smaller number of examinations of these

problems have been undertaken from a poststructuralist perspective. The latter approach

offers the perspectives that have been absent in social psychological literature: Identities of

HLLs are multiple, shifting, contradictory, and socially constructed. However, the

poststructuralist concept of multiple identities without foundational basis has its limitations

(Luke, 2009). The assumption that human identity is wholly malleable and that the body can

be styled to assume an invented identity runs into problems when faced with the durability of

human beings’ internal schemata (Luke, 2009). The body does remember so that human

beings remain in many ways the products of kinship and blood (Luke, 2009). In line with

Luke’s perspectives, this study introduces the concept of habitus, which will be developed

further in Chapter Three. The notion of habitus underpins people’s dispositions rooted in

62

mind and body, as well as acquired characteristics that are the product of social conditions

and may be totally or partially common to people who have been the product of similar social

conditions (Bourdieu, 2005). As such, Bourdieu’s sociological notion of habitus offers a

useful theoretical lens to examine CHLLs’ ethnic identity.

2.4 Chapter summary

HLLs’ commitment to HL learning and the role played by ethnic identity in HL learning have

been understood differently by different theoretical approaches, varying from the traditional

social psychological approach to the poststructuralist approach. Both approaches offer

meaningful perspectives in certain contexts, but are open to critique in other contexts. The

social psychological camp contends that learners’ motivation and ethnic identity are

structured as static individual traits, whereas the poststructuralist camp considers learners’

investment and ethnic identity multiple, dynamic, and socially constructed. To a certain

extent, these two schools stand in opposition to each other in terms of their different

understanding of HLL’s commitment and ethnic identity. Bourdieu’s sociological notions of

capital and habitus could dissolve a plethora of oppositions between these two camps of

scholarship.

The summary of the literature review is outlined in Figure 2.1. The classical social

psychological notion of motivation considers learners’ commitment to language learning as a

dichotomous individual trait. It puts the full blame for unsuccessful learning on the learners

themselves without taking account of the complex social conditions that structure learners’

language practices. The feminist poststructuralist scholarship recasts ‘motivation’ into

‘investment’, which accounts for the complex relationship between the individual learners

and the social world; however, it does not adequately consider the structural and agentive

63

forces that shape learners’ language development. Since investment demands capital, which

has a capacity to produce new capital or reproduce itself as a return on the investment, the

thesis proposes the use of Bourdieu’s notion of capital to complement the notion of

investment. Likewise, ethnic identity is understood as a set of internal static attributes and

dynamic socially constructed dispositions by the social psychological and poststructuralist

schools respectively. As such, these two schools are largely opposed to each other. Some

physical and biological entities, as dimensions within ethnic identity, cannot be made over or

erased, and do stay with the language learners all through their life. Other dimensions of

ethnic identity, such as cultural awareness and values, are inculcated and cultivated through

language learning, and thus are socially constructed. To reconcile these two camps of

scholarship, the thesis proposes the use of Bourdieu’s notion of habitus to help understand

language learners’ ethnic identity as a set of embodied propensities that have the tendency to

internalise the social structures through language learning.

Figure 2.1. Summary of literature review

Social psychological perspective

on CHL learning

Poststructuralist perspective

on CHL Learning

Bourdieusian perspective

on CHL learning

Motivation

Habitus

Ethnic identity

Ethnic identity

Capital

Investment

Complement

Recast Reconcile

64

In summary, there is a dearth of research investigating HLLs’ commitment to, and their

ethnic identity in, HL learning from a Bourdieusian stance; HL research conducted outside

North America has a relatively small body of literature; and CHL in particular, with its

heterogeneity and complexity in its speakers and their associated social, cultural, and

historical ramifications, has not received its due scholarly attention (He, 2008). In this respect,

there is a demand for sociological investigation into the complexities of CHLLs, not only in

North American but also in other diasporic contexts worldwide where CHL has a relatively

significant population of learners and speakers, such as Australia, a complex lived social

world with an idiosyncratic historical and cultural background. By enquiring into CHL for

Chinese Australians from a Bourdieusian sociological perspective, this study offers a

theoretical and empirical analysis to capture the waxing and waning of Chinese Australians’

CHL learning orientations and stances to fill the gaps in the literature.

65

Chapter 3: Theoretical Framework

The current study draws on Bourdieusian theory to build a framework that underpins the

investigation of the research problems. Capital, habitus, and field are the three main “thinking

tools” (Wacquant, 1989, p. 40) of Bourdieu’s sociological approach. They form an

inter-dependent and co-constructed triad, with none of them primary, dominant, or causal

(Thomson, 2008). They need to be taken together to realise the full value of a Bourdieusian

perspective. Formally, Bourdieu (1989a, p. 101) has offered the equation: “[(habitus)

(capital)] + field = practice”, to summarise the conceptually and empirically essential

relationship amongst capital, habitus, and field, which informs individuals’ practices. The

equation implies that manners of being and thinking, routine behaviours, and patterned

sociocultural activities in which individuals engage (practices) result from the relations

between their dispositions (habitus) and their social resources (capital), within the current

state of play of a particular social arena (field). More specifically, Bourdieu (1991) theorises

that people make language choices as a form of social practice according to the amount of

different resources (capital) as well as the dispositions (habitus) that they have within a given

field. In this chapter, these theoretical insights will be developed to inform this study.

Young Chinese Australian adults may have different levels of CHL proficiency because

they may or may not have learnt or chosen to use their CHL. Their choices form part of their

practices in their social lives. Bourdieu’s three thinking tools offer a conceptual mechanism to

discover the deep-rooted reasons behind the CHL practices of young Chinese Australian

adults. Drawing insights from Bourdieu, this chapter constructs a theoretical framework for

the study to investigate the impact of capital and habitus on CHL proficiency of young

Chinese Australian adults, and the meanings that these Chinese Australians attach to CHL.

66

There are four sections in this chapter. The first three sections (Section 3.1, 3.2, and 3.3)

develop Bourdieu’s key concepts of capital, field, and habitus in turn. In order to relate these

three key concepts to the CHL context and the current study, these sections extend Bourdieu’s

notions of ‘linguistic capital’ and ‘culture’ developed within a single language and culture in

mid-20th

- century France into contexts with multiple languages and cultures in the

21st-century societies. Section 3.4 presents a complete theoretical framework by integrating

the three concepts as a theoretical lens on the research questions.

3.1 Capital

Language learners qua investors differ in terms of the quantity and quality of resources they

possess, and their opportunities for successful language learning vary in proportion to the

resources they possess and invest. It is the learners’ investment of different forms of resources

in the TL learning that will lead to TL proficiency. When learners invest in a TL, they hope to

access and acquire previously unattainable resources as a return on their investment

(Bourdieu, 1991; Norton, 2000). This return on the investment may give them access, not

only to material resources, such as capital goods, real estate, and overall financial success, but

also to symbolic and cultural resources, such as educational qualifications, literature, media,

or friendship. To clarify the concept of ‘investment’, it is useful to examine the concept of

‘capital’ because investment demands capital (Bourdieu, 1990) and the aim of investment is

to add value to one’s capital. According to Bourdieu (1986), capital, “in its objectified or

embodied forms, takes time to accumulate” and has “a potential capacity to produce profits

and to reproduce itself in identical or expanded form” (pp. 241-242). For Bourdieu, capital

refers to all goods, or material and symbolic resources, which present themselves as rare and

worth seeking after in a particular social relation located within a system of exchange. As

67

developed by Bourdieu (1986), capital has four interchangeable and convertible forms:

economic, cultural, social, and symbolic.

3.1.1 Economic capital

‘Economic capital’ refers to forms of financial wealth. It can be “immediately and directly

convertible into money and may be institutionalised in the form of property rights” and other

material objects (Bourdieu, 1986, p. 243). It refers to command over economic resources,

such as cash and assets. Economic capital is at the root of all the other forms of capital

(Bourdieu, 1986) in that other forms of capital are based on economic capital but not simply

reducible to it (Bourdieu, 1986; Dovey, 2010). By this is meant that there is convertibility

between economic and other forms of capital in both directions and at different rates of

exchange in different situations.

3.1.2 Cultural capital

‘Cultural capital’ refers to valued cultural property; legitimate knowledge, behaviour, and

modes of thought; and institutionally recognised capacity that characterise different classes

and groups in relation to specific sets of social forms (Bourdieu, 1986). It is about having the

‘right’ cultural resources and assets. The forms of cultural capital include, but are not limited

to, cultural goods, cultural awareness, language competence, behavioural manner, knowledge,

skills, credentials, and qualifications, acquired through education and upbringing (Bourdieu,

1993a) and accumulated over an extended period (Bourdieu, 1986). These forms can exist in

three states: “the objectified state”, “the embodied state”, and “the institutionalised state”

(Bourdieu, 1986, p. 243).

68

In the objectified state, cultural capital usually appears in the material form of cultural

goods that can be concretely displayed such as books, dictionaries, works of art, and musical

and scientific instruments, which are transferable by legal ownership (Bourdieu, 1986). Use

of objectified cultural capital requires specialised cultural abilities. It is not only ownership of,

but also capacity to choose and consume the cultural goods that really counts (Dovey, 2010).

For Bourdieu (1986), objectified cultural capital is only of value when individuals have a

certain amount of embodied cultural capital to consume the cultural goods. Consequently,

these cultural goods are critical when individuals wield power and obtain profits

proportionate to their mastery of the objectified cultural capital, as well as to the requisite

amount of their embodied cultural capital (Bourdieu, 1986).

In the embodied state, cultural capital refers to cultural inclinations, integrated within

the lasting dispositions of mind and body, existing as behavioural patterns, such as accent,

body language, intonation, facial expression, stance, and attitude, which can engender

authority in social situations and act in a way that shows class or manners. It exists “in the

form of schemata of perception and action, principles of vision and division, and mental

structures” (Bourdieu, 1996, p. 5). It is these embodied dispositions that appreciate and

understand objectified cultural goods (Bourdieu, 1986). The accumulation of embodied

cultural capital demands a process of embodiment and incorporation, and a labour of

inculcation and assimilation; it thus costs time, which must be invested personally by the

investor (Bourdieu, 1986). Hence, embodied cultural capital needs accumulation over time

and cannot be transmitted instantaneously.

In the institutionalised state, cultural capital exists as a connection to certain

institutions (e.g. university and professional associations). It takes an object form, such as

qualifications, certificates, educational degrees, academic titles, and credentials that can be

69

institutionally recognised (Bourdieu, 1986, 1991). Institutionalised cultural capital

symbolises a cultural competence and confers on its holder a conventional, constant, legally

guaranteed value with respect to cultural resources (Bourdieu, 1986).

3.1.3 Social capital

‘Social capital’ is the aggregate of the actual or potential resources that accrue to an

individual or a group by virtue of the possession of a durable network of more or less

institutionalised relationships of mutual acquaintance and recognition, such as family, friends,

clubs, schools, communities, and society (Bourdieu, 1986; Bourdieu & Wacquant, 1992). It is

constituted through “contacts and group memberships, which, through the accumulation of

exchanges, obligations, and shared identities, provide actual or potential support and access

to valued resources” (Bourdieu, 1993a, p. 143). In simple words, social capital is the benefits

that one derives from membership in a group. It is the benefits that one draws from social

cohesion and personal investment in a community and any connections within and between

social networks.

According to Bourdieu, there are three key components constructing different

dimensions of social capital: networks, interactions, and recognised values. Firstly, social

capital is a system of social networks inherent within a community that promotes cooperative

behaviour and serves the specific needs of its members. The size and structure of the network

are related to the volume of social capital contained within it. Secondly, social capital exists

in patterns of social relations, or in processes of interactions between individuals and the

social world. Social capital has value and adds value through these interactions. The strength

of the interactions largely decides the value of social capital. Thirdly, social capital is about

connections to and relationships with less, equally, or more powerful others. The amount of

70

social capital is also associated with the social capital of others with whom individuals are in

relation. The greater the recognised social capital of these others, the greater the social capital

accrued. In brief, the volume of social capital ultimately depends on the size and structure of

the network, the strength of the interactions, and the volume of capital of those to whom the

individual is connected (Bourdieu, 1986).

3.1.4 Symbolic capital

‘Symbolic capital’ is the most elusive form of capital to define. In Bourdieu’s early work, it

denoted the aesthetics of the owner and was a form of cultural capital (Dovey, 2010). In later

writings, symbolic capital appeared to be distinct from any definition of cultural capital as an

individually held resource. Symbolic capital was then defined as follows:

Symbolic capital is any property (any form of capital whether physical, economic,

cultural, or social) when it is perceived by social agents endowed with categories of

perception, which cause them to know it and to recognise it, to give it value. For

example, the concept of honour in Mediterranean societies is a typical form of symbolic

capital which exists only through repute, that is, through the representation that others

have of it to the extent that they share a set of beliefs liable to cause them to perceive

and appreciate certain patterns of conduct as honourable and dishonourable. (Bourdieu,

1998, p. 47)

For Bourdieu, symbolic capital refers to “a reputation for competence and an image of

respectability and honourability” (Bourdieu, 1989a, p. 291). It is known and recognised to be

“more or less synonymous with: standing, good name, honour, fame, prestige, and reputation”

71

(Bourdieu, 1993a, p. 37). Specifically, the manifestations of this capital include accumulated

resources and various key markers of standing, such as prestige, reputation, fame, title,

celebrity, and honour that are founded on the recognition granted by others. This capital

refers to the form assumed by different kinds of capital when they are perceived and

recognised as legitimate (Bourdieu, 1985b, 1989b, 1991, 2000).

Capital, in whatever form, may be convertible into or institutionalised in a title of

nobility (Bourdieu, 1986), existing and acting as symbolic capital (Bourdieu, 2000). As

Bourdieu (1998) further explains:

Symbolic capital is the form taken by any species of capital whenever it is perceived

through categories of perception that are the product of the embodiment of divisions or

of oppositions inscribed in the structure of the distribution of this species of capital. (p.

47)

In this sense, symbolic capital is nothing more than other forms of capital when these are

known and recognised by others (Bourdieu, 1985b, 1989b). Capital, converted into symbolic

capital from whatever form, can be reconvertible back into that capital itself (Bourdieu,

1977b). It is the exhibition of symbolic capital that is one of the mechanisms to make “capital

go to capital” (Bourdieu, 1977b, p. 181). For Bourdieu, symbolic capital acts as a social

mechanism within a system of exchange, specifically, exchange of value and forms among

capital. Symbolic capital can give added value and recognition to all other forms of capital.

As such, symbolic capital is perhaps the most valuable form of accumulation in a society

(Bourdieu, 1977b) because holders of large amounts of symbolic capital are in the position to

impose the scale of values most favourable to their products (Bourdieu, 1989b). Therefore,

72

symbolic capital is the power granted to those who have obtained sufficient recognition to be

in a position to impose recognition (Bourdieu, 1989b). It is subjective and is perceived as

making legitimate demands for recognition (Bourdieu, 1990).

3.1.5 Capitalising on language: Linguistic capital

The concept of capital captures the relevance of the economic metaphor of investment.

Language learners’ investment in a TL demands capital, and in return, these learners hope to

acquire a wider range of resources in the forms of economic, cultural, social, and symbolic

capital through language learning (Norton, 1995). That is to say, the invested capital produces

TL competence and TL competence can reproduce capital when it is valued in a given

situation. Norton’s (1995) interpretation of ‘investment’ resonates with Bourdieu’s (1986)

conceptualisation of ‘capital’, which has the capacity to produce profits and to reproduce

itself in an identical or expanded form. The reproduction of these resources depends on the

value of the profits that this reproduction can expect from its investments of the amount and

the structure of capital in different forms (Bourdieu, 1996). In line with this Bourdieusian

perspective, language competence can be conceptualised as a form of capital because it is

both the outcome of the investment in language learning and the generator of new capital

through the process of language learning. In Bourdieu’s term, it is linguistic capital, and can

be understood as a form of cultural capital (Bourdieu, 1999a), more specifically, as embodied

cultural capital (Bourdieu, 1977a).

Linguistic capital has two dimensions: the mastery of, and relation to, language

(Bourdieu & Passeron, 1990). The mastery of a language can produce linguistic competence

in that language, while the relation to a language can produce legitimate competence in that

language. Linguistic competence is the capacity for “infinite generation of grammatically

73

regular discourse”, which may not be valued either “in the social conditions of its constitution”

or “in the social conditions of its operation” (Bourdieu, 1977a, p. 646). In contrast, legitimate

competence is the capacity to produce judicious and appropriate language, which is used in

strategies invested with all possible functions rather than only communicative functions. To

clarify, what is often rare in practice is not the competence to speak, which, being part of our

biological capacity, is “universal and therefore essentially non-distinctive” (Bourdieu, 1991, p.

55), but rather the competence to speak the legitimate language, which produces “a profit of

distinction on the occasion of each social exchange” (Bourdieu, 1991, p. 55). In this way, in

the mid-20th

French context, Bourdieu differentiated between a literary orientation, which

may include Latinate vocabulary and constructions, as well as a striving for rare and novel

expressions, and a situational orientation, which may include vocabulary that is acquired

rather than learned and reliance upon figures of speech shared by a particular group. In short,

the concept of linguistic capital makes language competence move “from syntax to semantics

and pragmatics” (Bourdieu, 1977a, p. 646). Valued vocabulary, accent, and parlance

(Bourdieu, 1991, 1999b), together with legitimate linguistic strategies, such as “tension or

relaxation”, as well as “vigilance or condescension” (Bourdieu, 1977a, p. 654), can be

valuable aspects of linguistic capital.

Linguistic capital can reflect class-linked traits of speech (Bourdieu & Passeron, 1990).

More generally, linguistic capital reflects the power relationship between a speaker and a

listener. A speaker has to be not only understood but also believed, obeyed, respected, or

distinguished. A listener has to not only understand but also entitle the speaker to speak and

empower the speaker with the right to the legitimate language. As Bourdieu (1977a, p. 648)

explained, “those who speak regard those who listen as worthy to listen and those who listen

regard those who speak as worthy to speak.”

74

Language competence becomes linguistic capital only when it is valued and recognised

in a specific language market. For Bourdieu, there are “hierarchies of legitimacy” or “the

hierarchy of the arts, of genres etc.” (Bourdieu, 1989a, p. 86). As such, there is a hierarchical

situation in terms of the legitimised languages and their value in the linguistic market. That is

to say, different ways of speaking in a language can accrue different amount of linguistic

capital when being legitimised with different values in different situations. Along the

spectrum of a given language competence, there is a hierarchy of different ways of speaking

with different values, from the most legitimised to the least legitimised.

Bourdieu’s concept of linguistic capital also sheds light on people’s choice and pursuit

of different languages in a multicultural society where there is a spectrum of different

languages existing with different legitimised values. This is particularly true for HLs in a

dominant language context. Gogolin (2002) argued that German was positioned at the top of

the hierarchy in the German school education system, followed by different “layers” (p. 126)

of HLs, such as English, Turkish, HLs spoken by legal immigrants and their descendents, and

HLs spoken by illegal immigrants and their descendents. Likewise, there seems to be a

perceived hierarchy among different varieties of Chinese language. In Li and Zhu’s (2011)

study, British Chinese pupils, their parents, and their teachers considered Mandarin most

popular, Cantonese better than Hakka or Hokkien, and other regional varieties “rough” and

“uncultured” (p. 18). Given its potential to illuminate multilingual settings, Bourdieu’s

concept of ‘linguistic capital’ offers a meaningful theoretical tool for the current study to

examine CHL for Chinese immigrants and their descendents in Australia, where linguistic

plurality amongst settlers has been a reality since the first cohort of white immigrants landed

on this continent in the late 18th

century. When CHL is recognised and valued in a given

situation, Chinese Australians may invest their available capital to cultivate their CHL. At the

75

same time, they may expect a return on their investment, an access to a wider range of

different forms of capital. In contrast, Chinese Australians may not engage in CHL learning

when and where their CHL does not accrue any recognised value.

3.1.6 Section summary

As Bourdieu (1986) argues, capital, in its different forms, has the potential to produce profits,

which determine the chances of success for certain practices, and to reproduce itself in an

identical or expanded form through successful practices. Accordingly, linguistic capital can

realise the exchange of value and form through successful language learning. Drawing

insights from Bourdieu’s key concept of capital, the first part of the theoretical framework of

the current study can be modelled (see Figure 3.1).

Figure 3.1. Forms of capital and language

As shown in Figure 3.1, capital has four different forms, namely economic, cultural,

social, and symbolic capital, all of which play a salient role in the process of language

cultivation, a form of practice in Bourdieu’s term. The opportunities for language learners to

cultivate a TL are determined by the quantity and quality of various forms of capital available

to them. When the TL accrues value and becomes linguistic capital in the process of language

cultivation, capitalisation on the TL helps language learners gain access to a wider range of

Capital

Symbolic

capital

Social

capital

Cultural

capital

Economic

capital

Language as

linguistic capital

76

resources in the forms of economic, cultural, social, and symbolic capital. In this respect,

language learning and investment of various forms of capital are related reciprocally. The

reciprocal relationship between capital and language learning is demonstrated by the

double-headed arrow in Figure 3.1.

The reciprocal relationship between capital and language learning only makes sense at

a certain place and time. On the one hand, language learning can be understood as a form of

practice. As Bourdieu (1986) argues, the chances of success for practices are determined by

the structure and the distribution of different forms of capital at a given moment in a given

social situation. According to this perspective, the language market for the investment of the

amount and forms of capital may decide whether a given individual will choose to cultivate a

given language (Garnham & Williams, 1993). On the other hand, the language market

decides the value of the capitalised language, or linguistic capital, and decides the amount

and forms of other capital that this linguistic capital can produce as a return on the investment.

To understand this ‘reciprocal relationship’, it is useful to unpack Bourdieu’s notion ‘field’.

3.2 Field

As argued earlier, all forms of capital are necessarily interrelated because capital is

potentially convertible from one form into another in a given situation. Hence, the concept of

‘field’ is central to understanding Bourdieu’s uses of ‘capital’ in that capital does not exist and

cannot function except in relation to a field (Bourdieu & Wacquant, 1992). A classic

definition of ‘field’ was offered by Bourdieu (2011):

A field is a structured social space, a field of forces, a force field. It contains people

who dominate and people who are dominated. Constant, permanent relationships of

77

inequality operate inside this space, which at the same time becomes a space in which

various actors struggle for the transformation or preservation of the field. All the

individuals in this universe bring to the competition all the relative power at their

disposal. It is this power that defines their position in the field and, as a result, their

strategies. (pp. 40-41)

One of the essential aspects of field is Bourdieu’s view of individuals, groups, and

institutions as agents. As agents, human beings and institutions act consciously with

intentions. By virtue of agents’ consciousness and intention, a field becomes a configuration

of positions comprising agents, who struggle to defend or improve their positions (Bourdieu,

1993b) according to a specific interest of the field, or “practical mastery of its rules”

(Bourdieu & Wacquant, 1992, p. 117). In other words, the position of each particular agent in

a given field is a result of interaction within the specific rules of the field. The positioning

rules in fields are associated with forces acting on the resources at stake. Agents struggle and

compete to take up and occupy these positions in fields according to how many different

forms of symbolic and material resources they have (Bourdieu, 1993b; Bourdieu & Wacquant,

1992; Kramsch, 2008). The field structures strategic action for control over resources that are

construed as forms of capital (Bourdieu, 1991, 1993b). That is to say, to enter a field requires

capital and that capital in turn, enables positions in the field (Bourdieu, 1986, 1996; Lingard,

Rawolle, & Taylor, 2005). Agents within the field compete for control over the capital valued

by the field, and accumulate, conserve, convert, or utilise their different forms of capital in

this competition in order to secure a position within the field. Agents are positioned within a

given field in the first instance according to the overall amount and value of the capital they

possess; in the second instance, they are distributed according to the composition of their

78

capital, that is, the relative weight of different forms of capital in their total set of assets

(Bourdieu, 1985b, 1989b, 1991). Therefore, a field contains a network or a configuration of

relations between positions imposed upon agents by a) their present and potential distribution

of forms of capital, the possession of which commands access to specific profits that are at

stake in the field; and b) their relations to other positions, which could be dominant,

subordinate, or equivalent according to the rules of the field (Wacquant, 1989). In brief, a

field is a structured social space of relational, intentional, and multidimensional activities,

interactions, and transactions that contain agents with different positions according to the

rules of the field.

Not only does capital enable agents to enter a field and place them in a favourable

position in that field, but also capital grants agents ‘symbolic power’ to influence the rules of

the field. According to Bourdieu (1989b), symbolic power is the power to preserve or to

transform objective principles of union and separation, of association and dissociation; and to

conserve or to transform current classifications in matters of gender, nation, region, age,

social status, and language. Agents who capture more capital in various forms can wield their

symbolic power and influence the rules of the field according to their own interest. To this

end, they impose ‘symbolic violence’ on the less powerful others. According to Bourdieu

(1989a), symbolic violence is fundamentally the imposition of thoughts and perceptions upon

the dominated agents that tends to perpetuate the structures of the legitimated actions of the

dominant agents. As such, symbolic violence generates power relations that require the

dominant to dominate and the dominated to accept the domination.

All capital does not necessarily have the same value in a given field because different

fields with which the agents come into contact do not operate under the same rules; capital

only exists and only produces its effects in the field in which it is produced and reproduced

79

(Bourdieu, 1989a). That is to say, each field only generates and consolidates the values

recognised in the field and rewards the relevant capital by adding value to it. It is the field

that produces, reproduces, and legitimises what counts as resources and how they are

allocated. A given field may place a higher value on one form of capital than on another,

although that form of capital may be worth less in another field. The existence, efficacy, and

value of capital depend upon the nature of the rules within the field (Bourdieu, 1993b) in

such a way that capital is produced differently in different fields (Bourdieu, 1996). As

Bourdieu and Wacquant (1992, p. 117) further explain, “each field simultaneously

presupposes and generates a specific form of interest incommensurable with those that have

currency elsewhere.”

A field does not have a fixed boundary. It is dependent on the existence of other fields.

Consequently, a field lies along a continuum between autonomy and heteronomy (Bourdieu,

2000). Each field has enjoyed a certain degree of autonomy, defined by its ability to reject

external interventions and obey its own specific logic, interest, and rules (Bourdieu, 1993a).

When agents operate in a given field, they have to follow the internal rules of that field.

Agents without certain forms of capital valued by the field can be denied entrance into that

field. This demonstrates the autonomy of fields. While fields are relatively autonomous, they

are also constantly (re)shaped by both internal struggles and external developments in related

fields (Jackson, 2009). Internally, various agents struggle for transformation or preservation

of the field (Bourdieu, 1998). In other words, fields do not exist independently of the

consciousness of the agents; rather, fields are (re)produced by agents’ intentions. Externally,

every field may be part of one or several other larger fields, or may contain subfields itself.

No field enjoys complete autonomy or exists independently of other fields. In brief, a given

field imposes on, and is imposed on by both internal dynamics and external interventions.

80

In line with Bourdieu’s theorisation of field, the one degree of language proficiency

may be linguistically equal but not socially equal across different fields, in other words, may

be worth more or less as linguistic capital. In a social field where multiple languages co-exist,

when one language dominates the linguistic community, the linguistic field will make that

language the norm against which the value of other languages is measured (Bourdieu, 1977a).

This recognised domination will, in return, structure the rules of the field to measure other

languages against the legitimate language. Given the essential interdependent relationship

between capital and field, the theoretical framework demonstrated in Figure 3.1 can now be

elaborated into Figure 3.2 (Note the addition of the concept of field).

Figure 3.2. Forms of capital and language in fields

Although there is no research conceptualising Bourdieu’s ‘capital’ and ‘field’ in CHL

contexts, some empirical studies have implied the mutual constitution between ‘capital’ and

‘field’. For example, CHLLs may use CHL in personal conversations among themselves so

that no one else can understand them (J. Zhang, 2009). Entry into this field of conversation

demands CHL as linguistic capital, without which entry will be rejected. Furthermore, CHL

can be used to favourably position CHLLs in the economic field (Wong & Xiao, 2010). For

Fields

Capital

Symbolic

capital

Social

capital

Cultural

capital

Economic

capital

Language as

linguistic capital

81

example, CHL can be valuable cultural capital and symbolic capital in the global job market,

enabling CHLLs to become members of the dominant economic group. In the context of the

current study, Bourdieu’s notion of field helps explain how the structures within Chinese

Australians’ lived worlds shape their CHL learning and legitimise the value of their CHL

learning.

3.3 Habitus

The concept ‘field’ cannot be understood independently of the concept ‘capital’. Nor can it be

divorced from the concept ‘habitus’ (Bourdieu & Wacquant, 1992). A precursor of Bourdieu’s

concept of habitus can be found in Durkheim’s (1964) conceptualisation of the ‘collective

habit’. For Durkheim, the collective habit exists not only in an immanent state in the

successive acts that it generates, but also in the biological realm where it is permanently

expressed for all in a formula that is repeated from mouth to mouth, transmitted by education,

and fixed in writing. Similarly, Bourdieu (1998) argues that agents are endowed with durable

cognitive structures and a dispositional sense of action that direct them to appropriate

responses to given situations. This is what Bourdieu means by ‘habitus’. The habitus

integrates past experiences and functions at every moment as a matrix of perceptions,

appreciations, and actions, making possible the achievement of infinitely diversified tasks

(Bourdieu, 1977b). It refers to the internalisation of the external (Bourdieu, 1988, 1989a) by

which basic dimensions of external conditions in social life, such as gender, race, ethnicity,

norms, class, and occupation come to shape agents’ internal attitudes, values, perceptions,

and dispositions in ways of which they are seldom aware. This latter point is crucial. The

continuity and regularity of agents’ habitus is not a rational mechanism and action is not

principally a matter of rational choice (Bourdieu, 1977b). It is a generative rather than

82

determining mechanism. It is a mechanism that underpins agents’ being, doing, and thinking;

but does not decide them. This mechanism will now be described in some detail.

3.3.1 Conceptualising habitus

Bourdieu used a variety of wordings to explain what he meant by habitus. The following,

commonly cited definition is used here:

Systems of durable, transposable dispositions, structured structures predisposed to

function as structuring structures, that is, as principles of the generation and structuring

of practices and representations which can be objectively regulated and regular without

in any way being the product of obedience to rules, objectively adapted to their goals

without presupposing a conscious aiming at ends or an express mastery of the

operations necessary in order to attain them and, being all this, collectively orchestrated

without being the product of the orchestrating action of a conductor. (Bourdieu, 1977b,

p. 72)

In this classic definition, the term ‘disposition’ seems particularly suited to express what is

covered by the concept of habitus (Bourdieu, 1977b). A system of dispositions encompasses

several distinct meanings: the “result of an organising action” or a set of outcomes of internal

and external “structure”; a “way of being” or a “habitual state”; and a “predisposition”,

“tendency”, “propensity”, or “inclination” (Bourdieu, 1977b, p. 214). This classic definition

highlights the key features of habitus, namely its durable and transposable dispositions and its

structured and structuring structures (Bourdieu, 1977b). Elaboration of these attributes

follows.

83

Firstly, habitus is durable and transposable but not immutable (Bourdieu & Wacquant,

1992). Habitus is durable in that it lasts over time, and transposable in being capable of

becoming active within a wide variety of fields (Bourdieu, 1993b). This is demonstrated in

the way in which the relationship between ‘field’ and ‘habitus’ functions to produce agents’

deportment or manner. As the product of habitus, bodily dispositions are individual, subject to,

belonging to, and characteristic of the agents themselves. In other words, habitus is embodied

and expressed through dispositions, or durable and transposable ways of “standing, speaking,

walking, and thereby of feeling and thinking” (Bourdieu, 1990, p. 70). Habitus is the bodily

inscription of agents’ present and past positions in the social structure carried by individuals

at all times and in all places (Bourdieu, 1977b). Although habitus is a past that survives in the

present and tends to perpetuate itself into the future (Bourdieu, 1977b), it is not immutable.

Habitus can change constantly in response to new experiences (Bourdieu, 2000). As Bourdieu

and Wacquant (1992, p. 133) argue, “being the product of history, it is an open system of

dispositions that is constantly subjected to experiences, and therefore constantly affected by

them in a way that either reinforces or modifies its structures.” The body is “open to the

world, and therefore exposed to the world, and so capable of being conditioned by the world,

shaped by the material and cultural conditions of existence in which it is placed from the

beginning” (Bourdieu, 2000, p. 134). Dispositions, knowledge, and values are always

potentially subject to modification, rather than being passively consumed or reinscribed. This

can occur when narratives, values, and explanations generated by a habitus no longer make

sense; or when agents use their understanding and feel for the rules of the game as a means of

furthering and improving their own standing and capital within a field, as well as adapting

their dispositions. These dispositions therefore behave in ways that necessarily reproduce the

structures but in a modified form.

84

Secondly, Bourdieu defines habitus as a property of social agents, such as individuals,

groups, or institutions, that comprises a “structured and structuring structure” (Bourdieu,

1994, p. 170). Habitus is structured by agents’ past circumstances, such as cultural history,

family upbringing, and educational experiences. Existing knowledge, or the way agents

understand the world, their beliefs, and values, is always structured through habitus, rather

than merely recorded passively. Agents are disposed towards certain attitudes, values, or

ways of behaving because of the influence exerted by their cultural trajectories. Habitus is

structured by these shared cultural trajectories and the conditions associated with a particular

type of existence. As Durkheim (1938) argued earlier,

In each of us, in varying proportions, there is part of yesterday’s man [sic]; it is

yesterday’s man who inevitably predominates in us, since the present amounts to little

compared with the long past in the course of which we were formed and from which we

result. Yet we do not sense this man of the past, because he is inveterate in us; he makes

up the unconscious part of ourselves. (p. 70)

In this sense, habitus is a structured structure. At the same time, habitus is structuring agents’

present and future practices. It is always structuring the moments of practice when a set of

dispositions meets a particular problem or choice in a given context or social space. In this

regard, habitus generates practices, beliefs, perceptions, and feelings in accordance with its

own structure (Maton, 2008). These practices, beliefs, perceptions, and feelings in return

shape agents’ future possibilities and set agents on a particular path that further shapes their

understanding of themselves and of the world (Maton, 2008). As such, habitus is a structuring

structure.

85

Thirdly, habitus is durable and transposable because it is a structured structure. By this

expression, Bourdieu refers to the largely unconscious acceptance of rules, values, and

dispositions as ‘habitus’. This habitus is largely unconscious because it is structured inside

the agents’ body. It is internalised and preconscious, and will be primarily responsible for

generating many of the actions that agents take. The agents rarely have to think rationally

about how to behave or what to say. Their bodies know. As Bourdieu (1987) further states,

Habitus entertains with the social world which has produced it a real ontological

complicity, the source of cognition without consciousness, of an intentionality without

intention, and a practical mastery of the world’s regularities which allows one to

anticipate in the future without even needing to posit it as such. (p. 12)

In other words, habitus is the source of series of actions by agents “which are objectively

organised as strategies without being the product of a genuine strategic intention” (Bourdieu,

1977b, p. 73). Therefore, the durable and transposable habitus is unconsciously structured

and generative.

Fourthly, habitus is not immutable because it is also a structuring structure. Values,

tendencies, and dispositions allow people to respond to shifting cultural rules and contexts in

a variety of ways because habitus allows for improvisations. Hence, Bourdieu defines habitus

as “the durably installed generative principle of regulated improvisations” (Bourdieu, 1977b,

p. 78) and considers the work of habitus as “the intentionless invention of regulated

improvisation” (Bourdieu, 1977b, p. 79). The ‘regulated improvisation’ suggests that habitus

can be transformed and modified (Pahl, 2008). It is the improvisation that offers the

possibility of change within a field (Pahl, 2008). Every agent, wittingly or unwittingly, is both

86

a producer and a reproducer of social structures. Agents’ responses are always largely shaped

and regulated by their habitus. When agents’ habitus, or natural self, is affirmed by people

around them in conformity with the expected norms, their habitus coincides with that of

others who belong to the same culture, class, and way of life in the situation. Therefore,

habitus is not immutable because it is structuring the consciousness in an ongoing process.

Putting these attributes together, habitus can be thought of as the underpinning

mediation that produces agents’ practices, without either explicit reason or signifying intent,

to be none the less sensible and reasonable (Bourdieu, 1977b). Habitus serves as a cognitive

map that routinely guides and evaluates agents’ choices and options. Habitus operates at a

level that is simultaneously conscious and unconscious. Agents do think and act in strategic

ways, and try to use the rules of the field to their advantage, but at the same time they are

influenced, or almost driven, by the values and expectations that they have internalised in

their habitus. Though they may be conscious of making moves and acting strategically, they

are unaware that their motives, goals, and aspirations are not spontaneous or natural, but are

derived from the habitus. Habitus represents a set of dispositions and taken-for-granted ways

of thinking and behaving that agents gradually develop over time. It is reflective of agents’

day-to-day experiences and the routine set of social relationships within which they are

embedded. As such, habitus can be regarded as the internalisation or embodiment in

individuals of the social environment within which they are located (Connolly, 2011).

Habitus is not a random or unpatterned structure but a structure systematically ordered in the

embodied state, persisting across time as a system of dispositions that generate perceptions,

appreciations, and practices (Bourdieu, 1990).

Of particular relevance to this study, habitus can be understood as values, tendencies,

deportment or manner, and dispositions gained from cultural history and social experience in

87

fields. It makes agents think and act in certain ways that are inculcated and structured by

cultural and social history. Bourdieu (1993b) has explained:

The habitus, as the word implies, is that which one has acquired, but which has become

durably incorporated in the body in the form of permanent dispositions. So the term

constantly reminds us that it refers to something historical, linked to individual history,

and that it belongs to a genetic mode of thought, as opposed to essentialist modes of

thought. (p. 86)

Habitus captures how people carry their culture, experience, and history within themselves,

and how they make choices to act in certain ways rather than others. As a system of

dispositions to certain practices, habitus constructs an objective basis for regular modes of

behaviour (Bourdieu, 1994). These modes of behaviour can be predicted by virtue of the

effect of the habitus because agents who are equipped with it will behave in a certain way in

certain circumstances (Bourdieu, 1994). In these behavioural practices, habitus has the

tendency to reproduce itself. It can produce agents endowed with the systems of dispositions

that are capable of engendering the behavioural practices adapted to the field and thereby

contributing ultimately to the reproduction of the habitus (Bourdieu, 1973).

As argued at the very beginning of this section, habitus cannot be understood

independently of a given field. Bourdieu and Wacquant (1992) explained the relationship

between habitus and field:

On the one side, it is a relation of conditioning: the field structures the habitus, which is

the product of the embodiment of the immanent necessity of a field. On the other side,

88

it is a relation of knowledge or cognitive construction. Habitus contributes to

constituting the field as a meaningful world: a world endowed with sense and value, in

which it is worth investing one’s energy. (p. 127)

The relationship between habitus and field helps make sense of agents’ language choices in

certain situations, which can be understood as the product of a “sub-set of dispositions

acquired in the course of learning to speak in particular contexts (the family, the peer group,

the school, etc.)” (Bourdieu, 1991, p. 17). These socially constructed dispositions imply a

certain capacity to speak and a certain propensity to say given things in a particular situation.

This capacity can be understood as the control over linguistic capital, which involves both the

linguistic capacity to generate an infinite number of grammatically correct utterances, and the

social capacity to use this competence adequately in a given situation. This argument

illuminates the way that habitus is the generative basis for language choices and also suggests

a causal link between habitus and language choices in a given field (Bourdieu, 1991).

At the same time, there is a widespread belief that embodied dispositions were

produced through usage of a shared language (Joseph, 2009). An empirical study involving

interviews with Chinese parents in a region of central Scotland (A. Hancock, 2006) echoed

Joseph’s argument. Acquiring CHL literacy was perceived by the parents to be linked with

the transmission of traditional Chinese cultural beliefs and values. When asked about the

attitude towards children’s learning Chinese literacy, one parent replied: “Poor Chinese, poor

Chinese person” (A. Hancock, 2006, p. 363). What emerged from this investigation was that

CHL literacy was considered to be inextricably bound together with the maintenance and

development of Chinese cultural dispositions. The fact that agents spend formative years

learning language results in their acquiring a habitus that will endure into their adult life

89

(Joseph, 2009). Therefore, habitus is structured or restructured through language and

language is central to habitus (Joseph, 2009).

For Bourdieu, the mutually constitutive effect between habitus and language choice

and practice existed within a given language, namely French. The current study extends this

theoretical perspective to a mutually constitutive effect between habitus and language choices

and practices across different languages. As such, habitus can be used as a theoretical lens to

trace the language development of learners who encounter structural forces while enacting

their agentive will in a field (Costa, 2010). The third part of the theoretical framework can

now be modelled. As shown in Figure 3.3, the double-headed arrow demonstrates the

mutually constitutive effect between habitus and language learning.

Figure 3.3. Habitus and language in fields

3.3.2 Interpreting ‘Chineseness’ through habitus

Ascribed attributes and identities, such as race, ethnicity, gender, and native language, are not

of people’s own choice, and thereby they remain an embodied presence and cannot be erased

or made over. Habitus explains the tendency to perpetuate these attributes and identities

(Bourdieu, 1996). But people may have their own intentions to hide, alter, redesign, or

garnish their identities by certain degrees (Luke, 2009). As such, certain aspects of identity,

although durable and transposable, are not immutable. Therefore, the concept of habitus, as a

foundational basis of kinship and culture, offers an opportunity to describe the tensions

Language as

linguistic capital

Fields

Habitus

90

between multiple positioning and identities (Luke, 2009) and to examine the relationship

between practices and identities (Pahl, 2008).

Though it is impossible for all individuals of the same social group to have had exactly

the same experiences and in exactly the same order, each individual of the same group is

more likely than individuals of another group to have encountered situations common

amongst members of the group. Accordingly, within a particular field, individuals who

occupy similar positions are likely to have similar dispositions and therefore to produce

similar practices (Bourdieu, 1985b). In line with this perspective, Costello (2005) conceives

of identity as a set of largely unintentional dispositions, using the notion of habitus to

explicate how identity affects cognitive style as well as embodied deportment. Similarly,

Holland and colleagues (Holland, Lachicotte, Skinner, & Cain, 1998) emphasise the

production and transformation of habitus, seeing it as a fundamental but not final or given

aspect of identity construction. This emphasis is shared by Bartlett and Holland (2002), who

use the concept of habitus as a way of describing and analysing identity formation in practice.

Rowsell (2008) also emphasises the interplay between habitus and identity practices. In brief,

identity construction is often less than fully conscious (Bucholtz & Hall, 2009) and is often

tied to habitual practice (Bourdieu, 1977b). Identity draws upon and reflects habitus (Zacher,

2008). Habitus, in this way, comes to generate identity (Rowsell, 2008). Therefore,

Bourdieu’s key concept ‘habitus’ offers a theoretical instrument to examine people’s identity.

One of the ways in which attempts have been made to make sense of ethnic identity has

also been through Bourdieu’s notion of habitus (Connolly, 2011). Drawing insights from this

perspective, Connolly (2011) demonstrates how children have already begun to embody and

internalise the cultural dispositions and ethnic awareness of their respective ethnic groups.

These cultural propensities of the ethnic groups may not stem from people’s own conscious

91

choice, and hereby may remain durable and transposable across different times and places in

people’s lives. These embodied dispositions, such as affiliated cultural, experiential, and

historical memories, stay (Luke, 2009; J. Webb, Schirato, & Danaher, 2002). In this respect,

ethnic and racial dimensions are constitutive of habitus (Cockerham & Hinote, 2009;

Diamond, Randolph, & Spillane, 2004; Horvat & Antonio, 1999) and habitus can be shaped

by these ethnic and racial dimensions (McClelland, 1990; Reay, 2004). This argument

augments Bourdieu’s own focus on class in habitus.

Although the potential of employing Bourdieu’s concept of habitus in ethnicity

research is evident, few studies have related ‘habitus’ to an analysis of ethnicity, Asian

ethnicity in particular. Given this, the current study examines Chinese Australians’ ethnic

identity of ‘Chineseness’ with reference to Bourdieu’s concept of habitus, interpreting it as a

set of durable and transposable tendencies to think and act in such a way that has been

inculcated by their Chinese heritage, in which they share the same tastes, behaviours, values,

and way of life. Although this ethnic identity may change constantly in response to new

experiences, the change is “never radical, because it (habitus) works on the basis of the

premises established in the previous state” (Bourdieu, 2000, p. 161). In short, the construction

of an ethnic identity makes sense through habitus (Jenkins, 1992) and derives from the shared

cultural history (Rowsell, 2008).

Culture is one of the most complicated concepts in the English language mainly

because it has come to be used across several distinct intellectual disciplines (Williams, 1988).

For Bourdieu, culture does not only refer to the objectified cultural capital, or cultural goods

belonging to consumers, but also refers to consumers’ embodied competence in the

consumption of these cultural goods. Culture thereby constitutes appropriate and legitimate

tastes for these cultural goods, where taste is “the faculty of perceiving flavours” and “the

92

capacity to discern aesthetic values” (Bourdieu, 1989a, p. 474). It is a “gift of nature” born

from a family and “the product of upbringing and education” (Bourdieu, 1989a, p. 1). Due to

the socially recognised hierarchy of family origin and educational system, taste, “a system of

classificatory schemes” (Bourdieu, 1989a, p. 174) or “an acquired disposition to differentiate

and appreciate” (Bourdieu, 1989a, p. 466), is predisposed to classify people into different

cultural groups. Consequently, people with different tastes are marked to have different

culture, from the “authentic” culture to the “imitation”, from the “true” culture to the

“popularisation”, or from the “high culture” to the “middle-brow culture” (Bourdieu, 1989a,

p. 250). This hierarchy is what Bourdieu meant by “distinction” (Bourdieu, 1989a). In this

sense, to have the authentic, true, and high culture is to “know the best that has been said and

thought in the world” (Arnold, 1999, p. xxxiii).

However, Arnold’s elitist conception of culture is concise but problematic. What is

considered ‘the best’ is not fixed but highly contingent on time and space. While the colonial

age purported European culture to be the most advanced form of civilisation, the two World

Wars shook this complacency. Those wars in concert with subsequent anti-colonialist

movements contributed to changes in attitudes toward culture and how it was understood (K.

Louie, 2008). As Johann Gottfried von Herder (1744-1803), the German philosopher, argued

against the high and dominant position of European culture purported by the colonial age,

“the very thought of a superior European culture is a blatant insult to the majesty of Nature”

(Williams, 1988, p. 89). Within European culture, different tastes may be used to distinguish

culture considered to be of higher value from that considered to be of lower value. However,

these tastes should not give more value to European culture than non-European cultures.

Along the cultural spectrum, there are different cultures of the same value. Within each

culture, there will be a hierarchy of ‘high’ to ‘low’ culture. In Williams’ term, this is culture

93

conceived as “a general process of intellectual, spiritual, and aesthetic development” (1988, p.

90). This study speaks of culture not only in the hierarchical sense of certain social and

economic groups within a particular field, but also in the specific and variable cultures of

different nations and periods across fields. In Williams’ term, this is culture conceived as “a

particular way of life, whether of a people, a period, a group” (1988, p. 90).

Parallel to Bourdieu’s ‘culture’ in the French context, there exists Confucian culture in

the Chinese context, where to have ‘culture’ generally refers to having a certain amount of

formal education, being at a certain level of social and economic status, and perpetuating

such social norms. Confucianism has traditionally valued education, thereby constructing a

hierarchy of social and economic positions, and legitimate social norms (K. Louie, 2008).

Accordingly, there are high or low forms of culture within Confucian thought. Adherence to

this Confucian value system, or ‘taste’ in Bourdieu’s term, can distinguish a strong sense of

‘Chineseness’ from a weak one.

Attributes associated with ‘Chineseness’ result from both intentional and unintentional

learning. As Bourdieu suggested, both types of learning are made possible by habitus

acquired through culture (Bourdieu, 1989a) and produced through history (Bourdieu, 1990).

Because Confucianism is the bedrock, even the definitive core, of Chinese culture (Tan,

2008), it can therefore be understood to constitute a cultural history or “previous state”

(Bourdieu, 2000, p. 161) for ‘Chineseness’. Since Confucianism is deeply rooted in Chinese

societies and highly valued in Chinese social fields, it has become a generative mechanism

behind Chinese people’s thinking, being, and doing. In this respect, “history turned into

nature” (Bourdieu, 1977b, p. 78) because what historically needed to be durable and

transposable through a process of continuous reproduction is now inscribed through social

regulations, forms, and norms.

94

As a set of embodied dispositions, ‘Chineseness’ is understood as both the

consciousness and the unconsciousness of the Confucian way of doing and understanding

things. As habitus, ‘Chineseness’ rooted in the Confucian culture is durable and transposable.

It is durable because the core Confucian values have an enduring impact on Chinese people

today (W. O. Lee, 1996), being the dynamic force that directs Chinese life and generates

forms of Chinese life (Tan, 2008). It is also transposable because Confucian culture can be

carried out and enacted by Chinese people who reside not just in China but also throughout

the world. Chinese people, wherever they are in the world, “represent a general ‘Chineseness’,

deriving from the Confucian heritage itself”, which helps the understanding of their being and

learning (Biggs & Watkins, 1996, p. 269). For the purpose of this thesis, it is helpful to

introduce the core values of Confucianism that underpin this habitus of ‘Chineseness’.

3.3.3 Confucian dispositions

Confucianism developed from the ethics and philosophy of the Chinese philosopher

Confucius (551-478 BC), whose principal concept was to maintain harmony and order and

thus keep society together without the undue exercise of force (Clayre, 1984). In theory,

Confucianism is a complex system of moral, social, political, philosophical, educational,

quasi-religious, and ideological thought that has influenced the culture of China and some

countries in East Asia and Southeast Asia. It plays an important role in Chinese civilisation

and has a deep impact on Chinese society, Chinese education, Chinese culture, and Chinese

people.

The dimensions of Confucianism have accumulated over the past 2,500 years.

Different historical eras have reflected different interpretations of these dimensions, but there

is a core set of values consisting of Wuchang (五常, Five Constants) and Sizi (四字, Four

95

Virtues). During the Western Han (206 BC-8 AD) Dynasty, Zhongshu Dong (179-104 BC), in

his book called 《春秋繁露》 (Rich Dew of Spring and Autumn), described the classical

dimensions of Wuchang (五常, Five Constants), namely Ren (仁, Benevolence), Yi (义,

Righteousness), Li (礼, Ritual or Propriety), Zhi (智, Knowledge or Wisdom), and Xin (信,

Integrity or Trustworthiness). During the Ming (1368-1644) Dynasty, Zhonglin Xu

(1567-1620), in Chapter 20 of his book called 《封神演义》 (The Legend of Deification),

described the classical dimensions of Sizi (四字, Four Virtues), namely Zhong (忠, Loyalty),

Xiao (孝, Filial Piety), Jie (节, Continence), and Yi (义, Righteousness). These values were

elaborated in ancient publications about Confucianism. The most authoritative ones are Sishu

(《四书》, Four Books) and Wujing (《五经》, Five Classics). The Sishu (《四书》, Four Books)

are Daxue (《大学》, Great Learning), Zhongyong (《中庸》, Doctrine of the Mean), Lunyu

(《论语》, Analects), and Mengzi (《孟子》, Mencius). The Wujing (《五经》, Five Classics)

are Shijing (《诗经》, Classic of Poetry), Shangshu (《尚书》, Classic of History), Liji (《礼

记》, Classic of Rites), Zhouyi (《周易》, Classic of Changes), and Chunqiu (《春秋》, Spring

and Autumn Annals). The texts of these ancient publications consist of fragments not written

by Confucius himself but compiled by his disciples and second-generation disciples and

collated over a long period of time (Lai, 2008). A core set of Confucian values consisting of

Wuchang (五常, Five Constants) and Sizi (四字, Four Virtues) elaborated in these ancient

publications will be explained in the following sections.

The Wuchang (五常, Five Constants)

Ren (仁, Benevolence) and Yi (义, Righteousness) are fundamental values within

Confucianism. Philanthropy is the key practice of Ren (仁, Benevolence) and Yi (义,

96

Righteousness), by which people can establish healthy and harmonious relationships.

Confucianism argues that people should love each other and should “老吾老以及人之老，幼

吾幼以及人之幼” （《孟子·梁惠王上》, Mencius）. This means people should respect not

only their own parents, but also others’ parents; and should look after not only their own

children, but also others’ children. Confucius also said that “君子成人之美” (《论语·颜渊》,

Analects). This means that a good person has the virtue to oblige others; in other words, a

good person will pursue every effort to help others achieve their goals or meet their demands.

This can explain why Chinese people often feel uncomfortable about refusing requests from

others. Otherwise, it is hard to maintain their face, or a good impression. These dispositions

of ‘Chineseness’ are associated with the values of Ren (仁, Benevolence) and Yi (义,

Righteousness).

Li (礼, Ritual or Propriety) refers to secular ceremonial behaviour including the

propriety or politeness of everyday life. It was codified and treated as a comprehensive

system of norms. Shaping rituals in a way that leads to the health of society and its people is

one purpose behind Confucian philosophy. With Li (礼, Ritual or Propriety), propriety of

behaviour is internalised and exerts its influence before peoples’ actions are taken. People

behave properly because they want to avoid losing face. With Li (礼, Ritual or Propriety),

people also understand and acknowledge each person’s correct social position. Li (礼, Ritual

or Propriety) builds hierarchical relationships through protocols, assigning everyone a place

in society with a proper form of behaviour. It is used to distinguish between the younger and

the older, or between the student and the teacher. Social disorder is considered to stem from

the failure to call things or people by their proper names. The Confucian solution to such

disorder is Zhengming (正名, rectification of terms) (《论语》, Analects), by which Confucius

was concerned with the use of titles to establish normative relationship obligations (《论语》,

97

Analects). This ideal form of social interaction shaped by Zhengming (正名, rectification of

terms), an aspect of Li (礼，Ritual), is associated with the Chinese disposition that people

often address those who are senior in age and/or position by their title plus surname rather

than their first name. This form of social interaction “is not lubricated with the democratic oil

of warmth and first names, but with the oil of respect”, which functions as an effective

lubricant in a hierarchical Confucian culture (Biggs, 1998, p. 730).

Confucianism pays great heed to Zhi (智, Knowledge or Wisdom). Confucius said that

“朝闻道，夕死可矣” (《论语·里仁》, Analects), which means, “If I had mastered the

knowledge this morning, I would not have regretted it even if I had to die this evening.”

Accordingly, most essential is the Confucian belief in human self-perfection pursued as the

highest purpose of life through personal commitment to learning. Influenced by this

Confucian belief, Chinese people consider learning honourable and firmly believe that

education is of paramount importance in people’s life (Biggs, 1998). In order to be at the top

of society, one must be a scholar. Chinese students’ motivation for socioeconomic

advancement through education leads them to study harder (Salili, Chiu, & Lai, 2001). In

Australia, a case study (Zhao & Singh, 2011) demonstrated that Chinese-Australian parents

have higher academic expectations for their children than Anglo-Australian parents; and they

also attach a higher value to the academic aspects of their children’s life. To some extent, the

Chinese-Australian parents’ approach to their children’s academic education can be attributed

to their ‘Chineseness’ in pursuing Zhi (智, Knowledge or Wisdom).

Xin (信, Integrity or Trustworthiness) can be literally understood as credibility, which

is valued not only in Confucianism but also in all cultures. To be trustworthy, people should

do what they say and keep their promises. Like the other dimensions of Wuchang (五常, Five

Constants), Xin (信, Integrity or Trustworthiness) is an embodied disposition inculcated by

98

Confucian heritage; unlike the other dimensions, Xin (信, Integrity or Trustworthiness) has to

be granted by others. People with Xin (信, Integrity or Trustworthiness) are trusted only when

others think these people are worthy of their trust.

The Wuchang (五常, Five Constants), as its name suggests, is the constant belief

system of Confucianism. It has constructed the foundation of Chinese ethics, considered the

root of humanity. Due to this constant Confucian belief system, the Wuchang (五常, Five

Constants) has an enduring impact on Chinese people, embedded in their dispositions. As

such, the Wuchang (五常, Five Constants) can be considered as the underpinnings of the

habitus of ‘Chineseness’.

The Sizi (四字, Four Virtues)

Zhong (忠, Loyalty) emphasises obligations of the ruled to the ruler, the dominated to the

dominant, and the less powerful to the more powerful, but puts less emphasis on the

obligations of the ruler to the ruled. With Zhong (忠, Loyalty), the social hierarchy is

maintained. Zhong (忠, Loyalty) is also an extension of one’s duties to the country, the nation,

the family, and friends.

Xiao (孝, Filial Piety) characterises the respect that children should show to their

parents. Confucius said that “父母在，不远游，游必有方” (《论语·里仁》, Analects), which

means, “When your parents are alive, do not travel too much; if you have to, you must let

your parents know where you are.” Chinese people often stay close to their family members,

especially close to the older generation. This can be attributed to their ‘Chineseness’ to fulfil

their filial piety to the older generation. Mencius (372-289 BC), one of the disciples of

Confucius, said that “不孝有三，无后为大” (《孟子·离娄上》, Mencius), which means,

99

“There are three things that can challenge Filial Piety. Being without children, in particular

without sons to continue the patrilineal line, challenges Filial Piety the most.” This

disposition of ‘Chineseness’ is unique as it relates to gender, because traditional Chinese

genealogical notions of identity exclude women.

Jie (节, Continence) is an important Chinese virtue. In Shangshu (《尚书·大禹谟》,

Classic of History), Confucianism argues that “克俭于家”, which means, “Frugality is

important in family life.” In Zhouyi (《周易·否》, Classic of Changes), Confucianism argues

that “俭德避难”, which means, “The virtue of frugality overcomes difficulty.” In Zuozhuan

(《左传·庄公二十四年》, Zuo’s Commentary), a book written by Qiuming Zuo (556-451 BC)

to interpret Chunqiu (《春秋》, Spring and Autumn Annals) written by Confucius’ disciples,

Confucianism argues that “俭，德之共也”, which means, “Frugality is a common character of

people with virtue.” Traditionally, Chinese people live frugal lives. This disposition of

‘Chineseness’ can be attributed to the value of Jie (节, Continence).

Yi (义, Righteousness), as one of the dimensions in Sizi (四字, Four Virtues), is

equivalent of that in Wuchang (五常, Five Constants) as explained above.

The Sizi (四字, Four Virtues), deeply rooted in the Confucian cultural history, has

served Chinese people as a guide to moral living and shaped their doings in a particularly

Chinese way. Chinese people’s propensities associated with their habitus of ‘Chineseness’ can

be largely attributed to the core values in the Sizi (四字, Four Virtues).

The Liuyi (六艺, Six Skills)

Besides Wuchang (五常, Five Constants) and Sizi (四字, Four Virtues), Confucianism

requires people to grasp Liuyi (六艺, Six Skills), namely Li (礼, Rites), Yue (乐, Music), She

100

(射, Archery), Yu (御, Equestrianism), Shu (书, Calligraphy), and Shu (数, Mathematics).

These six skills were first documented in Zhouli (《周礼》, The Rites of Zhou Dynasty): “养

国子以道，乃教之六艺：一曰五礼、二曰六乐、三曰五射、四曰五御、五曰六书、六曰

九数”, which means, “Men who excel in these six skills were considered to have reached the

state of perfection.” These skills were also documented in 《三字经》 (Three-Character

Classic) (Wang, 1223-1296): “礼乐射，御书数，古六艺”, which means, “The traditional six

skills are rites, music, archery, equestrianism, calligraphy, and mathematics.” These skills

incorporated both military and civil components. This strongly supports the observation that

Chinese students often excel at music, calligraphy, and mathematics. This argument is

consistent with the Australian data. For instance, Chinese Australian students have been

shown to outperform white Australian students in mathematics (Lokan, Ford, & Greenwood,

1997; Mu, 2012; Zhao & Singh, 2011). Besides the pedagogical and psychological

explanations to this better performance, the ‘Chineseness’ associated with the value of Liuyi

(六艺, Six Skills) may serve as a generative mechanism underpinning Chinese students’

relative strength in mathematics (Mu, 2012).

The Liuyi (六艺, Six Skills) focuses on the cultivation of a life of perfect goodness. It

serves as a foundation of Chinese people’s cultural dispositions associated with their habitus

of ‘Chineseness’. In this respect, the Liuyi (六艺, Six Skills) largely explains Chinese

people’s perceptions and behaviours in relation to their cultural tastes.

The crisis and rejuvenation of Confucianism

Confucianism has developed and evolved over the past 2,500 years. However, several

historical events nearly devastated this tradition. Fenshu Kengru (焚书坑儒, Burning of the

101

books and burying of the scholars) refers to a policy and a sequence of movements in the Qin

Dynasty, between the period of 213 BC and 206 BC. Shiji (《史记·卷 006·秦始皇本纪》,

Chapter 6: The Basic Annals of the First Emperor of Qin, in Records of the Grand Historian)

documents:

臣请史官非秦记皆烧之。非博士官所职，天下敢有藏《诗》、《书》、百家语者，悉

诣守、尉杂等烧之。有敢偶语《诗》、《书》者弃市。以古非今者族。吏见知不举

者与其同罪。令下三十日不烧，黥为城旦。

This historical record can be translated as follows: After Emperor Qin (秦始皇, the first

emperor of China) unified China in 221 BC, his Prime Minister Si Li proposed that all

histories in the imperial archives except those written by the Qin historians be burned; that

Shijing (《诗经》, Classic of Poetry) and Shangshu (《尚书》, Classic of History) be collected

by the local authorities for burning; that anyone discussing these two books be executed; that

anyone using history to criticise the present be put to death, along with their families; that

authorities who failed to report cases that came to their attention be equally guilty; and

anyone who had not burned the two books within 30 days of the decree be banished to build

the Great Wall. Politically, Emperor Qin (秦始皇, the first emperor of China) strengthened

his central power and dictatorship by the movement of Fenshu Kengru (焚书坑儒, Burning

of the books and burying of the scholars). However, culturally, he wreaked irreversible

damage by destroying most collections of ancient Chinese books published prior to the Qin

Dynasty. As Chapter 121 of Shiji (《史记·卷 121·儒林列传》, Records of the Grand Historian)

documents, “及至秦之季世，焚诗书，坑术士，六艺从此缺焉。” This can be translated as

follows: By the end of Qin dynasty, documents of Liuyi (六艺, Six Skills) have nearly

102

disappeared due to the burying of the scholars and the burning of Shijing (《诗经》, Classic of

Poetry) and Shangshu (《尚书》, Classic of History).

Another blow to Confucianism was caused by the Cultural Revolution. This

socio-political movement (1966-76) did significant economic, social, and cultural damage to

the People’s Republic of China (Xi & Gao, 2005). One of the stated goals of the Cultural

Revolution was to bring an end to the Four Olds: Old Ideas, Old Cultures, Old Customs, and

Old Habits (Chen, 1966). Old Ideas refers to Confucian thought and the works of Confucius

and his successors (Chen, 1966). Hence, the slogan accompanying this movement was ‘to

criticise Confucius’. This movement was devastating. At the material level, it caused

dramatic damage to Confucian heritage. Architecture was ransacked; literature and the

classics were burned; paintings were torn apart; and antiques were shattered (Xi & Gao,

2005). At the spiritual level, it attempted to sweep away social norms and human virtues

established by Confucianism over the previous 2,500 years. People arbitrarily used violence

to break rules and laws. Without a spiritual home, China became a chaotic society at that

time.

Starting in the early 1990s and continuing into the 21st century, there has been a

massive rebuilding effort to restore the culture that was destroyed or damaged during the

Cultural Revolution (Xi & Gao, 2005). This includes the re-instatement of the value system

of Confucianism. The resurgence of Confucianism continues into the new century. Today,

Confucianism enjoys a robust rejuvenation as “an indispensable cultural force” indicated in

the “culture craze” and “national learning craze” (Deng, 2011, p. 563). Accompanied by a

strong interest in the rediscovery of Chinese tradition and history, China is returning to its

own roots for inspiration (Deng, 2011).

103

Confucianism has survived the Fenshu Kengru (焚书坑儒, Burning of the books and

burying of the scholars) and the Cultural Revolution. Despite several attempts to destroy it,

Confucianism remains an integral part of the psycho-cultural construct of Chinese people

today (Tan, 2008) and the texture of contemporary Chinese life (Clayre, 1984). It has had a

deep and extensive impact on various dimensions of Chinese society, including history,

education, arts, ethics, religion, law, politics, and the military (Q. Y. Zhang, 2009). Two and a

half thousand years after his death, Confucius could still represent an entire way of thinking

and living: Confucianism. Therefore, Confucianism continues to provide the basis of the core

value of Chinese people (Q. Y. Zhang, 2009).

People of Chinese ancestry, such as Chinese Australians, no matter how much capital

they capture or invest in whatever fields, are likely to retain aspects of their ‘Chineseness’ as

habitus, at least to a certain extent. But the dispositions brought to a given field do not

preclude deliberate remaking of the habitus. Therefore, the habitus of ‘Chineseness’ may vary

in anticipation of, and in response to, the positioning that occurs through structural

distinctions and categories of discourse that constitute rules of exchange within fields

(Bourdieu, 1998). Therefore, when examining the interactions between Chinese Australians’

‘Chineseness’ and their CHL practice, their ‘Chineseness’ should be interpreted as habitus

animated within a particular field and interacting with the power of the field.

3.4 Chapter summary

With the concept of habitus, Bourdieu seeks to grasp the generative principles that underlie

agents’ practices in specific cultural contexts and social settings, or ‘fields’. Hence, particular

practices should be understood as not only the product of the habitus but also the product of

these fields, as well as the product of the relations between the habitus and the fields. These

104

specific fields are structured spaces of interrelations where agents’ positions are determined

by the distribution of different kind of resources, or ‘capital’. Agents struggle to maintain or

change the distribution of the forms of capital according to the rules of the fields. Those rules

determine the value of capital and allow one form of capital to be converted into another.

Therefore, habitus, field, and capital are three entangled concepts when interpreting agents’

practices.

Language learning and its usage, as forms of practices, receive their value and make

sense only in their relations to a field (Bourdieu, 1991). Hence, the positions and the

dispositions of language learners within a given field are essential to understanding of their

language choices. Bourdieu (1991) has theorised that people make language choices

according to the amount of different resources (capital) as well as the dispositions (habitus)

that they have with reference to a particular situation (field). For Bourdieu, these language

choices refer to a particular accent, a certain vocabulary, or an appropriate way of speaking in

a given language that would be legitimised within a given situation. This study extends this

Bourdieusian stance to choices between different languages. As such, the choice of a

particular language in a given field is then affected by the quantity and the quality of capital

and habitus that are available to the agents in that field. Possession of certain forms and

combinations of capital and a certain habitus lead to the ability to anticipate the ‘language

game’. This ability determines whether an agent has a “basic minimum of chances in the

game, and therefore power over the game” and hence can invest in the potential profits of that

game (Bourdieu, 2000, p. 220).

In the CHL context, some empirical studies have indicated the essential roles played by

CHLLs’ commitment and their ethnic identity in CHL learning processes. Bourdieu’s

concepts of ‘capital’, ‘habitus’, and ‘field’ will shed light on these empirical findings.

105

However, there is scant research that integrates capital, habitus, and field into the

investigation of CHLLs’ language choices. In light of the Bourdieusian sociological

perspectives presented here, a theoretical framework was constructed to guide the research

reported in this thesis. The theoretical framework is demonstrated in Figure 3.4, which is the

synthesis of Figure 3.1, 3.2, and 3.3.

Figure 3.4. Theoretical framework

This framework talks back to the informing theories and the existing literature, and

guides the examination of the interdependence of Chinese Australians’ ‘Chineseness’ qua

habitus, their various resources as capital, and their CHL proficiency as a form of linguistic

capital that results from their practice of CHL learning within particular fields. Specifically,

the theoretical framework underpins the investigation of the two research questions, which

can be posed in more theoretical terms as follows:

RQ1. Is CHL proficiency of young Chinese Australian adults influenced by their

investment of capital, the strength of their habitus of ‘Chineseness’, or both?

RQ2. How do young Chinese Australian adults understand their CHL learning in

relation to (potential) profits produced by this linguistic capital in given fields?

The next chapter explicates the methodology that underpins the investigation of these two

research questions.

CHL proficiency as

linguistic capital

Four forms of

capital

Chineseness

qua habitus

Fields

106

107

Chapter 4: Methodology

This chapter explains the methodological pluralism that underpins the research methodology,

details the research design of the current study, and addresses ethical considerations. Section

4.1 presents an introduction to the philosophical position that combines a positivist ontology

and a constructivist epistemology. It talks back to the informing theory and guides the

research design. Section 4.2 argues the importance of mixed methods design, which enhances

the scope and depth of the study and increases the validity and reliability of the findings. This

section also emphasises the theoretical and practical appropriateness of mixed methods to the

study. Section 4.3 delineates the rationale of the quantitative phase and argues its function in

terms of operationalising the theoretical framework to address RQ1: Is CHL proficiency of

young Chinese Australian adults influenced by their investment of capital, the strength of

their habitus of ‘Chineseness’, or both? The design of the pilot study and the main study are

then presented. Section 4.4 describes the design for the qualitative phase, which functions as

a complement to the initial quantitative phase. The qualitative phase addresses RQ2: How do

young Chinese Australian adults understand their CHL learning in relation to (potential)

profits produced by this linguistic capital in given fields? Section 4.5 addresses ethical

considerations in this project. This chapter concludes with a brief summary in Section 4.6.

4.1 Philosophical position

The positivist ontology is arguably the most prevalent research paradigm in most natural and

many social sciences since the 1930s. Although it considers reality to be independent of the

observer (Burrell & Morgan, 1979), it does not necessarily exclude the input of human minds.

As Albert Einstein argued, physical concepts are the creation of human minds and are not

108

uniquely determined by the external world (Von Glasersfeld, 2001). In this sense, human

beings sometimes describe their world of reality according to their paradigmatic lenses,

taking the position that both the external world and their minds contribute to reality. This

position is crucial. It posits that, firstly, there is a real world independent of human beings’

knowledge and consciousness; and secondly, there is a dimension that includes human beings’

socially constructed knowledge about reality and its effects. This philosophical position

indicates the existence of a real world independent of human beings’ perceptions, theories,

and constructions; and concomitantly assumes that understanding of this real world is

inevitably a construction from human beings’ own perspectives and standpoint (Creswell &

Clark, 2011).

The current study draws on the philosophical position that combines the positivist

ontology and the constructivist epistemology because this particular way of thinking

resonates with the Bourdieusian theoretical stance. Bourdieu’s theoretical approach is

intended to bypass or dissolve a plethora of oppositions and antinomies (Bourdieu, 1991). His

theory is a systematic attempt to move beyond a dichotomy between ‘objectivism’ and

‘subjectivism’. By ‘objectivism’ Bourdieu means “an intellectual orientation to the social

world which seeks to construct the objective relations which structure practices and

representations” (Bourdieu, 1991, p. 11), “independent of individual consciousness and wills”

(Bourdieu, 1981, p. 87). Objectivism presupposes a break with subjective experience. It

attempts to elucidate the structures and principles upon which subjective experience is

dependent but which it cannot grasp. In contrast, by ‘subjectivism’ Bourdieu means “an

intellectual orientation to the social world which seeks to grasp the way the world appears to

the individuals who are situated within it” (Bourdieu, 1991, p. 11). Subjectivism presupposes

the possibility of the comprehension of the lived experience that is a form of knowledge

about the social world where subjective lived experience is central to comprehension and

109

knowledge. Traditionally, objectivism and subjectivism have been opposed to each other.

However, “the coexistence of two opposing truths defines the full truth” of the social world

(Bourdieu, 1981, p. 89). This Bourdieusian dualist perspective indicates that neither

objectivism nor subjectivism alone is an adequate intellectual orientation (Bourdieu, 1991).

For Bourdieu, subjectivism concentrates too much on individual experience and perceptions

of the social world and overlooks the power of structures within the social world, while

objectivism refuses to take account of individual capacity, relegating and shackling

individuals to objective relations of social structure. Therefore, Bourdieu chooses the term

“structuralist constructivism” or “constructivist structuralism” to stress the dialectical

articulation of the two moments (objectivism and subjectivism) in his theory (Bourdieu,

1989b, p. 14; Bourdieu & Wacquant, 1992, p. 11).

One of the ways that Bourdieu thinks through objectivism and subjectivism is his

conceptualisation of ‘agency’ and ‘structure’. ‘Agency’ refers to the capacity of individuals to

act independently and to make their own free choices (Barker, 2005). By contrast, ‘structure’

refers to the recurrent patterned arrangements which seem to influence or limit the choices

and opportunities that individuals possess (Barker, 2005). Social structure establishes

objective regularities, laws, mechanisms, and systems that are capable of guiding or

constraining agents’ practices and representations. At the same time, social structure does not

exist independently of agents’ consciousness and wills. Their subjective agency shapes their

worldview and seeks to grasp the way that the social world appears to them. Consequently,

agency presupposes the possibility of the comprehension of agents’ lived experience that is a

form of knowledge about the structures within the social world. This intellectual orientation

produces what agents come to know as the reality of the social world.

Bourdieu’s work on the conceptualisation of capital, habitus, and field attempts to

reconcile agency and structure, rather than conflate the two (Reay, 1995). In the Bourdieusian

110

nexus of capital, habitus, and field, agents are socialised by an evolving set of roles and

relationships in social domains (fields), where their dispositions (habitus) and various

resources (capital) are at stake. Agents’ capital and habitus are valued and re-valued by the

field and their agency is structured by the field. At the same time, various agents struggle for

the transformation or preservation of the structure of the field (Bourdieu, 1998). In other

words, fields are structured and restructured by the agents who operate their habitus and

capital within the fields. Thus, the field is a “field of forces”, which imposes on agents and

agents in turn conserve or transform the structure of the field by their agency (Bourdieu, 1998,

p. 32). Bourdieu theorises the necessary relations between the concepts of capital and habitus

in the existing fields in which capital and habitus are contextualised. This relational concern

constructs a theoretical foundation for the current study to examine how Chinese Australians’

‘Chineseness’ qua habitus and their various resources as capital are related to their CHL

learning in their lived social fields. As agents, Chinese Australians inhabit the social world

and therefore tend to adjust themselves to the structure of the social world. At the same time,

they create their social lives and generate the structure of the social world, so the social world

is continuously changing due to their dynamic agency. There is therefore a mutually

influential relationship where Chinese Australians shape society by virtue of their agency,

which in turn is structured by society.

4.2 Mixed methods research

Mixed methods research includes both qualitative and quantitative approaches in research

design, data production, and data analysis to answer research questions in a single study

(Tashakkori & Teddlie, 2003). There are contrasting views towards mixed methods research.

On the one hand, in the past there was considerable hostility between the quantitative camp

111

and the qualitative camp (Bourdieu, 1999c; Bourdieu & Wacquant, 1992). These two

approaches were viewed from purist perspectives as distinct and incompatible. Advocates of

quantitative approaches criticised qualitative approaches, considering them imprecise,

subjective, and unfit for making predictions. Alternatively, the qualitative side argued that

quantitative approaches are based on a naïve theory of objectivity and cannot describe the

complexity of social reality. In concert with these disputes, the appropriateness of combining

quantitative and qualitative methods within a single study was questioned (Guba & Lincoln,

1994; Tashakkori & Teddlie, 2003). On the other hand, Bourdieu (1999c) advises against the

opposition between quantitative methods and qualitative methods. He goes on to argue that

the traditional opposition conceals the fact that both quantitative and qualitative approaches

are based on social interaction, taking place within the constraints of social structures that

exert effects not only on the interactions amongst participants in a quantitative study, but also

on the interactions between the researcher and the researched in a qualitative study. In line

with this Bourdieusian perspective, an increasing number of scholars have articulated less

rigid views. Although the ‘either qualitative or quantitative’ perspective has been prevalent

for a long time, a mixture of ‘both qualitative and quantitative’ approaches has gained rapid

ground in recent years (Danermark, Ekström, Jakobsen, & Karlsson, 2002; Haralambos &

Holborn, 2004). There is a need to transcend the polarisation of methods in quantitative and

qualitative terms (Danermark et al., 2002; Luke, 2010), particularly in sociological research

(Sadovnik, 2011).

‘Methodological pluralism’ (Danermark et al., 2002) suggests that there is no

‘universal method’. Both quantitative and qualitative methods have their domains and

relevance. A particular method cannot be excluded a priori. The key issue is the ability to

judge the strength and the weakness, respectively, of a method for a given research problem.

Qualitative and quantitative methods are not mutually exclusive but complementary research

112

strategies. There is value in multimethodological approaches and it is profitable to combine

methods in practical work.

Following the route of ‘methodological pluralism’, the current study adopted a mixed

methods design. The choice of this design was governed, on the one hand, by the research

questions; and on the other hand, by what can be learnt from social complexity with the help

of different methods. It will be recalled that RQ1 asks about general relationships, tendencies,

and predictions, while RQ2 asks about individual perceptions and meanings of each

participant. The mixed methods design can gain a broader and deeper understanding of

different levels of research problems and better facilitate the examination of complex

phenomena than would be possible through a single method of research. Therefore, mixed

methods research has gained increasing popularity and appeared more frequently in a wide

range of social science disciplines (Mactavish & Schleien, 2000). The current project takes

account of both theoretical and practical considerations in its mixed methods design.

From a theoretical perspective, the study draws heavily on Bourdieusian theory.

Bourdieu’s logic of research is “inseparably empirical and theoretical” because “one cannot

think well except in and through theoretically constructed empirical cases” (Bourdieu &

Wacquant, 1992, p. 160). The goal of such research is “to grasp the particularity within the

generality and the generality within the particularity” (Bourdieu & Wacquant, 1992, p. 75)

and to uncover “the universal buried deep within the most particular” (Bourdieu & Wacquant,

1992, p. 44). By ‘generality’, Bourdieu refers to the kinds, facts, and patterns sought by

quantitative approaches; by ‘particularity’, Bourdieu means the contextualised experiences of

particular subjects sought by qualitative approaches. This viewpoint demonstrates Bourdieu’s

attempt to think through the split between quantitative and qualitative positions (J. Webb et

al., 2002). Bourdieu has shown the way for a kind of research that manages to combine both a

quantitative and a qualitative methodology (Kramsch, 2008; Thomson, 2008). He argues that

113

“statistics are bound to be abstract” but with the complement of interviews and texts,

profound phenomenon can be revealed “even more clearly” (Bourdieu, 1996, p. 174). In the

operationalisation of theoretical concepts such as cultural capital and social capital, there is a

need for a combination of various methods (Vryonides, 2007). Moreover, a fuller

understanding of habitus and how it works in identity construction would benefit from the use

of both qualitative and quantitative methods (Connolly, 2011). Thus, there is theoretical

support for the current study to adopt a mixed methods design to investigate the research

problems.

From a practical point of view, a combination of qualitative and quantitative research

strategies enhances the depth and scope of the investigation. ‘Numbers’ not only add

precision to ‘words’ but can also justify or falsify them. Simultaneously, ‘words’ can add

meaning to ‘numbers’ and explain them. The research questions addressed by the current

study are multi-faceted and the phenomena investigated by the study are complicated. The

complicated linguistic histories, profiles, and needs of CHLLs, their diverse language

learning and socialisation processes and outcomes, and the benefits and challenges of

developing their CHL proficiency call for a variety of methods, across the quantitative and

qualitative spectrum (He, 2006). These complex interactions might well demand a mix of

methods. Proper application of mixed methods in the study not only addresses this

complexity but also supports stronger scientific inferences than would be the case were either

qualitative or quantitative research tools to be applied in isolation. The mixed methods design

makes good use of the strengths of both quantitative and qualitative approaches, while the

weaknesses of each approach are potentially minimised. Practically, the mixed methods

design enables the study to depict a more complete portrait of Chinese Australians,

embracing how they negotiate their habitus of ‘Chineseness’ and capitalise on various forms

of capital through learning CHL in Australia.

114

An explanatory mixed methods design was adopted in the current study. The design

consisted of first producing quantitative data to test and explain the theoretical framework

and then producing qualitative data to complement and enrich the quantitative results. This

design can enhance confidence in the trustworthiness, credibility, and reliability of the

research findings, and add depth and scope to the project. An explanatory mixed methods

design can be especially helpful when unexpected results arise from the initial quantitative

phase. The subsequent qualitative phase can examine and interpret these results in detail.

4.3 Quantitative phase

Quantitative research is a type of approach that frames what is to be studied, asks specific and

narrow research questions, collects quantifiable data, analyses the data by using statistics, and

conducts the inquiry in an unbiased and objective manner (Creswell, 2008). In line with these

traits, a quantitative study addresses research questions that seek to establish relationships or

variance. In order to generalise this relationship or variance, a quantitative study provides a

description of trends, attitudes, or opinions of a population by investigating a sample of that

population.

In a quantitative study the literature functions to identify key issues related to the

research and document the importance of these issues (Creswell, 2008). It will be recalled

that the existing knowledge in the literature argues the interdependence amongst the ethnic

identity that CHLLs carry, the various forms of resources that they accrue, and the CHL

proficiency that they achieve. In light of Bourdieu’s key concepts of habitus, capital, and field,

the interaction of these dimensions was articulated in the theoretical framework. Quantitative

research can verify or reject this framework deduced from existing knowledge and theories. If

‘habitus’ and ‘capital’ do represent the internal agency operating in social fields and do

115

represent a set of dispositions and positions that are manifest, to some extent, in the

regulation of particular ways of thinking and behaving, these patterns should be discernible

and thus ultimately measurable to some degree. However, there is little documentation of a

comprehensive quantitative interrogation on the interactions amongst the habitus of

‘Chineseness’ of young Chinese Australian adults, various forms of capital accessible to them,

and their CHL learning practices within certain social fields. The initial quantitative phase

may add new knowledge through the analysis of the relatedness of these concepts.

A survey design has been selected. It offers a systematic method for gathering

information from a sample of entities for the purposes of constructing quantitative descriptors

of the attributes of the larger population of which the entities are members (Groves et al.,

2004). It is one of the most commonly used methods in quantitative studies to collect data to

test the theoretical models (Groves et al., 2004). In the current study, quantifiable data were

analysed by using statistical modelling to allow generalisations and claims to be articulated

about the population.

4.3.1 Research questions and hypotheses

In the quantitative phase, the main aim of the survey is to collect data to help understand the

relationship amongst young Chinese Australian adults’ ‘Chineseness’ qua habitus, their

various resources as capital, and their CHL learning as practice. Bourdieu (1989a) argues that

practice results from the relations between one’s dispositions (habitus) and one’s position in a

social world (capital), within the current state of play of that social arena (field); and that

people make choices about languages according to the amount of different resources (capital)

as well as the dispositions (habitus) they have within certain fields (Bourdieu, 1991). This

theoretical perspective informs RQ1: Is CHL proficiency of young Chinese Australian adults

116

influenced by their investment of capital, the strength of their habitus of ‘Chineseness’, or

both? Corresponding to this research question, the null hypotheses (H0) are:

H01. The habitus of ‘Chineseness’ of young Chinese Australian adults has no

statistically significant impact on their CHL proficiency.

H02. The investment of capital of young Chinese Australian adults has no


Deduced from the overall theoretical framework constructed in Chapter Three (see Figure

3.4), Figure 4.1 serves as the first part of the theoretical framework that drives examination of

RQ1 and hypothesis testing.

Figure 4.1. Theoretical framework guiding the quantitative phase

4.3.2 Target population

A population is a group of individuals who share particular characteristics pertinent to

research questions (Groves et al., 2004). As argued in Chapter One, this study identifies

young Chinese Australian adults as the population under investigation. A target population is

a group of individuals with some common defining characteristics that can be identified and

CHL proficiency as

linguistic capital

Four forms of

capital

Chineseness as

habitus

Fields

117

studied (Groves et al., 2004) and is therefore available to the research. This section reports on

the rationale for framing the target population of the initial quantitative phase.

According to the ABS 2011 Census, 94% of Chinese Australians and 97% of Chinese

Australians who spoke Chinese at home lived in the eight capital cities. It can be argued that

an overwhelming proportion of Chinese Australians live in urban areas. As such, the

quantitative phase targeted young Chinese Australian adults in urban Australia.

Identity develops over time, beginning in childhood, through a particularly salient

process during adolescence and young adulthood (Oh & Fuligni, 2010; Phinney & Ong,

2007). Levinson (1978) called this life period the “novice phase” and argued that the

overriding task of this phase is to move into the adult world and build a life structure (pp.

322-323). During this phase, or specifically the age range between 17 and 33, according to

Levinson, young adults experience a considerable amount of change and instability while

sorting through various possibilities in their social lives. In line with these arguments, the

quantitative phase targeted young Chinese Australian adults, ranging in age between 18 and

35. If born outside Australia, they had to have moved to Australia before the age of 13, as

children below 13 are considered less shaped by their learning experiences (Bhatti, 2002). In

this respect, their language learning experiences in Australia would be more salient in their

lives, and they would have been educated primarily in English, with Chinese becoming their

HL. The Australian-born and the non-Australian-born groups can be included within the same

target population frame because they share many linguistic, cultural, and developmental

experiences (J. Zhang, 2009). They are both CHLLs, as people who immigrated at an early

age and were primarily educated in English in their host country, or children born in the

English speaking country but exposed to their immigrant parents’ or ancestors’ native

language in a community, are likely to become HLLs (Cho, 2000; Cho et al., 1997). In short,

the target population of the quantitative phase refers to young Chinese Australian adults in

118

urban Australia, ranging in age between 18 and 35. If born outside Australia, they had to have

moved to Australia before the age of 13.

This target population consists of a diverse group of people in terms of their citizenship

status, the languages that they speak, and their countries of origins. They could hold

Australian citizenship, dual citizenship, or Australian permanent residency. They may or may

not be able to speak a Chinese language. They could be born in the Chinese Mainland, Hong

Kong, Macau, or Taiwan. They could be born in other Asian countries where Chinese is

widely spoken, such as Singapore, Malaysia, Indonesia, or Vietnam. They could also be born

in Australia or any other Western country.

4.3.3 Instrument design and development

In a survey design, a questionnaire is a form of instrument that participants complete and

return to the researcher (Creswell, 2008). It is widely used because it has several advantages

(Haralambos & Holborn, 2004). Firstly, data analysis is time effective. Data produced by

questionnaires can be analysed in an effective way, using computer software packages.

Secondly, questionnaires are familiar to most people and they generally do not make people

apprehensive. Nearly everyone has had some experience completing questionnaires. Finally,

questionnaires are considered less intrusive. The respondents are free to complete the

questionnaire in their own time. Based on the above advantages, a questionnaire was

designed and applied in the initial quantitative phase.

Online questionnaires are becoming increasingly popular in quantitative research

(Groves et al., 2004). This is especially relevant for the current study conducted in urban

areas, given the widespread use and ease of access to the internet. The technology enables

electronic data collection and data entry. E-based data production is less time consuming and

119

increases efficiency. Data entry errors are reduced because data entry occurs automatically

and technologically rather than manually. Moreover, the online questionnaire was considered

cost-effective for the current sole-researcher and unfunded study involving a large sample

size and a scattered population across varied geographic areas. Therefore, an online

questionnaire was conducted in the quantitative phase by using the survey generation

software, Key Survey.

The questionnaire for the survey was structured as follows. A participation statement

that outlined the project and explained ethical considerations was presented at the very

beginning of the questionnaire. In this way, the researcher introduced the proposed research

to the participants and obtained their informed consent for participation. At the end of the

questionnaire, participants were asked to choose whether they would like to participate in an

interview for the subsequent qualitative phase. Participants who were interested were asked

to leave their names, email addresses, or phone numbers for further contact. An

acknowledgement was presented to the participants upon completion of the questionnaire.

There were four sections in the main body of the questionnaire. The first section collected

demographic and background information from the participants. The next three sections

measured participants’ attitudes towards attributes often associated with ‘Chineseness’,

various resources they invested in CHL learning, and their perceptions of their CHL

proficiency respectively. Accordingly, ‘Chineseness’ qua habitus, four forms of capital,

namely economic, cultural, social, and symbolic capital, and CHL proficiency were treated as

variables in the quantitative phase, reflected by their corresponding questionnaire items. An

overview of the questionnaire items will be presented in Chapter Five.

To measure items mapping to these variables, a 7-point Likert-type scale was used. The

major statistical models applied to the data analysis in the main study were Factor Analysis,

Structural Equation Modelling (SEM), and Regression. These models require the variables to

120

be measured at the interval level at least, whilst a Likert-type scale is ordinal in nature

(Creswell, 2008). However, the errors for treating the Likert-type scale results as interval data

are minimal especially when there are more categories or choices in this scale and the data

exhibits a symmetrical distribution (Binder, 1984; Zumbo & Zimmerman, 1993). Therefore, a

7-point Likert-type scale is widely accepted as a proxy interval level of measurement in line

with common practice in educational research (Lehman, 1991; Tabachnick & Fidell, 2007).

When there is an existing questionnaire testing the variables investigated in a given

study, it is good practice to adapt the questionnaire to the current context if the original

questionnaire has proven reliable and valid in certain research contexts. Self-reporting is a

commonly used approach to measure language proficiency in large-scale questionnaire

studies where the direct testing of language proficiency is difficult (Phinney, Romero, Nava,

& Huang, 2001). Moreover, self-reporting measures of language proficiency have been found

to correlate highly with direct measures of language ability (Oh & Fuligni, 2010). Chinen and

Tucker’s (2005) ‘can do’ questionnaire was originally used to gauge the self-reported

language proficiency of Japanese HLLs. After permission from the authors was obtained, this

instrument was modified to meet the requirement of the quantitative phase and to measure

participants’ perceptions of their CHL listening, speaking, reading, and writing skills.

When there is no existing standardised questionnaire specifically designed to plumb the

variables of the study, items to measure these variables need to be designed. This is the case

for Chinese Australians’ ‘Chineseness’ qua habitus and possession and investment of various

resources as capital. Therefore, an instrument to measure these concepts was developed. This

was a challenging process consisting of planning, constructing, evaluating, and validating.

Detailed steps and strategies were undertaken as follows.

The theoretical ideas underpinning each concept need to be translated into items within

the questionnaire, which makes it possible to take measurements relating to those abstract

121

concepts (Danermark et al., 2002). This process is called ‘operationalisation’.

Operationalisation is an important stage in quantitative studies. This involves breaking the

theoretical concept down into various components or dimensions in order to specify exactly

what is to be measured. The relations between concepts, formulated in the theory, are then

studied in the form of quantitative relationships between variables (Danermark et al., 2002).

Chinese Australians’ ‘Chineseness’ qua habitus, their various resources as capital, and their

CHL proficiency were regarded as ‘latent variables’ in the quantitative analysis. A ‘latent

variable’ refers to a variable that is not directly measurable but related to several variables

that can be measured directly (Field, 2009). These directly measurable variables are called

‘indicators’ or ‘measures’ and can be operationalised through corresponding questionnaire

items that provide quantifiable data for measuring each theoretical concept. As required by

SEM, a minimum number of indicators are needed to represent a latent variable. There is no

upper limit in terms of the indicator numbers. But too many indicators make it difficult to fit

a model to data (Bentler, 1990). As a practical matter, there is a preferred minimum number

of indicators, and in most cases, four is the absolute minimum required for model fit testing

with five to seven indicators considered to adequately reflect most latent variables (R. Ho,

2006). A good measurement needs to be accurate and consistent. Otherwise, the findings will

not be valid and reliable. Therefore, operationalisation requires attention to two core

methodological concerns, namely ‘reliability’ and ‘validity’.

‘Reliability’ refers to the stability and consistency of a measurement (Creswell, 2008).

‘Test-retest reliability’ and ‘inter-item reliability’ are two commonly used measures of

reliability (Hair, Black, Babin, Anderson, & Tatham, 2006). ‘Test-retest reliability’ is the

expectation that there will not be different findings each time the measures are used,

assuming that nothing has changed in what is being measured. For practical reasons, the

quantitative phase will not use ‘test-retest reliability’. The second and more commonly used

122

measure of reliability is ‘inter-item reliability’, which evaluates whether a set of indicators

coheres and maps to one underlying latent variable. Often reliability cannot be determined

until the data are collected and analysed. A pilot study can however be used to generate some

top-level indicators of reliability. Since the variables of the quantitative phase were

considered to be measured at the interval level, the pilot study used Cronbach’s alpha

coefficient to test ‘inter-item reliability’ (R. Ho, 2006).

Reliability does not guarantee validity. It is a necessary but not sufficient condition of

validity (Hair et al., 2006). ‘Validity’ here refers to accuracy and whether the

operationalisation is correctly indicating what it is supposed to (Creswell, 2008, 2009). It is

the extent to which the measurement relates to the underlying latent variable; that is, it argues

whether the data make sense, and are meaningful. This enables the researcher to draw good

conclusions from the sample to the population. ‘Face validity’ indicates whether the

questionnaire items seem to ask what they are supposed to and seem to be getting the desired

results (Nardi, 2006). It indicates the extent to which the content of the items is consistent

with the construct definition, based solely on researcher’s judgment, experts’ advice, and

research literature (Hair et al., 2006). As such, face validity concerns the acceptability of the

test items. This is a legitimate, but not very rigorous, way of assessing validity. ‘Construct

validity’ signals how well a set of questionnaire items as indicators are measuring the

complexity of a variable, that is, a concept being studied (Nardi, 2006). It indicates the extent

to which a set of indicators reflects the theoretical constructs specified as latent variables that

those indicators are designed to measure (Hair et al., 2006).

Satisfying validity requires an in-depth review of literature to examine how the concept

is defined and described from a theoretical point of view. To this end, when designing the

measurements of habitus of ‘Chineseness’ and various forms of capital, literature about

Confucianism and Bourdieu’s interpretation of capital and habitus were reviewed. The key

123

dimensions within the theoretical concepts were identified. This careful consideration of the

literature also revealed how the concepts of habitus and capital have been operationalised in

previous empirical studies. A detailed review follows.

Habitus is a nebulous concept. Accordingly, it is difficult to measure in empirical

research (Sullivan, 2002). Despite this challenge, existing studies have attempted to

operationalise it. Habitus can explain how individuals formulate their expectations and beliefs

(Dumais, 2002; McClelland, 1990). Following this perspective, both Dumais (2002) and

McClelland (1990) operationalised habitus with a set of measures of occupational aspirations.

More relevant to the current study, Cockerham and Hinote (2009) suggested that measures of

collectivities can be especially useful to quantify habitus, where collectivities refer to

collections of agents linked together through particular social relationships and shared norms,

values, ideals, and social perspectives (Cockerham, 2005). In particular, religion and ideology

are examples of collective perspectives (Cockerham, 2005). Ho and colleagues (2012) have

made an attempt to compile a pool of items to measure Chinese collective perspectives by

consulting Confucian classics and sayings associated with Confucianism. Cockerham, Ho,

and others have informed the operationalisation of the habitus of ‘Chineseness’. As argued in

Chapter Three, Confucianism can be understood as a Chinese quasi-religious and ideological

system. Key dimensions of Confucian norms, values, ideals, and social perspectives

documented in Confucian classics and popular classical sayings associated with

Confucianism were used as measures to quantify habitus of ‘Chineseness’.

According to Bourdieu (1990), “the distribution of capital in its different kinds among

the individuals” is quantifiable (p. 135). Following this advice, attempts have been made to

quantify capital. Highest personal educational attainment (Bourdieu, 1973, 1989a; Veenstra,

2009), attendance at high cultural events (such as concerts, galleries, and museums) (De

Graaf & De Graaf, 2000; DiMaggio, 1982; Dumais, 2002; Marks, 2009; Sullivan, 2001;

124

Yamamoto & Brinton, 2010), and ownership of cultural or educational objects (Roscigno &

Ainsworth-Darnell, 1999; Sullivan, 2001; Yamamoto & Brinton, 2010) could be used as

indicators of the presence of cultural capital. Participation in different kinds of social

associations and frequency of contacts with acquaintances, friends, and family members

could be used as indicators of presence of social capital (Veenstra, 2009).

A key issue in the achievement of valid and reliable operationalisation is the

construction of a well-written and manageable questionnaire. Frustrations and noncompliance

can result from an unclear, biased, wordy, or poorly designed questionnaire. Several strategies

to minimise these dangers are reported below.

Firstly, not all questions are necessarily applicable to all participants. Filter questions

are asked to avoid forcing participants to provide opinions when they do not have any

relevance. These are also called branching or contingency questions (Nardi, 2006). For

example, before asking the question “except Australia and your country of origin, where else

have you lived and for how long”, a filter question of “have you ever lived in a country other

than Australia and your country of origin” was asked first. Similarly, before asking “how long

have you formally studied Chinese”, “have you ever formally studied Chinese” was first

asked. Likewise, “have you ever visited China” was asked before asking “how many times

have you visited China”. This filtering technique ensured that branching or contingency

questions were answered only by those who responded in a relevant way to a previous item.

Participants were filtered towards questions applicable only to them.

Secondly, sometimes the answers to a question depend on whose opinions are under

scrutiny. To ask whether the respondents think “family members should live close to one

another”, different results could be prompted by asking “I always prefer to live close to my

family members”. Similarly, to ask “people always prefer to say yes to save face”, different

results could be prompted by asking “to save face I always prefer to say yes”. In the above

125

examples, the original wording was more about a general belief, but the second wording put

the spotlight on the participants’ own personal views. For another example, one original item

was “I am very popular among my Chinese peers”. This item directly scrutinised participants’

own opinions about how popular they were and could push those participants who were

popular to say so, even though they could normally feel uncomfortable about saying this.

This item was reworded as “people consider me very popular among my Chinese peers”. This

item scrutinised participants’ popularity through the eyes of others to avoid any personal

discomfort of self-reporting on this.

The third strategy is to avoid ‘double-barrelled’ questions that actually measure two

things within one item. Such questions often contain the word ‘and’. For example, an original

item “I always keep up to date with current Chinese affairs by watching TV, listening to the

radio, reading newspaper, and surfing online” was replaced by the item “I always keep up to

date with current Chinese affairs by watching TV, listening to the radio, reading newspaper,

or surfing online”. The original item was difficult to answer for someone who used only one

of the media mentioned not the rest. For another example, to measure whether participants’

Chinese language competency was valued, the original item was “I am well regarded because

of my Chinese language competency”. Further consideration revealed that it actually

contained two questions, namely “are you well regarded or not?” and “what is the level of

your Chinese language competency?” People may be well regarded due to their Chinese

language proficiency or something else, and people may not be well regarded due to their

Chinese language proficiency or something else. It was hard to tell whether people’s Chinese

language proficiency was valued. The original item was then simply reworded as “my

Chinese competency is well regarded”.

Fourthly, enough variance in responses is important for most commonly used statistical

analyses. A split strategy was used to ‘push’ participants to choose different answers in order

126

to get enough variance. Adverbs such as ‘very’, ‘really’, ‘completely’, ‘easily’, and ‘always’

as well as auxiliary verbs such as ‘must’, ‘have to’, and ‘do’ were useful in this regard. By

using these words, participants could be split between those who really agreed and those who

agreed but not that much.

Fifthly, negatives in a statement may lead to errors in responses. It becomes hard for

participants to know whether agreeing with a negatively worded statement might actually

mean they are disagreeing with it. Therefore, negatively worded questions should be restated

or reworded to eliminate this complexity. For example, the original item “I won’t have a good

time if I don’t celebrate Chinese festivals” was first replaced by “I will have a good time if I

celebrate Chinese festivals”. However, the entire sample might be happy to celebrate Chinese

festivals due to the popularity of Chinese culture in Australian society. Therefore, the

participants’ responses may not offer enough variance. Eventually, the item was reworded as

“I feel very sad if I miss celebrating Chinese festivals”.

Finally, jargon needs to be eliminated and replaced by words familiar to all participants.

It is also of great value to work with colleagues and test alternative readings of draft

questions. It is beneficial to check the self-designed questionnaire with individuals

knowledgeable in the topic before using it in the survey. Numerous meetings and discussions

with the supervisory team and experts were held. They have either theoretical knowledge of

the concepts to be operationalised or experience in instrument design and development.

Through meetings and discussions with them, the quality of the instrument improved.

4.3.4 Pilot study

A good research strategy requires careful planning and a pilot study will often be a part of

this strategy. A ‘pilot study’ is a small-scale preliminary study conducted before the main

127

research in order to check the feasibility or to improve the research design (Haralambos &

Holborn, 2004). It is also a trial of a particular research instrument (T. L. Baker, 1994) to

pre-test whether the proposed instrument is appropriate. A pilot study can reveal deficiencies

in the initial design. These can then be addressed before time and resources are expended on a

main study. A pilot study was designed to test logistics and operationalisation in order to

improve the quality and efficiency of the main study. Part of the questionnaire was adapted

from the existing instrument in the literature and part of the questionnaire was designed and

developed by the researcher. It was considered critical to evaluate the feasibility of the

instrument before it could be applied in the main study.

The benefits of the pilot study and its evaluation are two-fold. First of all, face validity

and the appropriateness of the language used for the questionnaire were assessed, according

to the feedback from participants in the pilot study. Ambiguous wording or phrasing was

clarified. Any distress or uncertainty inadvertently elicited by sensitive questions was

decreased to a minimal level by using more comfortable language. Redundant questions were

either deleted or integrated into questions with more succinct language. Inadequacies in the

instructions were also identified. Secondly, reliability testing included in the pilot study is a

prerequisite for reliable findings. Cronbach’s alpha estimates the internal consistency of a set

of items and can be used as a measure of reliability. Though the alpha value is item number

and sample size dependent, an alpha level higher than .80 indicates that items are reliable and

the set of questions is internally consistent (Kline, 1999). An alpha level lower than .80

indicates that there may be problems with internal consistency. This procedure represents a

refinement by identifying problematic items in the test. Though deleting these items from the

questionnaire may affect the validity, it will increase the internal consistency of the

instrument. Apart from the alpha value, corrected item total correlation is another relevant

measure. Its value lower than .33 indicates that less than 10% of the variance in the scale is

128

accounted for by that corresponding item (R. Ho, 2006) and that this item is of concern. It is

also useful to determine whether Cronbach’s Alpha value increases if a particular item is

deleted. In short, the pilot study serves to plumb the validity and reliability of the instrument.

A pilot study is normally small in comparison with the main study. It requires a smaller

number of participants in contrast to the main study. The sample size for Cronbach’s

coefficient alpha test could be as small as 30 (Duhachek, Coughlan, & Iacobucci, 2005;

Yurdugul, 2008). Convenience sampling is widely used to get an inexpensive approximation

of the truth (Walonick, 1993). As the name implies, the sample is selected because it is

convenient for the researcher to approach participants. The participants are often known to

the researcher and are willing and available to be studied. This non-probability method is

often used in research to get a gross estimate of results, without incurring the cost or effort

required to select a random sample. Therefore, 38 participants were approached by

convenience sampling for the pilot study. The participants in the pilot study were excluded

from the final sample of the main study, because their participation in the pilot study could

have influenced their behaviour in the main study (Haralambos & Holborn, 2004).

4.3.5 Main study

Sample size

Different statistical models have different rules to identity the optimal sample size required to

produce reliable results. SEM is the major statistical model applied to the data analysis in the

quantitative phase. SEM in general requires a large sample size, which varies as a function of

choice and value of fit index, number of variables / degrees of freedom, relation among

variables, and statistical power (Kim, 2005). Consulting the algorithm developed for sample

size planning in SEM (Kelley & Lai, 2011; Kim, 2005) and the conventional minimal sample

129

size required for SEM (Bentler & Chou, 1987), this study was conducted with the intent of

gaining 200 complete cases as the sample size for SEM.

Multiple regression models were also constructed in this study. A rule-of-thumb that

provides a rationale to estimate sample size for regression analysis is proposed by Green

(1991). When significance level and power value were set at .05 and .80 as appropriate for a

wide range of behavioural research (Cohen, 1992; Hair et al., 2006), the functional

relationship between sample size (N) and effect size (f2 or R

2) as well as the number of

predictors (k) was established (Green, 1991). Although Cohen (1988) argues that the choice

of values for effect size (f2 or R

2) should depend on the research area, he proposes, as a

convention, f2 values of .02, .15, and .35 to serve as operational values for small, medium,

and large effect size respectively, when the corresponding R2 values are .02, .13, and .26

(Cohen, 1988). The functional equation is shown below:

N ≥ L / f2, where f

2 = R

2 / (1 - R

2) and L = 6.4 + 1. 65 × k - .05 × k

2

In the regression analysis, there were two predictors, namely ‘capital’ and ‘Chineseness’.

Accordingly, L valued at 9.5 (L = 6.4 + 1. 65 × 2 - .05 × 22 = 9.5). For a medium effect size

of .15, the minimum sample size for the regression analysis was 63 (N ≥ 9.5 / .15 = 63). This

value was verified by using the software G-Power. When significance level, power value, and

effect size were set at .05, .80, and .15, G-Power reported the calculated sample size of 68 as

the minimum sample size required by the regression analysis, which was very close to the

manually calculated value 63. Taking account of the minimal sample sizes required by SEM

and multiple regression, the study sought at least 200 complete responses in order to produce

reliable results.

130

Sampling

It is characteristic of most sampling techniques in the quantitative approach that the aim is to

provide information about the total population. The point of sampling techniques is that

conclusions regarding the total population and not just the sample can be drawn with a

recognised degree of uncertainty (Danermark et al., 2002). Therefore, it is imperative to

sample in a careful and explicit manner.

Every sampling method has its strengths and weaknesses. The selection of sampling

method is contingent upon the desired sample size, the characteristics of the target population,

and the availability of participants (Creswell, 2008). Simple random sampling and systematic

sampling could not be used in the current study because it was impossible to identify each

member in the target population. Stratified sampling and cluster sampling are often used

when there is a target population that is large and easy to access. Though Chinese constitute

the third largest group among all immigrants in Australia, just behind people from the United

Kingdom and New Zealand, Chinese Australians only account for 4.3% of the Australian

national population (ABS 2011 Census) and they are a numerical minority. This increases the

difficulty of sampling because the smaller the population, the harder it becomes to approach

the participants to represent that population. As Groves et al. (2004) argues, what makes a

population rare sometimes is not its absolute size but its size relative to other coexisting

populations. In this regard, stratified sampling and cluster sampling were not possible.

Snowball sampling avoids the challenges that applying other sampling methods raises

in the real world. It occurs when the researcher asks participants to recommend other

individuals to participate in the study. It is commonly used when the participants are rare and

thus difficult to access or identify (Nardi, 2006), for example, in heritage research (Gibbs &

Hines, 1992; Hall, 1992; Kiang, 2008; Pao, Wong, & Teuben-Rowe, 1997; Root, 1992).

Therefore, snowball sampling was used to approach participants in the current study.

131

To ‘get the snowball rolling’, a request was posed in the participation statement of the

online questionnaire and through informal conversation with individuals through personal

networks, asking participants to invite other participants whom they know and meet the

participation requirements. This led to sampling of individuals who had not been anticipated

when the project began because the snowball sampling proceeded after the study began.

Although the research started with a relatively short list of participants, the list grew like a

snowball as names were added through the referral of participants. Eventually, snowball

sampling netted 230 respondents. A detailed description of these participants is presented in

Chapter Five.

Data analysis

The instrument for the main study was the refined and modified version of the original online

questionnaire trialled in the pilot study. Responses to the online questionnaire provided by the

participants were transferred directly into SPSS to produce the raw data set. Given this

electronic process of data production, data entry error should not be a concern. In the raw

data set, each row recorded all the variables corresponding to each participant. Each line

recorded each variable across all the participants. The production of single-factor congeneric

measurement models and structural models investigating the relationships amongst constructs

is pivotal to successful analysis. To this end, SEM was the major statistical technique used in

this study. IBM SPSS AMOS 21 was used to run SEM. The advantages of SEM are reported

below.

First and foremost, SEM is theory-driven. Bourdieu’s theory of habitus, capital, and

field (Bourdieu, 1989a) argues that practice results from the relations between one’s

dispositions (habitus) and one’s position in a social world (capital), within the current state of

132

play of that social arena (field). Specifically, Bourdieu (1991) theorises that people make

choices about language according to the amount of different resources (capital) as well as the

dispositions (habitus) that they have within given social fields. As argued in Chapter Three,

this study extended Bourdieu’s explication of agents’ language choices within a particular

language to agents’ choices from different languages. SEM was applied to test the complex

entanglement amongst the habitus of ‘Chineseness’ of young Chinese Australian adults,

various forms of capital available to them, and their CHL proficiency resulting from their

CHL practice. SEM provides the justification for the specification of this theoretical

dependence relationship. The theory-based approach to SEM is a distinct strength of this

technique, as it entails a mode of thinking that specifies the theoretical framework more

exactly, tests the theory more precisely, and yields a more thorough understanding of the data

(R. Ho, 2006).

Secondly, SEM has the capability to assess how well the indicators reflect their

respective latent variables by Confirmatory Factor Analysis (CFA) and how well the latent

variables are related in a series of dependence relationships between ‘exogenous variables’

and ‘endogenous variables’. An exogenous variable is one whose variability is assumed to be

determined by factors outside the model under consideration; an endogenous variable, on the

other hand, is one whose variability is to be explained by exogenous and other endogenous

variables within the current model (Hair et al., 2006). SEM can assess how well each

theoretical concept is operationalised by their corresponding questionnaire items and how

well these theoretical concepts are related as respective exogenous and endogenous variables.

Moreover, SEM improves statistical estimation by accounting for unique disturbances

in the estimation process (R. Ho, 2006). Many other statistical models assume that variables

in the analyses are error-free. However, concepts can seldom be measured perfectly, from

either theoretical or practical perspectives. This is either ascribed to external errors associated

133

with inaccurate responses by the participants, or attributed to internal errors associated with

imperfect operationalisation of the concepts. Consequently, measured variables usually

contain at least moderate amounts of error (R. Ho, 2006). However, in SEM, scores on the

indicators are used to develop estimates of their respective latent variables. As these estimates

are derived on the basis of the common or shared variance among the indicators, scores on

the latent variables are unaffected by measurement errors (R. Ho, 2006). Therefore,

measurement errors are removed and offset in SEM.

It is important to distinguish a reflective model from a formative model before an SEM

is applied because SEM applied to formative models can lead to large standard errors and

unstable parameter estimates. According to Coltman, Devinney, Midgley, and Venaik (2008),

reflective models are distinct from formative models in the following ways. In reflective

models, causality runs from the latent variable to its indicators. Causality runs in an opposite

direction in formative models. As a result of causality direction, variation of the latent

variable causes the variation of the indicators in reflective models, while it is the other way

around in formative models. In reflective models, the latent variable exists as a theoretical

construct independent of the indicators used to measure it. Indicators are represented by the

latent variable. In formative models, the latent variable is formed as a combination of its

indicators. Indicators define the latent variable. Consequently, indicators in reflective models

are interchangeable. Adding or dropping an indicator does not change the conceptual domain

of the theoretical construct. This is not the case in formative models. Moreover, indicators in

reflective models share a common theme and they should have high positive intercorrelations.

In contrast, this is not necessarily the case in formative models. Therefore, these two models

are conceptually, substantively, and psychometrically different. Bearing this in mind, the

design of the online questionnaire has taken careful consideration of the basis for the

construction of SEM. Questionnaire items did not function to form and define their

134

corresponding theoretical constructs. Rather, they were designed to reflect the attributes

associated with their corresponding theoretical constructs. As a result, variance of the existing

constructs caused that of the item measures. In line with the above perspectives, reflective

models were established and SEM applied to these reflective models was considered

appropriate for the current study.

Two critically important assumptions associated with SEM are: data are measured at an

interval level and have a multivariate normal distribution (Byrne, 2001). It has been argued

earlier that the data measured by a 7-point-Likert-type scale are widely accepted as proxy

interval level of measurement in practice, which satisfies the first assumption. However, most

data in practice fail to meet the assumption of normal distribution and the presence of

excessive kurtosis is particularly troublesome for the normality assumption (Byrne, 2001). As

a rule of thumb, if the value of skewness (kurtosis) falls beyond the range from minus twice

standard error of skewness (kurtosis) to plus twice standard error of skewness (kurtosis), the

univariate normal distribution assumption is considered to be violated (Field, 2009). That is,

to satisfy the univariate normal distribution assumption, the critical ratio between the value of

skewness (kurtosis) and that of the standard error of skewness (kurtosis) should be less than

|2|. Alternatively, a Kolmogorov- Smirnov Test can be used to test the normality. A significant

Kolmogorov-Smirnov value indicates the violation of the normality assumption. Even if all

univariate distributions are normal, the joint distributions of the variables may depart

substantially from multivariate normality (Stevens, 2002). Mardia’s coefficient was used as

an indicator of the degree of multivariate normality. A Mardia’s coefficient greater than 3

implies the violation of multivariate normality (Yuan, Marshall, & Bentler, 2002). One

approach to handling the presence of non-normal data is to use a procedure known as

‘bootstrapping’ (Byrne, 2001). The decision as to whether or not bootstrapping needs to be

used will be based on the degree of multivariate normality indicated by Mardia’s coefficient.

135

SEM is a multivariate technique that can be described as a combination of CFA and

path analysis. Correspondingly, SEM consists of two parts: the ‘measurement model’ and the

‘structural model’. The measurement model is concerned with the relations between

indicators and latent variables (R. Ho, 2006). It represents the degree to which the indicators

capture the essence of the latent variable (Hair et al., 2006). It specifies the relations between

a set of questionnaire items and the theoretical concepts these items were designed to

measure. This is achieved by CFA. The structural model is a flexible and comprehensive

model that specifies the pattern of relationships among latent variables, namely exogenous

and endogenous variables (R. Ho, 2006). This will be achieved by path analysis. Both models

were distinctive in the current study. The measurement model provided a test for validity and

reliability of the questionnaire items employed to measure the theoretical concepts. This

further improved the validity and reliability of the instrument after the pilot study. The

structural model offered a direct test of the theoretical framework and hypotheses, and led to

the answers to RQ1. The rationale for the initial measurement model and the subsequent

structural model are explicated below.

The proper specification of the measurement model is necessary before meaning can be

assigned to the analysis of the structural model. Therefore, a measurement model was firstly

specified. The indicator variables or the questionnaire items were assigned to the

corresponding latent variables or the theoretical constructs. The overall model fit of the

measurement model was assessed by CFA. Several approaches were used to check the overall

model fit. First of all, the chi-square goodness-of-fit statistic was computed. The chi-square

test assesses an exact null hypothesis that the data included in the model do not significantly

deviate from the population. A chi-square value corresponding to an insignificant p value

indicates that there is not enough statistical evidence to reject the null hypothesis, and

therefore demonstrates an acceptable model fit. However, chi-square value is sensitive to

136

sample size and increases as a direct function of sample size. In cases of large samples

needed by SEM, the power of the statistical test underlying the SEM approach is very high.

With a large sample size and a great deal of statistical power, almost every reasonable model

will be rejected if only the chi-square value and the associated p value are considered (R. Ho,

2006). Bentler (1990) advises against the sole use of the chi-square value in judging the

overall fit of the measurement model. To this end, absolute fit measures, such as RMSEA

(Root Mean Square Error of Approximation), and some incremental fit measures, such as CFI

(Comparative Fit Index), IFI (Incremental Fit Index), NFI (Normed Fit Index), RFI (Relative

Fit Index), and TLI (Tucker-Lewis Index) are also computed to augment the chi-square

goodness-of-fit. Among these, CFI and TLI are reportedly used most often (Hair et al., 2006).

Although there are no clearly established rules as to what constitutes a good fit, a widely

applied guideline suggests that the incremental fit measures should be higher than .90 to

indicate a good overall fit of measurement model (Bentler, 1990; Hair et al., 2006) and that

RMSEA values below 0.08 are deemed acceptable to represent the goodness-of-fit when the

proposed model is estimated in the population (Hair et al., 2006; R. Ho, 2006).

CFA also provides additional diagnostic information that may suggest modifications for

improving the model. These diagnostic cues include regression weights, standardised

residuals, and modification indices. To test whether each item in the questionnaire is

significantly represented by their respective latent variable, the unstandardised and the

standardised regression weights are calculated. The unstandardised regression weights are

checked by the critical ratio and its corresponding significance level. A significant (p < .05)

critical ratio indicates the corresponding item in the questionnaire is a significant indicator of

the respective latent variable, and vice versa. The standardised regression weights should be

at least .5 and preferably .7 to consider the corresponding item as a significant indicator of the

respective latent variable (Hair et al., 2006). Typically, standardised residuals less than |2.5|

137

do not suggest any problem (Hair et al., 2006). Conversely, residuals greater than |4.0| serve

as a warning and suggest a potentially unacceptable problem (Hair et al., 2006). Standardised

residuals between |2.5| and |4.0| deserve some attention but may not suggest any changes to

the model if no other problems are associated with these two items (Hair et al., 2006).

Modification indices demonstrate how much the chi-square value would be reduced.

Modification indices of greater than |4.0| suggest that the model fit would be improved by

allowing corresponding variables and error terms to correlate. However, given the strong

theoretical basis of SEM, changes to the measurement model cannot be solely based on the

diagnostics provided by CFA. The decision whether or not to implement these modifications

rests on the theoretical justification (R. Ho, 2006) along with those suggested diagnostics

(Hair et al., 2006). As mentioned earlier, SEM is a theory-based approach. Without strong

theoretical justification, employing the modification to improve model fit only capitalises on

the uniqueness of the data set but the results will most likely be atheoretical.

Traditional approaches to reporting reliability and validity are not easily transferred to

SEM because they do not take account of the congeneric nature of the model and assume that

the factor loadings are considered equal. As a result, traditional approaches underestimate

both reliability and validity measures (Graham, 2006). To offset the drawbacks of traditional

approaches, four measures of reliability, namely Squared Multiple Correlations (SMCs),

construct reliability, variance extracted, and coefficient H, as well as two measures of validity,

namely construct validity and convergent validity were reported. The rationale behind these

reliability and validity measures will be explained in Chapter Five.

After checking and improving the fit of the measurement model as well as confirming

the reliability and validity of the measurement model, the structural model that assesses the

relationships amongst latent variables can be constructed. The regression weights matrices

demonstrate how strong and significant the relationships amongst these latent variables are.

138

In this study, the structural model talked back to the theoretical framework, tested the two

hypotheses, and helped to answer RQ1. Attuned to the theoretical framework, the following

two structural models (see Figures 4.2 and 4.3) were constructed for hypothesis testing.

Straight arrows depict a dependence relationship, which means the impact of one variable on

another. In a measurement sense, dependence relationships exist from latent variables, or

theoretical constructs, to their indicators or the questionnaire items by which latent variables

are measured. In a structural sense, dependence relationships occur between latent variables,

with arrows flowing from the exogenous variable to the endogenous variable. In Figures 4.2

and 4.3, latent variables are represented by ovals while indicators are represented by squares.

‘I’ stands for ‘indicator’ or ‘item’ in the questionnaire and ‘e’ stands for measurement error or

more correctly ‘unique disturbance’.

Figure 4.2. Structural relationship between ‘Chineseness’ and ‘CHL proficiency’

Figure 4.3. Structural relationship between ‘capital’ and ‘CHL proficiency’

I1 I2 I3 …

e e e

Chineseness

I1 I2 I3 …

CHL proficiency

e e e

e

e

Cultural

capital

e

Capital

I1 I2 I3 …

e e e

CHL proficiency

Symbolic

capital

Social

capital

Economic

capital

e e e

139

4.4 Qualitative phase

The initial quantitative phase of this project sought general explanations and systematic

relations within the phenomena under investigation. Accordingly, the quantitative phase was

not and could not be concerned with the individual and exceptional case, as well as

participants’ detailed and deep perceptions and viewpoints. To tackle these dimensions of the

phenomena, the subsequent qualitative phase was designed to function as a complement of

the initial quantitative phase.

Qualitative research seeks to gain a deep holistic perspective and understanding of

people’s experiences by asking broad and general questions (Fraenkel & Wallen, 1996). The

basic principle of a qualitative study is that there are in-depth meanings embedded in people’s

experiences and that these meanings are mediated through researchers’ own perceptions

(Merriam, 1998). As such, qualitative research is concerned with the nature, explanation, and

understanding associated with people’s experiences, views, and beliefs. Based on the findings

of the quantitative analysis, more analysis working with qualitative data is needed to extend,

probe, and illustrate the results of the quantitative phase. The initial quantitative phase

examines the research problems of whether the habitus of ‘Chineseness’ of young Chinese

Australian adults and their investment of different forms of capital can produce their CHL

proficiency. Subsequently, the qualitative phase addresses RQ2: How do young Chinese

Australian adults understand their CHL learning in relation to (potential) profits produced by

this linguistic capital in given fields?

In order to answer this question, the qualitative phase adopted an interview approach. A

research interview is a purposive conversation (Lodico, Spaulding, & Voegtle, 2006) and an

interchange of views between the interviewer and the interviewee(s) on a topic of mutual

interest, which constructs sites for knowledge production (Kvale, 1996). It is the most

common data collection method in qualitative research, particularly for studies like the

140

current project, when intimate, repeated, and prolonged involvement in the life and

community of the participants is not possible (Lodico et al., 2006). Thus, interviews become

the most pragmatic and productive way to understand what particular experiences mean to

participants.

Multiple interviews allow participants to reflect on their experiences and to add to what

they said in earlier interviews (Lodico et al., 2006). In addition, new questions may arise after

the first interview and a second interview is required (Bourdieu, 1999c). In order to explore

the deep meaning embedded in the participants’ experiences, multiple interviews were used to

collect the qualitative data for this study.

Interviews have often been criticised for lack of objectivity, due in particular to the

human interaction inherent in the interview process. Kvale (1996) argues for three

conceptions of objectivity suitable for interview research: freedom from bias, intersubjective

knowledge, and reflecting the nature of the object. Firstly, objectivity as freedom from bias

refers to “good, solid, craftsmanlike research” (Kvale, 1996, p. 64), producing reliable

knowledge that has been systematically cross-checked, controlled, and verified, undistorted

by personal bias and prejudice as far as possible. Secondly, this knowledge must be

intersubjectively testable and reproducible, which means that repeated research of the same

problem by different researchers should yield the same data. This is what Kvale (1996, p. 64)

refers to as “dialogical intersubjectivity”, which is an agreement through a rational discourse,

reciprocal critique, and communicative validation among researchers, as well as between

researchers and their participants. Thirdly, objectivity also means the reflection and

expression of the nature of the participants researched by treating the participants as the

speaking subjects existing in a linguistically constituted and interpersonally negotiated

interview process. As such, the interview, as a linguistic, interpersonal, and interpretive

method, becomes more objective.

141

To maximise these dimensions of objectivity, the interview also requires attention to

two methodological concerns, namely ‘reliability’ and ‘validity’. ‘Reliability’ refers to how

consistent the results are, and ‘validity’ means whether an interview investigates what is

intended to be investigated (Kvale, 1996). These two issues do not belong to some separate

stages of an investigation, but should be addressed throughout the entire interview research

process, starting from thematising, designing, interviewing, transcribing, and analysing, to

reporting. Each stage will be discussed in the following sections.

4.4.1 Thematising

Thematising refers to a conceptual clarification and a theoretical analysis of the theme(s) to

be investigated, and informs the formulation of the questions to be asked (Kvale, 1996). The

theoretical framework developed in Chapter Three (see Figure 3.4) has conceptualised the

main themes under investigation and has related these themes, or theoretical constructs, to

one another. As Bourdieu (1991) suggested, languages can produce profit in response to the

demands of a given field. Accordingly, the overarching theme to be investigated in the

qualitative phase is in relation to the profits produced through young Chinese Australian

adults’ CHL learning in certain fields. Attuned to this theme, the analytical questions driving

the qualitative phase include: Does young Chinese Australian adults’ investment in CHL

learning have any return at a given time and place? Does CHL proficiency produce any

valuable asset for young Chinese Australian adults in any situation? Does lack of CHL

proficiency impact on their life in any way? These questions align with the research interview

approach. They can return to the subjective experience of young Chinese Australian adults

regarding their CHL usage and learning trajectories over time. The research interview can

obtain comprehensive descriptions about the CHL usage and learning experience from the

142

participants’ perspective, and then provide the basis for a deep analysis to portray the essence

of the experience, the essential meanings of CHL to these participants.

If the structural models (see Figures 4.2 and 4.3) deduced from the first part of the

theoretical framework (see Figure 4.1) functioned to drive the initial quantitative phase, the

subsequent qualitative phase was guided by the second part of the theoretical framework (see

Figure 4.4) deduced from the overall theoretical framework (see Figure 3.4). It is worth

noting that the single-headed arrows demonstrated in Figures 4.1 and 4.4 were in the opposite

direction because the research questions in the initial quantitative phase and the subsequent

qualitative phase addressed relationships that went in different directions.

Figure 4.4. Theoretical framework guiding the qualitative phase

4.4.2 Designing

Designing consists of overall planning and preparing the methodological procedures for

obtaining the intended knowledge (Kvale, 1996). The following sections explain the rationale

behind the design for interview and sampling strategy in the qualitative phase.

CHL proficiency as

linguistic capital

Four forms of

capital

Chineseness as

habitus

Fields

143

Interview design

The one-to-one interview is purported to be the most commonly used data collection method

in qualitative social research (Ryan, Coughlan, & Cronin, 2009). It is a data production

process of asking questions and recording answers from only one participant at a time. It is a

valuable approach to gain insight into participants’ perceptions, understandings, and

experiences of a given phenomenon under certain circumstances (Lodico et al., 2006; Ryan et

al., 2009). The qualitative phase aimed to address participants’ perspectives and viewpoints

on their own CHL commitments, and thus a one-to-one interview method was selected.

Semi-structured interviewing provides the flexibility to combine an interview schedule

of prepared questions on predetermined topics with explorations of any unanticipated

responses or issues that emerge through the use of open-ended questioning (Tod, 2006). It has

a sequence of themes to be covered, as well as suggested questions, yet at the same time, it is

open to changes of sequence and further probes in order to follow up on answers given and

experiences explained by the participants. The flexibility of the semi-structured interview

allows the interviewer, who is the researcher himself in the current study, to pursue a series of

less structured questions and to explore unforeseen issues raised spontaneously by the

participants. Reponses to structured questions can net useful information. The open-ended

responses can explore details that may surface during the course of the interview. This

facilitates the production of richer and more textured data from the participants than from

formally structured questions in standardised interviews. Therefore, a semi-structured

interview design was adopted in the qualitative phase.

If residing in Brisbane, participants were interviewed face-to-face. Otherwise, they

were interviewed by telephone or via online software such as Skype or MSN, because

telephone and online interviews were more cost-and time-effective than face-to-face

interviews in this situation. All the participants have achieved native, native-like, or nearly

144

native-like command of English. In contrast, their Chinese functioned as their HL, the

proficiency of which was lower or much lower than that of their English. Therefore, the

interviews were conducted in English.

An interview protocol was prepared for asking questions and recording key points

during the interview. As suggested by Lodico, Spaulding, and Voegtle (2006), the interview

protocol should usually include a heading that records date, place, and interviewee code, a

brief introduction of the study, the preliminary questions to be asked, and possible probing

questions to follow up, or in Bourdieu’s term “constant improvisation of pertinent questions”

(Bourdieu, 1999c, p. 613). These interview questions have essential relationships with the

research questions. However, interview questions tend to be more contextual than research

questions. Their development requires creativity, sensitivity, and insight, rather than a

mechanical translation of the research questions into an interview guide.

Patton (1990) offers advice concerning the development of good interview questions to

address reliability and validity. Questions can be asked from a variety of angles, including

experience questions, opinion, feeling, or value questions, knowledge questions, and

background or demographic questions. Another piece of advice is to offer a quotation

selected from another source that contains ideas on which the participants may comment. The

value of the quotation is that it attributes to someone else ideas that are usefully provocative

but are not in the researcher’s own voice. In contrast to the above good practice for interview

questions, questions that are dichotomous ‘yes or no’ questions are not helpful for exploring

details because such questions precipitate short answers. Multiple questions should be

avoided because the participants will most likely talk more fully about one of the several

questions and forget the others.

Leading questions that suggest an answer or contain information expected by the

researcher (Patton, 1990) have pros and cons. On the one hand, the leading effects of such

145

questions are well documented (Kvale, 1996). They give the participants hints about what

would be considered a desirable or appropriate kind of answer. As a result, the answer may

not reflect the thoughts of the participants. On the other hand, they can be judiciously used to

clarify the exact meaning of the participants. The decisive issue for interview questions is not

whether to lead or not, but where the questions do lead, whether they lead in important

directions that yield new and worthwhile knowledge (Kvale, 1996).

Structured within the themes and guided by the research questions, the interview

questions in the current qualitative phase consisted of four sections, namely CHL learning,

CHL usage, attitudes towards CHL, as well as particular attributes for each participant (see

Appendix 4). The prepared probing questions were developed with reference to the

theoretical framework of the current study. Bourdieu’s concept of capital, habitus, and field

informed the probing questions. These probing questions were used to elicit information to

explore whether CHL proficiency can produce ‘capital’ and ‘habitus’ for the interview

participants in given fields, however these theoretical terms were expressed in everyday

terms, such as ‘resources’, ‘Chinese identity’, and different ‘contexts’ or ‘situations’.

To make sure that the interview questions made sense to the participants and the

participants felt comfortable with the interview questions, a pilot interview was conducted

before the main interview. The pilot interview served to reshape the original interview

questions and interview style. For example, in the original design one of the questions was “If

you were to have children, would you encourage them to learn Chinese?” One of the pilot

interviewees suggested that to certain extent, this was a leading question because anyone with

a Chinese ancestry tends to say yes to this question even if Chinese language is not the first

choice for their children. Following this suggestion, the original question was amended to “If

you were to have children, would you encourage them to learn another language other than

English?” If the participants said “Chinese”, the reasons behind this choice were probed. If

146

the participants said any other language, they were asked why they would not encourage their

children to learn Chinese. In this way, the primary and overall question was presented at the

very beginning and then the interview process was managed by using interview skills to elicit

information. When the information elicited appeared to be drying up, pre-prepared probe

questions planned on the interview protocol were used. As Cavana, Delahaye, and Sekaran

(2001) suggested, a semi-structured interview should commence with an open-ended question

then progress to a more closed-ended form. The pilot interview provided good practice to

improve the researcher’s interview skills. It also prompted the researcher to rework some of

the questions into a more open form.

Audio-recording was used to record the interviews. Two audio recorders were used in

case of equipment failure. In addition to the audio recording, field-notes were only used when

necessary to record key points of participants’ responses as faithfully as possible and to serve

as reminders for the researcher to return to any issue that needed more probing in the

interview. Therefore, digital recording and field-notes were used together to ensure high

quality data.

Before the interview, appropriate locations and time were found. A time block

convenient for the participants and available for both the researcher and the participants, as

well as a quiet and physically comfortable location were the overriding considerations.

Participants’ willingness to cooperate can be contingent on convenience. Therefore, it was

important to defer to the participants’ needs as much as possible.

At the very beginning of each interview, the context was introduced by a briefing about

the purpose of the interview, ethical considerations, the use of audio recorders, and the

general topics in the interview. Each participant was given an information sheet to read prior

to the interview (see Appendix 2). If participants decided to take part in the study, the

researcher asked them to sign a consent form (see Appendix 3). In this way, the researcher

147

introduced the proposed research to the participants and obtained their informed consent in

compliance with ethical requirements. Before starting the interview, the researcher also asked

whether the participant had any questions. Such briefing in the first minutes of an interview is

crucial (Kvale, 1996). This was a good opportunity to establish a friendly atmosphere and

mutual trust by attentive interactions, where the participants could have a sense of who the

researcher was before they talked freely about their lived world to a stranger, and where the

researcher could show interest, understanding, and respect for what the participants said.

By the end of the interview, the participants have given much information about their

lived world. They might have obtained new insights into aspects of their lived world but

sometimes they might not have received anything in return. Therefore, the interaction was

rounded off by the researcher reflecting on some of the main points learned from the

participants. The participants might want to comment on this feedback. The interaction was

thereafter concluded by the researcher saying, for example, “I have no further question. Do

you have anything more you want to bring up, or ask about, before we finish the interview?”

(Kvale, 1996, p. 128). This gave the participants an additional opportunity to deal with issues

they have been thinking or worrying about during the interview. It was worthwhile for the

researcher to set aside a short block of quiet time right after each interview to recall and

reflect on what has been learned from the particular interview. The immediate impressions, in

the form of field notes or audio records, provided a valuable resource for later analysis of the

transcripts.

Sampling

There is no definitive answer to the question of how many participants are needed in an

interview study (Mertens, 2005; Mertler & Charles, 2005). Within qualitative research,

148

although diverse samples might provide a broader range to distil the essence of the

phenomenon, data from only a few participants who have experienced the phenomenon and

can provide a detailed account of their experience about the phenomenon might suffice to

illuminate its core elements.

Unlike the sampling principles in the quantitative approach, where the samples usually

are statistically representative, the sampling principle in the qualitative approach is more

strategic. Therefore, purposeful sampling is more often used in qualitative research (Lodico et

al., 2006). The logic and power of purposeful sampling is the selection of cases that provide

rich and deep information central to the purpose of the research (Lodico et al., 2006; Patton,

1990). Participants in the initial survey were also invited to participate in an interview. Those

who were interested in the interview and who left their contact details became candidate

interview participants. The current study adopted purposeful sampling to select participants

from the interview candidate list.

Specifically, Patton’s (1990) extreme or deviant sampling and maximum-variation

sampling was used. Under the extreme or deviant sampling strategy, interviewees can be

selected from extreme cases that are unusual or special in some way. The designations imply

that they are odd phenomena or cases that represent the extremes. These types of cases often

provide considerably more insight than representative or average cases (Danermark et al.,

2002). In the qualitative phase, extreme cases were identified by revisiting the outliers in the

initial quantitative phase. Identification strategies of the outliers will be explained in Chapter

Six. Outliers who were also candidate participants were invited to participate in the interview.

New knowledge learnt from an intensive examination of the experiences of special or unusual

participants through the interview strategy in the qualitative phase were considered to offer a

good complement to the knowledge learnt from statistical depictions of what the more regular

cases were like through the survey in the initial quantitative phase.

149

Interviewees were also selected on the criterion of maximising variation. They were

selected from varied cases to gain insight into the importance of various conditions for

producing the particular phenomenon under investigation. This strategy can document what is

unique about each situation as well as what is common across diverse settings. In the

qualitative phase, participants’ self-reported CHL proficiency in the initial quantitative phase

was taken into account. To cover maximum variation, candidate participants with highly

divergent levels of CHL proficiency were identified and invited to participate in the interview.

Account was also made to maximum variation in participants’ demographic features.

Eventually, Patton’s (1990) extreme or deviant sampling and maximum-variation sampling

helped to net five interviewees. A detailed description of these interviewees is presented in

Chapter Six.

4.4.3 Interviewing

Interviewing is an engaging process, an interpersonal situation, and an interactional

conversation (Kvale, 1996). It is up to the researcher to create in a short time an interpersonal

situation that allows the interaction to go beyond merely a polite conversation or exchange of

ideas. The researcher must establish an atmosphere where the participants feel safe and

comfortable enough to talk freely about their lived world. In order to achieve this, several

important issues need due attention. They are addressed as follows.

There is a definite asymmetry of power during the interview process because it is the

researcher who defines the situation, introduces the topics of the conversation, and steers the

course of the interview through further questions. Bourdieu (1999c) has explained this

asymmetry:

150

It is the investigator who starts the game and sets up its rules, and is usually the one

who, unilaterally and without any preliminary negotiations, assigns the interview its

objectives and uses…This asymmetry is reinforced by a social asymmetry each time

the investigator occupies a higher place in the social hierarchy of different types of

capital, cultural capital in particular. (p. 609)

Due to the unequally distributed cultural capital, there was an asymmetry of power existing in

the context of the current study, where the researcher, an educated native Chinese speaker,

had conversations around topics associated with CHL learning experiences of the Chinese

Australian participants. In this social field, the participants might assume that the researcher

had better Chinese language proficiency or more cultural capital in Bourdieu’s sense than

themselves. This assumption entailed hierarchical field positions occupied by the researcher

and the participants based on different levels of Chinese language competence valued as

cultural capital in this field. Because of the recognised value, Chinese language proficiency

might accrue symbolic capital. Lack of this symbolic capital might yield embarrassment of

‘loss of face’ and generate an uncomfortable environment for the participants during the

interviews.

In order to control the effects of asymmetry of power, the “symbolic violence” exerted

through the asymmetric relationship should be reduced as much as possible (Bourdieu, 1999c,

p. 609). In practice, naivety characterises the researcher’s special learner role (Kvale, 1996).

It entails a frame of mind to set aside any assumptions that the meaning of the participants

has been known to the research. Such assumptions would preclude the researcher from

seeking explanations and shut down further probes. The learner role can be administered

through verbal “signs of feedback” or “response tokens” (Bourdieu, 1999c, p. 610), such as

“yes”, “right”, “oh”, and “ok”, as well as aiding explanations in a proposed rather than an

151

imposed way, such as suggestions to offer multiple and open-ended continuations to the

participants’ hesitations or searchings for appropriate expressions (Bourdieu, 1999c, pp.

614-615); and non-verbal means, such as an open posture, approving nods, appropriate facial

expressions, and good eye contact. These verbal and bodily signs of attention, interest,

approval, encouragement, and recognition are conditions for a continuing interchange that

encourages the collaboration of the participants interviewed. These practices ensure that the

researcher adopts the role of a listener and a seeker of knowledge during the whole process of

interviewing. In Bourdieu’s words, it is “active and methodical listening” that leads to

“adopting the interviewees’ language, views, feelings, and thoughts” (Bourdieu, 1999c, p.

609) and signals “the interviewer’s intellectual and emotional participation” (Bourdieu,

1999c, p. 10).

Apart from “active and methodical listening”, “proximity and familiarity” also provide

conditions of “nonviolent communication” during the interview (Bourdieu, 1999c, p. 610).

Common cultural and physical dispositions shared by the researcher and the interviewees

may contribute to this proximity and familiarity. For example, William Labov (1972) asked

young blacks to conduct the linguistic investigation of the speech used by blacks in Harlem.

In the same way, Bourdieu (1999c, p. 611) gave “training in interview techniques to

individuals who could have the kind of familiar access to certain categories of respondent”.

In the current study, the researcher, who is a native Chinese, is culturally and physically close

to the interviewees, or in Bourdieu’s sense “linked to them by close familiarity” (1999c, p.

611). When a young Chinese researcher interviewed these young Chinese Australians, the

conversation could spring from the dispositions attuned to each other. The interviewees

seemed to consider this situation as an exceptional opportunity to “make themselves heard”

and to “carry their experience over from the private to the public sphere” (Bourdieu, 1999c, p.

615). Their speech seemed to convey “a joy of expression” (Bourdieu, 1999c, p. 615).

152

However, this relatively easy access and affinity with the interviewees do not necessarily lead

to biased interpretation of the data to meet the researcher’s own assumptions and expectations.

As an international student from China, the researcher was more confined to the university

life and removed from the more complex lived worlds of Chinese Australians. Such social

difference enabled the researcher to step back as a listener from the interviewees’ experiences.

In brief, ‘proximity’ and ‘social difference’ avoid the two problematic extremes: “total

divergence” between the researcher and the interviewees, where understanding and trust are

not possible, and “total overlap”, where nothing can be said and questioned because

everything goes without saying (Bourdieu, 1999c, p. 612).

To sum up, the interview is a social interaction where the relationship between the

researcher and the interviewee(s) is of paramount importance. The researcher has the

responsibility of ensuring that the interviewee is at ease and also of taking the role of an

active learner. A relaxed, confident, and attentive approach demonstrated by the researcher

helps a good interview relationship to develop. However, the interview is more than just a

social interaction. It is a process with a distinct purpose to derive meaning from a particular

situation. It is a performance that demands skill and forethought. In Bourdieu’s sense,

interviewing demands a “craft”, a real disposition to pursue truth, which disposes the

researcher to improvise on the spot, in the urgency of the interview, strategies of

self-presentation and adaptive responses, encouragement and opportune questions to help

participants deliver up their truth (Bourdieu, 1999c, p. 621). It is “a sort of spiritual exercise

that, through forgetfulness of self, aims at a true conversion of the way we look at other

people in the ordinary circumstances of life” (Bourdieu, 1999c, p. 614).

153

4.4.4 Transcribing

Transcribing is the procedure of translation from oral conversations to written texts (Kvale,

1996). Raw interview data in this study consisted of audio recordings and field notes, which

were converted into transcripts. They were first organised and prepared for analysis by

transcribing audio recording and typing up field notes. The use of transcribed data enhances

the precision, and hence the reliability, of the qualitative analysis (Peräkylä, 2004; Silverman,

1993).

Reliable and valid transcriptions are consistent and accurate descriptions of the data.

The descriptive aspect of data is crucial so readers can make their own interpretations (Patton,

1990). Reliable and valid transcribing renders the interview conversations in a format

amenable to closer analysis.

Transcription reliability can be improved by listening to the audio recording a few

times. This increased the consistency of the transcripts. ‘Member checking’ was also used to

achieve transcription reliability. The supervisory team reviewed the transcripts and helped

identify some vague language that needed to be clarified. The researcher checked these

passages by listening to the audio recording again. Where the researcher struggled to capture

what exactly was said by the interviewees, native English speakers were invited to help with

the transcription. This ensured that the interviewees’ replies were transcribed as accurately as

possible.

Ascertaining transcription validity is more complex than assuring its reliability. Kvale

(1996) suggested the combination of several techniques to improve validity. Firstly, verbatim

descriptions respect the original meanings of the participants whose own words will be

transcribed as much as possible. The researcher of the current study transcribed the wording

of the participants as accurately as possible by listening to the audio recording many times.

Secondly, nonverbal communications were recorded on the field notes where necessary and

154

possible. These forms of communication, such as pauses, laughter, interruptions, and changes

in vocal tone or emotion, provided a richer access to the participants’ meanings. Thirdly,

transforming the conversation into a written form facilitates communication of the

participants’ meanings to readers. While transcribing, the researcher corrected superficial

conversational grammar mistakes of the participants for the sake of reading clarity. Care was

taken to preserve meaning. As oral language uses different clause patterns from written

language, the talk was broken up into sentences according to the researcher’s own

understanding and later checked by the participants. In brief, the combination of verbatim

descriptions, nonverbal communications, and written styles made the transcription more

meaningful and manipulable.

Kvale’s advice resonates with Bourdieu’s viewpoints that transcription is subject to the

reconciliation of two sets of constraints: the constraint of fidelity and “the constraint of

readability” (Bourdieu, 1999c, p. 622). Transcription means writing and rewriting. The

writing process is faithful to everything that came up in the interview, while the rewriting

process sometimes has to rid the transcribed text of certain confused phrases or verbal slips to

improve readability (Bourdieu, 1999c). Reconciliation of the two constraints is therefore

paradoxical. As Bourdieu (Bourdieu, 1999c, p. 622) explained, “the transition from the oral to

the written, with the changes in medium, imposes infidelities which are without doubt the

condition of a true fidelity.”

4.4.5 Analysing

Data analysis in qualitative studies is an ongoing process (Mertens, 2005). It does not occur

only at the end of the study as is typical in quantitative studies. Analysis in qualitative studies

is typically recursive (Mertens, 2005). By this is meant that findings are generated and

155

systematically built as successive pieces of data are produced. This analysis of the subjective

experience and the initial ‘objective’ analysis in the quantitative phase can fit within the same

methodological paradigm. Bourdieu (1981, p. 96) contended that objective analysis does not

contradict subjective analysis “of primary experience of the social world, and of the

immediate comprehension of the utterances, acts, or works of others”. It is necessary to pass

from the “statistical regularity or algebraic structure” to the principle of qualitative

investigation of subjective experience in the social world (Bourdieu, 1981, p. 94). The flow

of this qualitative approach to data analysis was administered as follows.

The first step was to obtain a general sense of the information by reading through all

the data. This initial review did not involve a careful reading for details. Instead, this initial

review was more comprehensive, examining all the different sources of data to understand

the scope of the data.

The next step was meaning condensation that entailed an abridgement of the meanings

expressed by the participants into shorter formulations (Kvale, 1996). The “natural meaning

units” (Kvale, 1996, p. 194), which were quite often long statements expressed by the

participants, were determined by the researcher. Long statements were compressed into

briefer statements in which the main sense of what was said was rephrased in a few words.

The main sense that dominated a natural meaning unit was stated as succinctly as possible.

The researcher attempted to read the participants’ main sense without prejudice or judgement

and to generalise participants’ main sense into briefer statements as understood by the

researcher. Meaning condensation thus involved a reduction of large interview texts into

briefer and more succinct formulations.

Coding was then conducted. It is the process of dividing the materials into manageable

chunks of meaning, identifying different segments, and labelling those categories with a term

that can be used across data sources (Lodico et al., 2006). A combination of predetermined,

156

theoretically informed, and emerging codes was used. Data were continually read, reread, and

re-examined. New codes were added as the data were reviewed. The initial codes were

gradually combined and reduced with the goal of eliminating overlap and producing a more

coherent view of the patterns in the data (Lodico et al., 2006). The software NVivo was used

to help with the coding process. This was time-effective compared to the traditional coding

on the hardcopy of the transcripts where texts were manually marked with different colours.

Categories for segments were described by rendering the information in detail. That

detailed information was used to generate a small number of essential and non-redundant

themes. These themes were described using a few words or phrases but they identified the

major issues to interpret and represent the data. These emergent themes served as major

findings in the qualitative phase and helped structure the discussion. They displayed

perspectives from participants by selected and representative quotations and interpretations.

By doing this, different participants’ experiences were reconstructed into a richer, more

condensed and coherent story than the scattered stories of the separate participants. More

sophisticated analysis was then able to proceed. That analysis went beyond information

description and theme identification to a search for theoretical connections.

4.4.6 Reporting

Reporting is the end product of a long analytical process in interview research. The main aim

of a report is to demonstrate the importance and the trustworthiness of the findings (Kvale,

1996). Reporting is not simply representing the views of the participants, accompanied by the

researcher’s viewpoints in the form of interpretations. Rather, it is a social construction

whereby the researcher’s voice, writing style, and literary devices construct a specific view

on the participants’ lived world (Kvale, 1996). What is worth reporting is to be conveyed in a

157

limited number of pages, presenting the main aims, methods, results, and implications of the

interview research, and is thus necessarily selective.

The common mode of presenting the findings of the interview research is through

selected quotes. Kvale (1996) suggested several strategies when reporting quotes. First of all,

quotes should be contextualised to render the interview context, which the research knows

well but which is unknown to the reader. Secondly, quotes should be interpreted by clear

statements of what viewpoint a quote illuminates, proves, or disproves. Moreover, quotes

should be rendered in a written style. Verbal transcriptions of oral speech, with repetitions,

digressions, and pauses, are difficult to grasp when presented in a written form. To facilitate

comprehension, the participants’ spontaneous oral speech should be rendered into a readable

and written textual form in the final report. The exception is when the linguistic form itself is

important to the study (Kvale, 1996). Last but not least, there is also an issue regarding the

balance between quotes and texts. The quotes should not make up more than half of the text

(Kvale, 1996) because many quotes with few connecting comments and interpretations can

appear chaotic and produce a linguistic flicker effect. Too many quotes mean that the data are

expected to speak for themselves. Instead, the researcher should have a theoretical frame to

inform the reader of the data by interpretation of the quotes.

The possibility of presenting the interview results in visual forms should not be

overlooked, although interview data are typically of a verbal nature. To this end, summary

tables were used in Chapter Six to report aspects of the interview results.

4.5 Ethical considerations

Ethics is not an afterthought or burden but an integral part of research design and

implementation. At all times, ethical considerations should be at the forefront of the

158

researcher’s agenda. Ethics functions as a form of guardianship against inappropriate research

behaviours, such as impropriety, deception, disturbance, offence, discrimination, and invasion

of privacy. Ethics also seeks to guarantee beneficence, respect, and justice in the research.

Therefore, sound ethics and sound methodology are mutually constitutive. Deeper concerns

about the rights, welfare, and betterment of participants tend to result in greater involvement

and engagement of the participants in the research process, and ultimately, richer data.

The current study included working and interacting with humans to conduct

questionnaires and interviews in Australia. This research was considered to pose a low risk to

participants. All participants were adults and the theme of the study was related to their

Chinese cultural heritage and CHL learning experiences. There was no intent to elicit

sensitive or intrusive data. Participation in the questionnaire or interview was informed and

voluntary. Participants were provided with detailed information about the study (see

Appendix 1 and 2). They could withdraw from the study if they felt uncomfortable when

doing the questionnaire or interview, which was an option but was not taken up by any

participant. The research was conducted in accordance with the National Statement on Ethical

Conduct in Research Involving Humans (2007) and the ethical guidelines of Queensland

University of Technology Human Ethics Committee. An application for ethical review was

submitted to the low risk review team in the Faculty of Education. The low risk application

was reviewed by the Faculty Research Ethics Advisor and confirmed as meeting the

requirements of the National Statement on Ethical Conduct in Human Research. The current

research has received approval (number 1100000165) from the Research Ethics Unit at

Queensland University of Technology.

Key risks related to the potential for distress if the questionnaire or interview questions

inadvertently elicited recall of unpleasant cultural or learning experiences, and to

inconvenience arising from the times, duration, and frequency of interviews. The researcher

159

carefully managed these potential risks. Consequently, no participants seemed to feel

uncomfortable about participating in the research.

An online questionnaire was used so that the privacy of participants was protected. The

data that participants provided were handled in such a way that no uniquely identifying

information was attached to the data and nobody could trace the data back to the participants

who provided them. There was an exception if the participants of the online questionnaire

were interested in participating in the interview afterwards. They were asked to leave their

contact details to arrange a comfortable time and place for interviews. The researcher thus

knew the identities of the participants in the interviews. However, an identity code was used

on any transcripts and pseudonyms were used in any reports. These participants’ identities

were only known and available to the researcher. They were not disclosed. Identification was

removed from the stored data. All paper-based documents were kept securely in a locked

filing cabinet in the researcher’s QUT office. E-files were stored on a QUT

password-protected network drive. USB drives were only used for data transfer. Only the

researcher had access to the raw data.

Ethics considerations promoted the integrity of the current research. However, ethical

considerations extended further than the promise of confidentiality and anonymity.

Participants volunteered to share their life stories. Therefore, they deserved the highest

standards of ethical treatment. The researcher endeavoured to protect the participants and

maintain their trust and respect. Not only participants but also research sites were respected.

The researcher displayed this respect by gaining permission before entering any site and by

being careful to minimise any disruption to the physical settings.

The whole process of this research paid great heed to ethical considerations. Research

design, data production, analysis and interpretation, final writing, and future dissemination of

the research all reflected cognisance of ethical practice.

160

4.6 Chapter summary

This chapter prepared the path for the research. The rich methodological pluralism wove the

initial quantitative phase and the subsequent qualitative phase together, helped think through

‘objectivism’ and ‘subjectivism’, and reconciled ‘agency’ and ‘structure’. As such, it

resonates with the Bourdieusian stances methodologically and theoretically. The selection of

an explanatory mixed methods design was based on philosophical, theoretical, and

methodological considerations. This approach established the basis for the detailed research

design that enabled data production, collection, and analysis to address the two research

questions. The two sub-theoretical frameworks, shown in Figure 4.1 and Figure 4.4, guided

the initial quantitative phase and the subsequent qualitative phase respectively, each of which

addressed one research question. The combination of quantitative and qualitative approaches

talked back to the overall theoretical framework and guided the data analysis and discussion.

Guided by the methodology chapter, the next two chapters report on the data analysis,

demonstrate the results, and reveal the findings. In the initial quantitative phase, SEM tests

the first part of the theoretical framework (Figure 4.1) by fitting the data into the statistical

models. The subsequent qualitative investigation is guided by the second part of the

theoretical framework (Figure 4.4). The detailed quantitative and qualitative data analysis and

results will be presented in the following Chapter Five and Chapter Six.

161

Chapter 5: Quantitative Data Analysis and Findings

This chapter reports the process of the quantitative analysis of the data. It also reports the

findings from the quantitative phase. The focus of this chapter addresses RQ1: Is CHL

proficiency of young Chinese Australian adults influenced by their investment of capital, the

strength of their habitus of ‘Chineseness’, or both? The first part of the theoretical framework

(Figure 4.1) guides the examination of this research question. Section 5.1 reports on the pilot

study, including issues related to face validity and internal consistency reliability of the

measurements. Section 5.2 details the quantitative data analysis of the main study and

explicates the results. EFA and CFA function to develop the measurement models for the six

theoretical constructs, namely ‘Chineseness’, ‘economic capital’, ‘cultural capital’, ‘social

capital’, ‘symbolic capital’, and ‘CHL proficiency’, specified as latent variables with their

corresponding indicators. They also function to validate the reliability and validity of the

measurement models. SEM constructs the relationship between participants’ ‘Chineseness’

and their ‘CHL proficiency’ as well as the relationship between their ‘capital’ and their ‘CHL

proficiency’, which serves the hypothesis testing. Multiple regression models function to

investigate whether participants’ habitus of ‘Chineseness’ and their investment of ‘capital’

co-contribute to their ‘CHL proficiency’. Section 5.3 highlights the key findings from the

quantitative data and concludes with a brief summary.

5.1 Pilot study

In the first stage of the pilot study, a printed copy of the questionnaire was distributed to five

participants, who were known to the researcher. Face validity and language appropriateness

were checked by virtue of the feedback from the participants. This ‘pre-pilot’ process was

162

important in that it identified some problems with the instrument. These problems and the

resultant modification of the instrument are discussed below.

In the original design, all participants were asked to choose which generation of

Chinese Australians they represented. A Malaysian-born Chinese Australian participant

commented that he considered all non-Australian-born Chinese Australians first-generation

Chinese Australians, no matter whether they moved to live in Australia by themselves or with

their parents. According to generation demarcation of the current study, participants born

outside Australia were treated as first-generation Chinese Australians. Therefore, they should

not have been asked about their generation. Only those born in Australia were asked this

question.

Participants were asked how many times they had visited China while living in

Australia. One participant had visited China while living in Singapore. She asked whether

this should count or not. The purpose of this item was to find out how frequently participants

had visited China. Therefore, this item was reworded as “How many times have you visited

China?”

Participants were also asked whether they had lived in a country other than Australia

and their country of birth. Some participants asked how long could be regarded as “have

lived”. This item was then clarified as “Have you ever lived in a country other than Australia

and your country of birth for more than one year?”

The above examples demonstrated how the pre-pilot process helped identify issues and

‘fine-tune’ the instrument. In this way, an online questionnaire was developed for the second

stage of the pilot study. Before reporting the second stage of the pilot study, the distinctions

between a reflective model and a formative model explained in Chapter Four were revisited.

The major statistical tests used in the current study are all reflective per se. These tests

included inter-item reliability test, Exploratory Factor Analysis (EFA), Confirmatory Factor

163

Analysis (CFA), and Structural Equation Modelling (SEM). Before conducting these tests,

the questionnaire items were revisited. Items inappropriate to plumb reflective models were

excluded from these tests. But some of the excluded items were still kept in the questionnaire

because they can capture meaningful information for the current study. For example, the item

(v10) measuring the individual gross annual income was not appropriate for reflective models

because the individual gross annual income may or may not vary with the variation of the

individual investment of economic capital in CHL learning. However, this item (v10) can

offer information about participants’ economic status. Therefore, it was kept in the online

questionnaire but was not included in the inter-item reliability test.

At the second stage of the pilot study, the link to the online questionnaire was sent to

38 participants who met the demographic requirements of the study. Given that these

participants were all known to the researcher, the return rate was 100%. It will be recalled

that there are six constructs as latent variables, namely ‘Chineseness’, ‘economic capital’,

‘cultural capital’, ‘social capital’, ‘symbolic capital’, and ‘CHL proficiency’. The online

questionnaire items were treated as indicator variables for these constructs. In the pilot study,

the item sets used to map the corresponding constructs are detailed in Table 5.1.

164

Table 5.1

Item sets and their corresponding constructs in the pilot study

Constructs Indicators Item description

Chineseness

v1 My mathematics was much better than that of my classmates.

v2 Handwriting tells you a lot about a person’s character.

v3 Children must be taught to paint, dance, or play musical instruments.

v4 Academic education is the most important thing in school.

v5 I prefer to live close to my family members.

v6 I hope to have sons to continue my family line or my husband’s family line.

v7 People who are senior in age and/or position should be addressed by their

title plus surname rather than their first name.

v8 To save face I always prefer to say ‘yes’.

v9 Confucianism values the virtue of frugality. Confucius said that frugality is

a common character of people with virtue. He also said that frugality is

very important in family life. In general, to what degree would you agree

with these values?

Economic

capital

v12 Money is not an issue if I wish to travel to China.

v13 I can afford to study in Chinese language schools as many hours as I wish.

v14 I can afford to employ private tutors to teach me Chinese if I wish.

v15 Money is never an issue if I want to buy Chinese language learning

materials, such as textbooks, dictionaries, and tapes.

Cultural

capital

v16 Learning Chinese is very important to me.

v17 I feel very sad if I miss celebrating Chinese festivals (Spring Festival,

Mid-Autumn, etc.).

v18 I always keep up to date with current Chinese affairs by watching TV,

listening to the radio, reading newspapers, or surfing online.

v19 I always go to venues, such as libraries, galleries, museums, theatres, or

concerts, if they feature Chinese culture.

v20 I read a lot of books about China.

v21 I have invested a lot of time in practising Chinese cultural activities, such as

learning musical instruments, calligraphy, or painting.

Social

capital

v24 Most of my friends are of Chinese descent.

v25 I am a member of social groups, such as a church or club, which mostly

include members of Chinese descent.

v26 I tend to mix exclusively with Chinese social groups.

v27 I think it is necessary to learn Chinese language in order to socialise with

my Chinese peers.

Symbolic

capital

v31 The wealth of my family is well-known in Chinese communities.

v32 My Chinese language competency is well regarded.

v33 Learning Chinese has increased my status in Chinese communities.

v34 People value my Chinese heritage.

v35 People consider me very popular among my Chinese peers.

CHL

proficiency

v36 I can read Chinese newspapers and magazines easily.

v37 I can read Chinese popular fiction easily.

v38 I can express my personal preferences and opinions in very clearly written

Chinese.

v39 I seldom make grammatical errors when writing in Chinese.

v40 I can easily understand Chinese conversations between people.

v41 I can easily understand Chinese language in the media, such as TV shows,

videos, and movies.

v42 I can talk about my life in detail in Chinese.

v43 I can handle complex situations in Chinese, such as banking, arguing,

purchasing a house or a car.

v44 I make appropriate use of popular words or idiomatic phrases common in

Chinese-speaking communities.

165

The value of Cronbach’s α corresponding to all six latent variables was above the

cutoff value of .80 (Kline, 1999). The value of corrected item total correlation corresponding

to all items was well above the cutoff value of .33, indicating that each item accounted for

more than 10% of the variance of the corresponding theoretical construct (R. Ho, 2006).The

overall internal consistency reliability of the instrument was considered to be reasonably high.

Table 5.2 summarises the results of internal consistency reliability tests for the six latent

variables.

Table 5.2

Summary results of internal consistency reliability test

Variables Chineseness Economic

capital

Cultural

capital

Social

capital

Symbolic

capital

CHL

proficiency

Cronbach’s α .91 .84 .91 .90 .87 .98

A close investigation of the inter-item correlation matrix identified that some indicators

mapping the construct ‘CHL proficiency’ were highly correlated (>.90), which could be

indicative of multicollinearity (Field, 2009). This was because the highly correlated items

were not distinguishable enough to stratify participants’ CHL proficiency into different levels.

For example, if participants could read Chinese newspapers and magazines easily, they were

likely to read Chinese popular fiction stories easily. Their CHL reading proficiency could not

be distinguished by these two items. Statistically, these two items were redundant and thus

were highly correlated. Based on extensive discussions with experienced Chinese language

teachers and CHL learners at various levels, the following two items were believed to be able

to distinguish participants’ CHL reading proficiency: (1) “I can read Chinese language

textbooks easily”; and (2) “I can read Chinese popular fiction stories easily”. Similar

strategies were applied to revise the indicators measuring CHL writing, listening, and

speaking proficiencies.

166

The distribution of the data points was also checked. For example, the data

corresponding to the item asking personal gross income was heavily positively skewed. There

was a larger concentration at the lower end of the scores. This indicated that the income

levels of the scale might have been too large. In order to better capture the spread of income,

the income levels within the scale were set to a lower level.

A 7-point bi-polar Likert scale was first used in the pilot study. The scale ranged from 1

(strongly disagree) to 7 (strongly agree). Compared to the bi-polar scale, the uni-polar scale

was later considered less mentally taxing because participants only have to consider one

attitude instead of balancing two opposing attitudes. As such, a uni-polar approach is

preferred wherever possible (Groves et al., 2004). Therefore, the original bi-polar scale was

changed to a 7-point uni-polar Likert-type scale in the main study. The scale ranged from 1

(not at all) to 7 (completely).

The pilot process shaped and cleaned the items that were used to measure their

corresponding constructs. Some of the items were reworded. Some of the items were

removed. Consequently, the items were renumbered. The item sets used in the main study to

map the corresponding constructs are detailed in Table 5.3.

Table 5.3

Item sets and their corresponding constructs in the main study

Constructs Indicators Item description

Chineseness

v1 My mathematics was much better than that of my classmates. To what extent

do you agree with this statement?

v2 Handwriting tells you a lot about a person’s character. To what extent do you

agree with this statement?

v3 Children must be taught to paint, dance, or play musical instruments. To

what extent do you agree with this statement?

v4 Academic education is the most important thing in school. To what extent do

you agree with this statement?

v5 I prefer to live close to my family members. To what extent do you agree

with this statement?

v6 I hope to have sons to continue my family line or my husband’s family line.

To what extent do you agree with this statement?


title plus surname rather than their first name. To what extent do you agree

167


v8 To save face I always prefer to say ‘yes’. To what extent do you agree with

this statement?

v9 Confucianism values the virtue of frugality. Confucius said that frugality is a

common character of people with virtue. He also said that frugality is very

important in family life. In general, how much degree would you agree with

these values?

Economic

capital

v11 Money is not an issue if I wish to travel to China. To what extent do you


v12 I can afford to study in Chinese language schools as many hours as I wish.


v13 I can afford to employ private tutors to teach me Chinese if I wish. To what

extent do you agree with this statement?

v14 Money is never an issue if I want to buy Chinese language learning

materials, such as textbooks, dictionaries, and tapes. To what extent do you


Cultural

capital

v15 Learning Chinese is very important to me. To what extent do you agree with

this statement?

v16 I feel very sad if I miss celebrating Chinese festivals (Spring Festival,

Mid-Autumn, etc.). To what extent do you agree with this statement?


listening to the radio, reading newspapers, or surfing online. To what extent



concerts, if they feature Chinese culture. To what extent do you agree with

this statement?

v19 I read a lot of books about China. To what extent do you agree with this

statement?


learning musical instruments, calligraphy, or painting. To what extent do you


Social

capital

v23 Most of my friends are of Chinese descent. To what extent do you agree



include members of Chinese descent. To what extent do you agree with this

statement?

v25 I tend to mix exclusively with Chinese social groups. To what extent do you


v26 I think it is necessary to learn Chinese language in order to socialise with my

Chinese peers. To what extent do you agree with this statement?

Symbolic

capital

v30 The wealth of my family is well-known in Chinese communities. To what


v31 My Chinese language competency is well regarded. To what extent do you


v32 Learning Chinese has increased my status in Chinese communities. To what


v33 People value my Chinese heritage. To what extent do you agree with this

statement?

v34 People consider me very popular among my Chinese peers. To what extent


CHL

proficiency

v35 I can easily understand my family members and friends when they talk to

me in Chinese. To what extent do you agree with this statement?

v36 I can easily understand Chinese language in the media, such as TV shows,

videos, and movies. To what extent do you agree with this statement?

v37 I can easily handle complex situations in Chinese, such as banking, arguing,

purchasing a house or a car. To what extent do you agree with this

168

statement?

v38 I can read Chinese language textbooks easily. To what extent do you agree


v39 I can have deep discussion and exchange thinking with people in Chinese.


v40 I can read Chinese popular fiction stories easily. To what extent do you


v41 I can always write Chinese characters and Chinese words correctly. To what


v42 I can express my personal preferences and opinions in very clearly written

Chinese. To what extent do you agree with this statement?

5.2 Main study

Two hundred and thirty people responded to the modified online questionnaire. Because the

‘not allow to skip question’ option was applied in Key Survey, respondents had to answer all

questions. By virtue of this strategy, all 230 cases were data complete. Due to the clearly

delineated eligibility for participation, all the respondents met the demographic requirements

of the study. They all ranged in age between 18 and 35. The age distribution demonstrated a

good spread (see Figure 5.1). Forty seven point eight percent (47.8%) of the respondents

were males and 52.2% were females. This gender distribution is comparable to ABS 2011

Census data, which reported 46.4% and 53.6% of Chinese Australians are males and females

respectively. As shown in Figure 5.2, all the respondents were from the capital cities where

the overwhelming proportion of the Chinese Australian population resided (93.6% according

to ABS 2011 Census data). The only capital city not represented in the study was Hobart.

This should not impact on the demographic diversity of the sample because the Chinese

Australian population in Tasmania was very small (according to ABS 2011 Census, 0.65% of

the Tasmanian population had Chinese ancestry). An overwhelming proportion (83.1%) of the

respondents lived in Sydney, Melbourne, and Brisbane. This proportion is comparable to ABS

2011 Census data, which reported 84.1%. In this vein, the demographic features of the

sample are largely consistent with those of the population.

169

Figure 5.1. Age distribution of participants

Figure 5.2. Distribution of participants’ resident cities

One hundred and nineteen respondents (51.7%) were born in Australia. Ninety five

respondents (41.4%) were born in China, including Chinese Mainland, Hong Kong, Macau,

and Taiwan. Sixteen respondents (6.9%) were born in other countries, including Indonesia,

Malaysia, New Zealand, Singapore, and Vietnam (see Figure 5.3). All the 111 non-Australian

born respondents (48.3% of the total respondents) came to live in Australia before the age of

13. They were regarded as first-generation Chinese Australians. Those born in Australia were

asked to designate their generations. Seventy-three of them (31.7% of the total respondents)

170

considered themselves to be second generation, 31 (13.5%) third generation, and 15 (6.5%)

fourth generation or further removed (see Figure 5.4).

Figure 5.3. Composition of participants’ birthplace

Figure 5.4. Composition of participants’ generation

One hundred and twelve respondents (48.7%) spoke both Chinese and English at home.

Sixty-four respondents (27.8%) only spoke Chinese at home and fifty-one respondents

(22.2%) only spoke English at home. Three respondents (1.3%) spoke other languages or

other languages mixed with Chinese and English at home, such as Indonesian, Indonesian

mixed with English and Chinese, or Vietnamese mixed with English and Chinese (see Figure

5.5). Of those who used Chinese at home, 109 (61.9%) used Mandarin, 55 (31.3%) used

Cantonese, and 12 (6.8%) used a mixture of Mandarin and Cantonese, or a mixture of

Mandarin and Hokkien, or other Chinese dialects, such as Fujianese, Hakka, Shanghainese,

Sichuanese, and Teochew (see Figure 5.6).

119, 51.7% 95, 41.4%

16, 6.9%

born in Australia

born in China

111, 48.3%

73, 31.7%

31, 13.5%

15, 6.5%

first generation

second generation

thrid generation

fourth generation and further removed

171

Figure 5.5. Composition of participant’s language usage at home

Figure 5.6. Composition of Chinese dialects used at home

An overwhelming proportion (87%) of the respondents had formally studied Chinese in

schools, universities, Confucius Institutes, and/or community schools. However, their

investment of time spent in studying Chinese varied from less than one year to over 15 years,

and their CHL proficiency varied remarkably. Data analysis reported in this chapter is to

explain how much variance of their CHL proficiency can be attributed to their habitus of

‘Chineseness’ and their investment of different forms of capital.

Descriptive statistics were applied before the main tests. Results are reported in Table

5.4. The standard deviations of the indicators ranged from 1.63 to 2.09 on a 7-point scale.

This suggested adequate variance in the responses.

112, 48.7%

64, 27.8%

51, 22.2%

3, 1.3% both English and Chinese

Chinese only

English only

other languages or other languages

mixed with English and Chinese

109, 61.9%

55, 31.3%

12, 6.8%

Mandarin

Cantonese

other dialects or a mixture of Mandarin and a dialect

172

Table 5.4

Descriptive statistics of indicators

Constructs Indicators N Mean SD Skewness Kurtosis

Stats Stats SE Stats Stats SE Stats SE

Chineseness

v1 230 4.55 .12 1.80 -.35 .16 -.84 .32

v2 230 4.11 .12 1.80 -.08 .16 -.92 .32

v3 230 4.12 .12 1.75 -.06 .16 -.82 .32

v4 230 4.40 .12 1.84 -.29 .16 -.95 .32

v5 230 4.37 .11 1.68 -.15 .16 -.72 .32

v6 230 3.62 .12 1.89 .20 .16 -1.04 .32

v7 230 4.26 .12 1.81 -.14 .16 -.97 .32

v8 230 3.68 .12 1.88 .15 .16 -1.07 .32

v9 230 4.18 .11 1.63 -.08 .16 -.67 .32

Economic

capital

v11 230 3.96 .12 1.82 .01 .16 -.98 .32

v12 230 4.02 .12 1.88 .03 .16 -1.08 .32

v13 230 3.88 .12 1.89 .10 .16 -1.09 .32

v14 230 4.22 .12 1.85 -.12 .16 -1.01 .32

Cultural

capital

v15 230 4.69 .11 1.74 -.42 .16 -.72 .32

v16 230 4.20 .12 1.84 -.11 .16 -.98 .32

v17 230 3.67 .12 1.87 .18 .16 -1.03 .32

v18 230 3.54 .12 1.83 .23 .16 -.98 .32

v19 230 3.35 .13 1.95 .40 .16 -1.06 .32

v20 230 3.52 .13 1.96 .30 .16 -1.12 .32

Social capital

v23 230 4.19 .12 1.79 -.07 .16 -.92 .32

v24 230 3.74 .13 1.92 .12 .16 -1.10 .32

v25 230 3.53 .12 1.76 .23 .16 -.90 .32

v26 230 4.07 .13 1.91 -.05 .16 -1.10 .32

Symbolic

capital

v30 230 3.36 .12 1.80 .31 .16 -.94 .32

v31 230 3.87 .12 1.88 .05 .16 -1.08 .32

v32 230 3.59 .12 1.83 .19 .16 -1.00 .32

v33 230 4.29 .12 1.77 -.12 .16 -.90 .32

v34 230 4.12 .11 1.71 -.06 .16 -.76 .32

CHL

proficiency

v35 230 5.16 .12 1.76 -.58 .16 -.90 .32

v36 230 4.40 .14 2.07 -.25 .16 -1.27 .32

v37 230 4.31 .13 2.00 -.14 .16 -1.26 .32

v38 230 3.82 .14 2.07 .12 .16 -1.31 .32

v39 230 4.04 .14 2.09 -.03 .16 -1.35 .32

v40 230 3.13 .13 2.01 .59 .16 -.94 .32

v41 230 3.63 .13 1.94 .26 .16 -1.10 .32

v42 230 3.06 .13 1.92 .65 .16 -.75 .32

5.2.1 Validity and reliability of measurement models: EFA and CFA

As EFA is being used in a descriptive way to summarise relationships, assumptions in regards

to normal distribution may be relaxed as long as the deviation is not too large (Tabachnick &

Fidell, 2007). Where possible a single factor congeneric measurement model was then

constructed for each latent variable. CFA was then conducted to demonstrate the degree to

173

which the indicators were reflective of their corresponding constructs. It reflected how well

the indicators captured the essence of the corresponding latent variables. Multivariate

normality was checked where it was required for the particular analysis adopted. When

multivariate normality was violated, bootstrapping techniques were applied to minimise the

impact of the non-normal distribution of the data. It is usual in a measurement model to set

the scale of the latent variable by fixing the variance of the construct to 1 rather than the usual

practice of setting a factor loading to 1. In this way, standard error estimates can be generated

for all factor loadings, which is of interest in a measurement model. However, this may

generate incorrect standard errors when bootstrapping techniques are applied. Consequently,

the factor loading of the indicator with the highest factor loading in EFA was set to 1. The

sections below report on the process of EFA and CFA that shape the instrument and improve

the reliability and validity of the measurement models.

‘Chineseness’

It will be recalled that Chapter Three introduced a core set of Confucian values consisting of

Wuchang (五常, Five Constants), Sizi (四字, Four Virtues), and Liuyi (六艺, Six Skills).

These Confucian values serve as the cultural and historical foundation of the habitus of

‘Chineseness’. As argued in Chapter Four, this habitus of ‘Chineseness’ can be quantified by

operationalising key dimensions of Confucian norms, ideals, and social perspectives

documented in Confucian classics and popular classical sayings. Originally, nine indicators,

reflective of durable and transposable dispositions rooted in Confucianism, were used to

measure the habitus of ‘Chineseness’. This section discusses whether the nine indicators fit

into the measurement model for the construct ‘Chineseness’, then reports on the modification

of the measurement model to improve the model fit.

174

Reflective models require intercorrelation between indicators. As such, the correlation

matrix was examined in the first step of EFA. It is desirable that the inter-item correlations

should be greater than .30 to demonstrate enough intercorrelation but less than .90 to avoid

multicollinearity (Field, 2009). The correlation matrix for indicators corresponding to

‘Chineseness’ demonstrated that all the indicators had a statistically significant small to

medium correlation with other indicators (p < .001 as shown in Table 5.5), varying within the

cutoff range between .30 and .90.

Table 5.5

Correlation matrix for the indicators within the construct ‘Chineseness’

Correlation v1 v2 v3 v4 v5 v6 v7 v8 v9

v1 1.00 .54 .52 .52 .49 .41 .43 .50 .49

v2 .54 1.00 .57 .62 .44 .54 .61 .56 .50

v3 .52 .57 1.00 .55 .43 .51 .48 .47 .51

v4 .52 .62 .55 1.00 .47 .48 .60 .57 .52

v5 .49 .44 .43 .47 1.00 .37 .49 .45 .52

v6 .41 .54 .51 .48 .37 1.00 .58 .66 .47

v7 .43 .61 .48 .60 .49 .58 1.00 .59 .59

v8 .50 .56 .47 .57 .45 .66 .59 1.00 .54

v9 .49 .50 .51 .52 .52 .47 .59 .54 1.00

Sig. v1 v2 v3 v4 v5 v6 v7 v8 v9

v1 .000 .000 .000 .000 .000 .000 .000 .000

v2 .000 .000 .000 .000 .000 .000 .000 .000

v3 .000 .000 .000 .000 .000 .000 .000 .000

v4 .000 .000 .000 .000 .000 .000 .000 .000

v5 .000 .000 .000 .000 .000 .000 .000 .000

v6 .000 .000 .000 .000 .000 .000 .000 .000

v7 .000 .000 .000 .000 .000 .000 .000 .000

v8 .000 .000 .000 .000 .000 .000 .000 .000

v9 .000 .000 .000 .000 .000 .000 .000 .000

The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy is an index for

comparing the magnitude of the observed correlation coefficients to the magnitude of the

partial correlation coefficients. When the sum of the squared observed correlation coefficients

is remarkably larger than the sum of the squared partial correlation coefficients, the value of

KMO approaches 1. The higher the KMO value, the more compact the correlation of the

indicators. The KMO measure should be greater than .50 for a satisfactory factor analysis to

175

proceed (Kaiser, 1974). The KMO measure of sampling adequacy was .92, which

demonstrated that the EFA should yield distinct and reliable factors (see Table 5.6).

Bartlett’s test of sphericity is used to test the null hypothesis that there is no statistically

significant difference between the observed correlation matrix and the identity matrix where

the indicators in the correlation matrix are uncorrelated. The result of the Bartlett’s test of

sphericity rejected the null hypothesis and confirmed the adequacy of the magnitude of the

correlations by presenting a statistically significant chi-square value of 1028.82 (p < .001 as

shown in Table 5.6).

Table 5.6

KMO and Bartlett's test for the indicators within the construct ‘Chineseness’

KMO and Bartlett's test

Kaiser-Meyer-Olkin measure of sampling adequacy. .92

Bartlett's test of sphericity Approx. chi-square 1028.82

df 36

Sig. .000

An examination of communalities indicated that a reasonable level of variation of the

indicators (between 39% and 60% as highlighted in Table 5.7) was explained by the construct

‘Chineseness’. The initial communalities demonstrated how much variance the indicator had

in common with other indicators. The extraction communalities demonstrated how much

variance the indicator had in common with the extracted factors in comparison to the variance

explained by a linear combination of the other indicators. The communalities demonstrated

that the extraction in most cases explained more variance in an indicator than the initial

model. This indicated the appropriateness of the solution for the current factor extraction.

However, of concern were v1, v5, and v6, whose variances explained by the extraction were

not more than those explained by the initial model. These indicators will be examined more

closely subsequently.

176

Table 5.7

Communalities for the indicators within the construct ‘Chineseness’

Initial Extraction

v1 .44 .44

v2 .56 .60

v3 .47 .49

v4 .54 .58

v5 .39 .39

v6 .52 .50

v7 .57 .60

v8 .57 .58

v9 .48 .51

Extraction method: Maximum likelihood

Eigenvalues represent the amount of variation explained by a factor and an eigenvalue

of 1 or greater represents a substantial amount of variation (Kaiser, 1960). Of all the nine

impact factors for ‘Chineseness’, there was only one factor with an eigenvalue above 1.

Therefore, only one factor was extracted. This factor accounted for 57.18% of the total

variance of the nine indicators (see Table 5.8).

Table 5.8

Total variance of the construct ‘Chineseness’ explained by its indicators

Factor Initial eigenvalues Extraction sums of squared loadings

Total % of Variance Cumulative % Total % of Variance Cumulative %

1 5.15 57.18 57.18 4.67 51.92 51.92

2 .75 8.32 65.50

3 .63 6.98 72.48

4 .53 5.86 78.35

5 .50 5.52 83.87

6 .45 5.02 88.90

7 .39 4.34 93.24

8 .32 3.56 96.80

9 .29 3.20 100.00


Examination of the scree plot also indicated the appropriateness of a one-factor

solution. The cutoff point for selecting the extracted factors should be at the point of inflexion

where the slope of the line changes dramatically (Cattell, 1966). The number of the extracted

factor(s) depends on the number of inflexion point(s). In Figure 5.7, the point of inflexion

177

occurred at the second data point, where there was a sharp descent in the curve followed by a

tailing off. Therefore, only one factor was extracted.

Figure 5.7. Scree plot of the ‘Chineseness’ scale

Before conducting CFA, univariate and multivariate normality were checked. As

highlighted in Table 5.9, all the individual indicators departed significantly from univariate

normal distribution by displaying the critical ratio of skewness, kurtosis, or both beyond the

cutoff value of |2| (Field, 2009). The value of Mardia’s coefficient was 16.14, which was

greater than the cutoff value of 3 (Yuan et al., 2002). This indicated significant deviation from

multivariate normality. Therefore, bootstrapping techniques were applied when conducting

CFA.

Table 5.9

Assessment of normality for the indicators within the construct ‘Chineseness’

Variable min max skew c.r. kurtosis c.r.

v9 1.00 7.00 -0.08 -0.52 -0.68 -2.10

v8 1.00 7.00 0.15 0.90 -1.07 -3.31

v7 1.00 7.00 -0.14 -0.87 -0.97 -3.01

v6 1.00 7.00 0.20 1.24 -1.04 -3.23

v5 1.00 7.00 -0.15 -0.90 -0.73 -2.27

v4 1.00 7.00 -0.29 -1.81 -0.95 -2.95

v3 1.00 7.00 -0.06 -0.37 -0.83 -2.56

v2 1.00 7.00 -0.08 -0.52 -0.93 -2.87

v1 1.00 7.00 -0.35 -2.18 -0.85 -2.62

Multivariate 16.14 8.70

178

A single-factor measurement model for the construct ‘Chineseness’ was specified as a

latent variable with nine indicators. The model with its standardised parameters is illustrated

in Figure 5.8.

Figure 5.8. Measurement model for the construct ‘Chineseness’

As highlighted in Table 5.10, the values of all the baseline comparisons fit indices (NFI,

RFI, IFI, TLI, and CFI) were above the cutoff value of .90 (Bentler, 1990), but the significant

chi-square value (CMIN = 66.91, p < .001) and the value of RMSEA (.08), which reached the

cutoff value of .08 (R. Ho, 2006), indicated that the model fit could be improved.

Table 5.10

Model fit indices for the construct ‘Chineseness’

CMIN

Model NPAR CMIN DF P CMIN/DF

Default model 18 66.91 27 .000 2.48 Saturated model 45 .000 0

Independence model 9 1046.34 36 .000 29.07 Baseline comparisons

Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 0.94 0.92 0.96 0.95 0.96

Saturated model 1.00 1.00 1.00

Independence model 0.00 0.00 0.00 0.00 0.00

RMSEA

Model RMSEA LO 90 HI 90 PCLOSE

Default model 0.08 0.06 0.11 0.02

Independence model 0.35 0.33 0.37 0.00

179

An examination of the standardised residual covariance matrix (see Table 5.11) and the

modification index (see Table 5.12) indicated that the model fit can be improved by deleting

v1, v5, v6, and v8 from the model. These indicators were either associated with high values

of standardised residual covariances, or associated with high values of covariances between

indicator error terms or regression weights between indicators (as highlighted in Tables 5.11

and 5.12). It will be recalled that v1, v5, and v6 were also of concern in EFA. The

questionnaire items corresponding to these indicators were revisited. On reflection, these four

items did appear different from the others because they directly scrutinised the respondents’

personal viewpoints. In contrast, the other items scrutinised respondents’ viewpoints about a

general statement. By virtue of the reflective nature of the model, deleting indicators will not

necessarily change the conceptual domain of the theoretical construct (Coltman et al., 2008).

Therefore, these four indicators were tentatively excluded from the model.

Table 5.11

Standardised residual covariances for the indicators within the construct ‘Chineseness’

v9 v8 v7 v6 v5 v4 v3 v2 v1

v9 0.00

v8 -0.03 0.00

v7 0.46 0.08 0.00

v6 -0.45 1.61 0.47 0.00

v5 1.03 -0.35 0.20 -1.00 0.00

v4 -0.25 -0.12 0.16 -0.82 -0.03 0.00

v3 0.19 -0.75 -0.73 0.23 -0.01 0.29 0.00

v2 -0.63 -0.37 0.19 -0.05 -0.50 0.39 0.44 0.00

v1 0.17 -0.05 -1.15 -0.78 1.05 0.24 0.70 0.39 0.00

180

Table 5.12

Modification indices for the construct ‘Chineseness’

Covariances Regression weights

M.I. Par change

e6 <--> e8 19.36 0.53

e5 <--> e9 4.82 0.23

e5 <--> e6 4.51 -0.26

e4 <--> e6 5.02 -0.26

e3 <--> e8 4.12 -0.23

e3 <--> e7 4.14 -0.22

e1 <--> e7 9.17 -0.34

e1 <--> e5 4.29 0.26

Before testing the fit of the modified model, multivariate normality was checked again

because four indicators had been dropped from the model (see Table 5.13). The Mardia’s

coefficient value of 3.59, which was greater than the cutoff value of 3.00 (Yuan et al., 2002),

indicated the violation of multivariate normality. Therefore, the model was tested for fit by

using bootstrapping techniques.

Table 5.13

Assessment of normality for the indicators within the construct ‘Chineseness’ (modified)


v9 1.00 7.00 -0.08 -0.52 -0.68 -2.10

v7 1.00 7.00 -0.14 -0.87 -0.97 -3.01

v4 1.00 7.00 -0.29 -1.81 -0.95 -2.95

v3 1.00 7.00 -0.06 -0.37 -0.83 -2.56

v2 1.00 7.00 -0.08 -0.52 -0.93 -2.87


A modified measurement model for the construct ‘Chineseness’ was specified as a

latent variable with five indicators. The modified model with standardised parameters is

illustrated in Figure 5.9.

M.I. Par change

v8 <--- v6 8.96 0.14

v7 <--- v1 4.80 -0.10

v6 <--- v8 7.28 0.13

181

Figure 5.9. Modified measurement model for the construct ‘Chineseness’

The modified model fitted the data well (see Table 5.14). It presented a non-significant

chi-square value of 10.54 (p = .061). All values for baseline comparison indices were above

the cutoff .90 (Bentler, 1990) and RMSEA value of .07 was below the cutoff .08 (R. Ho,

2006). The unstandardised regression weight of each indicator corresponded to a significant

critical ratio. All the values of the standardised regression weight (see Figure 5.9) were above

the cutoff value of .50 (Hair et al., 2006) and were above or the same as the preferred value

of .70 (Hair et al., 2006). This indicated that all five items were significant reflective

indicators of the corresponding construct ‘Chineseness’. The model converged.

Table 5.14

Model fit indices for the construct ‘Chineseness’ (modified)

CMIN


Default model 10 10.54 5 .061 2.11

Saturated model 15 0.00 0

Independence model 5 494.08 10 .000 49.41

Baseline comparisons

Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 0.98 0.96 0.99 0.98 0.99



RMSEA


Default model .07 .000 .13 .24 Independence model .46 .43 .50 .00

182

The modified model had a significantly better fit than the original model. The modified

model had a chi-square difference of 56.37 (66.91-10.54) from the original model. With a

difference in degrees of freedom of 22 (27-5), this chi-square difference was statistically

significant (p < .001).

The factor score weight and the proportional factor score weight are reported in Table

5.15. The proportional factor score weight of each indicator was calculated as the factor score

weight of each indicator divided by the sum of the factor score weight (Rowe, 2002).

Table 5.15

Proportional factor score weights for the indicators within the construct ‘Chineseness’

v9 v7 v4 v3 v2 Sum

Factor score weight .17 .20 .20 .15 .22 .94

Proportional factor score weight .18 .21 .21 .16 .23 .99

Taking account of individual and joint measurement error, the scale score for the

construct can be computed as a continuous variable by multiplying the individual’s raw score

on each indicator by the proportionally weighted factor score of each indicator and summing

(Rowe, 2002). The scale score for the construct ‘Chineseness’ then can be calculated by the

following equation:

v9 × .18 + v7 × .21 + v4 × .21 + v3 × .16 + v2 × .23

This approach ensures that the estimates of the scale score adjusted for measurement

error are proportionally weighted by the actual contribution made by each indicator (Rowe,

2002). The proportional factor score sums to 1 hence the composite score will range from a

minimum of 1 to a maximum of 7. This process ensures that the construct has the same

‘metric’ as that of the indicators for the construct (Rowe, 2002). In Table 5.15, the

proportional factor score weights amounted to .99 due to rounding errors.

183

Reliability and validity of the construct ‘Chineseness’ were further checked. The

following sections reported four measures of reliability, namely Squared Multiple

Correlations (SMCs), construct reliability, variance extracted, and coefficient H, as well as

two measures of validity, namely construct validity and convergent validity.

The SMC for an indicator represents the proportion of variance in the indicator that is

explained by the construct. The SMC for an indicator greater than .30 is considered

acceptable while the SMC greater than .50 is preferred for a construct to have a good

mapping of that indicator (Jöreskog & Sörbom, 1996). As demonstrated in Figure 5.9, the

SMCs for all the five indicators were above the cutoff value of .30 and most of them were

above the preferred value of .50. The exceptions were the values of .49 and .48, which were

only marginally lower than .50. Therefore, the construct ‘Chineseness’ accounted for a

reasonable level of the variance of its indicators. The reliability of all the indicators of the

construct ‘Chineseness’ represented by the SMC measures was good.

The construct reliability compares the variance of the indicators captured by the

construct with that due to the measurement errors. The computation of the construct

reliability was offered by Fornell and Larcker (1981) as follows:

2 2

1 1 1

/ ( )n n n

i i i

i i i

r r e

In the formula above, n represents the number of indicators, r represents the standardised

regression weight of the indicator, and e represents the measurement error associated with the

indicator. The cutoff value for the construct reliability is .50 (Fornell & Larcker, 1981), which

indicates that the variance of the indicators captured by the construct exactly equals that due

to the measurement errors. As such, the greater the value is larger than .50, the better the

construct reliability. The construct reliability for ‘Chineseness’ was .56, which indicated that

the variance of the indicators captured by the construct ‘Chineseness’ was greater than that

due to the measurement errors. The construct reliability was acceptable.

184

The variance extracted expresses the overall amount of variance in the indicators

accounted for by the construct. Variance extracted exceeding 50% is commonly considered a

good measure of reliability (Fornell & Larcker, 1981). The construct ‘Chineseness’ accounted

for 64.56% of the overall variation in the indicators. Therefore, the reliability of the indicators

of the construct ‘Chineseness’ represented by the variance extracted was good.

Coefficient H (G. R. Hancock & Mueller, 2001) is becoming a popular way of

reporting reliability for SEM. The computation of coefficient H was offered by G. R.

Hancock and Mueller (2001) as follows:

2

21

1/ (1 1/ )1

ni

i i

r

r

In the formula above, n represents the number of indicators and r represents the standardised

regression weight of the indicator. A coefficient H value above .70 is desirable to achieve

reasonable reliability (G. R. Hancock & Mueller, 2001). The coefficient H value for the

construct ‘Chineseness’ was .87, which indicated a high reliability.

Construct validity requires the unidimensionality of the indicators. The model fit

measures can be viewed as confirming construct validity. The modified measurement model

for the construct ‘Chineseness’ had a good model fit, which supported the claim for its

construct validity.

Convergent validity, operationalised through the factor loading, is a measure of the

direct structural relationship between an indicator and the construct. To achieve convergent

validity, the factor loadings must be significantly different from 0. The critical ratios of the

unstandardised regression weights of the indicators are used to test this significance. As stated

earlier, the unstandardised regression weight of all the indicators of the construct

‘Chineseness’ corresponded to a significant critical ratio. Therefore, all the indicators

demonstrated a good convergent validity.

185

Economic capital

Four indicators were used to operationalise the construct ‘economic capital’. These four

indicators were reflective of the investment of financial resources in (potential) activities

associated with CHL learning. The correlation matrix demonstrated that all the indicators had

a statistically significant medium to strong correlation with other indicators (p < .001 as

shown in Table 5.16), varying within the cutoff range between .30 and .90 (Field, 2009). This

suggested the basis for EFA.

Table 5.16

Correlation matrix for the indicators within the construct ‘economic capital’

Correlation v11 v12 v13 v14

v11 1.00 .63 .59 .64

v12 .63 1.00 .80 .72

v13 .59 .80 1.00 .72

v14 .64 .72 .72 1.00

Sig. v11 v12 v13 v14

v11 .000 .000 .000

v12 .000 .000 .000

v13 .000 .000 .000

v14 .000 .000 .000

As shown in Table 5.17, the KMO measure of sampling adequacy (.83) was well above

the cutoff value of .50 (Kaiser, 1974), which demonstrated the EFA should yield distinct and

reliable factors. Bartlett’s test of sphericity confirmed the adequacy of the magnitude of the

correlations by presenting a statistically significant chi-square value of 573.18 (p < .001).

Table 5.17

KMO and Bartlett's test for the indicators within the construct ‘economic capital’




df 6

Sig. .000

186

An examination of communalities indicated that a reasonable level of variation of the

indicators (between 50% and 80% as highlighted in Table 5.18) was explained by the

construct ‘economic capital’. Specifically, the extraction in all cases explained more variance

in an indicator than the initial model (see Table 5.18). This indicated the appropriateness of

the solution for the current factor extraction.

Table 5.18

Communalities for the indicators within the construct ‘economic capital’

Initial Extraction

v11 .47 .50

v12 .70 .80

v13 .69 .78

v14 .62 .68


Of all the four impact factors for ‘economic capital’, there was only one factor with an

eigenvalue above 1. Consequently, only one factor was extracted. This factor accounted for

76.46% of the total variance of the four indicators (see Table 5.19). Examination of the scree

plot also indicated the appropriateness of a one-factor solution (see Figure 5.10).

Table 5.19

Total variance of the construct ‘economic capital’ explained by its indicators



1 3.06 76.46 76.46 2.76 68.95 68.95

2 .45 11.28 87.74

3 .29 7.30 95.04

4 .20 4.96 100.00


187

Figure 5.10. Scree plot of the ‘economic capital’ scale

As highlighted in Table 5.20, all the individual indicators departed significantly from

univariate normal distribution by displaying the critical ratio of kurtosis beyond the cutoff

value of |2| (Field, 2009). Mardia’s coefficient value of 8.21was greater than the cutoff value

of 3.00 (Yuan et al., 2002). This indicated significant deviation from multivariate normality.

Therefore, bootstrapping techniques were applied when conducting CFA.

Table 5.20

Assessment of normality for the indicators within the construct ‘economic capital’


v14 1.00 7.00 -0.12 -0.73 -1.02 -3.14

v13 1.00 7.00 0.10 0.59 -1.10 -3.39

v12 1.00 7.00 0.03 0.20 -1.08 -3.35

v11 1.00 7.00 0.01 0.05 -0.98 -3.04


A single-factor measurement model for the construct ‘economic capital’ was specified

as a latent variable with four indicators. The model with standardised parameters is illustrated

in Figure 5.11.

188

Figure 5.11. Measurement model for the construct ‘economic capital’

Although the model had a significant chi-square value of 8.01 (p = .018), the ratio of

chi-square value and degree of freedom (4.00) was not very high. Although the RMSEA

value of .12 was above the cutoff value of .08 (R. Ho, 2006), all the baseline comparison fit

indices were above the cutoff value of .90 (Bentler, 1990). In particular, CFI and TLI, the two

most widely referenced indices (Hair et al., 2006), had relatively high values of .99 and .97

respectively (see Table 5.21). The unstandardised regression weight of each indicator

corresponded to a significant critical ratio. All the values of the standardised regression

weight (see Figure 5.11) were above the cutoff value of .50 (Hair et al., 2006) and the

preferred value of .70 (Hair et al., 2006). This indicated that all the four corresponding items

were significant reflective indicators of the respective construct ‘economic capital’. Therefore,

it can be argued that the model had an acceptable fit.

189

Table 5.21

Model fit indices for the construct ‘economic capital’

CMIN


Default model 8 8.01 2 .018 4.00

Saturated model 10 .000 0


Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 0.99 0.96 0.99 0.97 0.99



RMSEA


Default model 0.12 0.04 0.20 0.07


The proportional factor score weight of each indicator is demonstrated in Table 5.22.

Table 5.22

Proportional factor score weights for the indicators within the construct ‘economic capital’

v14 v13 v12 v11 Sum

Factor score weight .20 .30 .34 .11 .95

Proportional factor score weight .21 .31 .36 .12 1.00

The scale score for the construct ‘economic capital’ can then be calculated by the

following equation:

v14 × .21 + v13 × .31 + v12 × .36 + v11 × .12

Four measures of reliability, namely SMC, construct reliability, variance extracted, and

coefficient H, as well as two measures of validity, namely construct validity and convergent

validity, are reported below.

As demonstrated in Figure 5.11, the SMCs for all the four indicators were above the

cutoff value of .30 (Jöreskog & Sörbom, 1996) and were the same as, or above, the preferred

value of .50 (Jöreskog & Sörbom, 1996). Therefore, the construct ‘economic capital’

accounted for a reasonable level of the variance of its indicators. The reliability of all the

190

indicators of the construct ‘economic capital’ represented by the SMC measures was

considered good. The construct reliability for ‘economic capital’ was .69, which was greater

than the cutoff value of .50 (Fornell & Larcker, 1981). The variance of the indicators captured

by the construct ‘economic capital’ was greater than that due to the measurement errors. This

indicated that the construct reliability was acceptable. The construct ‘economic capital’

accounted for 76.46% of the overall variation in the four indicators. This was well above the

cutoff of 50% (Fornell & Larcker, 1981). Therefore, the reliability of the indicators of the

construct ‘economic capital’ represented by the variance extracted was good. The coefficient

H value for the construct ‘economic capital’ was .91, which was well above the cutoff value

of .70 (G. R. Hancock & Mueller, 2001). This represented a high reliability.

All the four indicators of the construct ‘economic capital’ loaded on one factor, which

indicated the unidimensionality of the four indicators. This supported the claim for its

construct validity. As stated earlier, the unstandardised regression weight of all the indicators

of the construct ‘economic capital’ corresponded to a significant critical ratio. Therefore, all

the indicators demonstrated a good convergent validity.

Cultural capital

Originally, six indicators were used to operationalise the construct ‘cultural capital’. These

indicators were reflective of the cultural dispositions in relation to CHL learning and ‘taste’

associated with various Chinese cultural activities. This section discusses whether the six

indicators fit into the measurement model for the construct ‘cultural capital’ and reports on

the modification of the measurement model to improve the model fit.

The correlation matrix demonstrated that all indicators had a statistically significant

medium correlation with other indicators (p < .001 as shown in Table 5.23), varying within

191

the cutoff range between .30 and .90 (Field, 2009). The strength of the intercorrelation was

considered enough, without the problem of multicollinearity. This suggested the basis for

EFA.

Table 5.23

Correlation matrix for the indicators within the construct ‘cultural capital’

Correlation v15 v16 v17 v18 v19 v20

v15 1.00 .61 .51 .45 .54 .56

v16 .61 1.00 .66 .63 .57 .63

v17 .51 .66 1.00 .67 .65 .55

v18 .45 .63 .67 1.00 .72 .66

v19 .54 .57 .65 .72 1.00 .66

v20 .56 .63 .55 .66 .66 1.00

Sig. v15 v16 v17 v18 v19 v20

v15 .000 .000 .000 .000 .000

v16 .000 .000 .000 .000 .000

v17 .000 .000 .000 .000 .000

v18 .000 .000 .000 .000 .000

v19 .000 .000 .000 .000 .000

v20 .000 .000 .000 .000 .000


the cutoff value of .50 (Kaiser, 1974), which demonstrated that the EFA should yield distinct

and reliable factors. Bartlett’s test of sphericity verified the adequacy of the magnitude of the


Table 5.24

KMO and Bartlett's test for the indicators within the construct ‘cultural capital’




df 15

Sig. .000

An examination of communalities indicated that a reasonable level of variation in the


192

construct ‘cultural capital’. The communalities in Table 5.25 demonstrated that the extraction

in most cases explained more variance in an indicator than the initial model. This indicated

the appropriateness of the solution for the current factor extraction. However, of concern was

v15, whose variance explained by the extraction was lower than that explained by the initial

model. This indicator will be examined more closely subsequently.

Table 5.25

Communalities for the indicators within the construct ‘cultural capital’

Initial Extraction

v15 .46 .43

v16 .60 .61

v17 .58 .62

v18 .65 .69

v19 .62 .67

v20 .58 .62


There was only one factor with an eigenvalue above 1, accounting for 67.19% of the

total variance of the six indicators (see Table 5.26). Consequently, only one factor was

extracted. The scree plot also supported a one-factor solution (see Figure 5.12).

Table 5.26

Total variance of the construct ‘cultural capital’ explained by its indicators



1 4.03 67.19 67.19 3.64 60.74 60.74

2 .62 10.27 77.46

3 .46 7.65 85.11

4 .39 6.48 91.59

5 .27 4.43 96.03

6 .24 3.97 100.00


193

Figure 5.12. Scree plot of the ‘cultural capital’ scale


univariate normal distribution by displaying the critical ratio of skewness, kurtosis, or both

beyond the cutoff value of |2| (Field, 2009). Mardia’s coefficient value (8.26) was higher than

the cutoff value of 3.00 (Yuan et al., 2002), indicating significant deviation from multivariate

normality. Therefore, bootstrapping techniques were applied when conducting CFA.

Table 5.27

Assessment of normality for the indicators within the construct ‘cultural capital’


v20 1.00 7.00 0.30 1.84 -1.12 -3.46

v19 1.00 7.00 0.39 2.44 -1.07 -3.30

v18 1.00 7.00 0.23 1.40 -0.98 -3.04

v17 1.00 7.00 0.18 1.12 -1.03 -3.19

v16 1.00 7.00 -0.11 -0.67 -0.99 -3.05

v15 1.00 7.00 -0.42 -2.58 -0.73 -2.27


A single-factor measurement model for the construct ‘cultural capital’ was specified as

a latent variable with six indicators. The model with standardised parameters is illustrated in

Figure 5.13.

194

Figure 5.13. Measurement model for the construct ‘cultural capital’

The model had a significant chi-square value of 52.48. The value of RFI (.89) did not

reach the cutoff value of .90 (Bentler, 1990). The value of RMSEA (.15) was much higher

than the cutoff value of .08 (R. Ho, 2006). These all indicated that the model did not fit the

data well (see Table 5.28).

Table 5.28

Model fit indices for the construct ‘cultural capital’

CMIN




Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model .94 .89 .95 .91 .95


Independence model .00 .00 .00 .00 .00 RMSEA




modification index (see Table 5.30) indicated the model fit could be improved by deleting

v15 and v16 from the model. These two indicators were associated with a high value of

standardised residual covariances, a high value of covariances between indicator error terms,

and high values of regression weights between indicators (as highlighted in Table 5.29 and

195

5.30). It will be recalled that v15 was also of concern in EFA. Therefore, v15 and v16 were

tentatively excluded from the model.

Table 5.29

Standardised residual covariances for the indicators within the construct ‘cultural capital’

v20 v19 v18 v17 v16 v15

v20 0.00

v19 0.17 0.00

v18 0.11 0.52 0.00

v17 -0.87 0.04 0.25 0.00

v16 0.24 -0.92 -0.29 0.55 0.00

v15 0.55 -0.01 -1.34 -0.07 1.29 0.00

Table 5.30

Modification indices for the construct ‘cultural capital’

Covariances Regression weight

M.I. Par change M.I. Par change

e18<-->e19 5.64 0.22 v19<---v16 4.44 -0.09

e17<-->e20 9.81 -0.33 v18<---v15 9.28 -0.13

e16<-->e19 12.84 -0.36 v16<---v15 6.36 0.12

e15<-->e18 17.31 -0.42 v15<---v18 4.41 -0.10

e15<-->e16 11.93 0.38 v15<---v16 4.02 0.10

Before testing the fit of the modified model, multivariate normality was checked.

Mardia’s coefficient was 3.60, which was greater than the cutoff value of 3.00 (Yuan et al.,

2002). This implied the violation of the multivariate normality (see Table 5.31). Therefore,

bootstrapping was applied when conducting CFA.

Table 5.31

Assessment of normality for the indicators within the construct ‘cultural capital’ (modified)


v20 1.00 7.00 0.30 1.84 -1.12 -3.46

v19 1.00 7.00 0.39 2.44 -1.07 -3.30

v18 1.00 7.00 0.23 1.40 -0.98 -3.04

v17 1.00 7.00 0.18 1.12 -1.03 -3.19


196

A modified measurement model for the construct ‘cultural capital’ was specified as a

latent variable with four indicators. The model with standardised parameters is illustrated in

Figure 5.14.

Figure 5.14. Modified measurement model for the construct ‘cultural capital’

The modified model had a non-significant chi-square value of 2.24 (p = .34). All the

baseline comparisons indices (NFI, RFI, IFI, TLI, and CFI) were well above the cutoff value

of .90 (Bentler, 1990). The value of RMSEA was .02, which was far below the cutoff value

of .08 (R. Ho, 2006). These implied a good model fit (see Table 5.32). The unstandardised

regression weight of each indicator corresponded to a significant critical ratio. All the values

of the standardised regression weight (see Figure 5.14) were above the cutoff value of .50

(Hair et al., 2006) and the preferred value of .70 (Hair et al., 2006). This indicated that all the

four corresponding items were significant reflective indicators of the respective construct

‘cultural capital’. The model converged.

197

Table 5.32

Model fit indices for the construct ‘cultural capital’ (modified)

CMIN




Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 1.00 0.99 1.00 1.00 1.00



RMSEA


Default model 0.02 0.00 0.14 0.51


Compared to the original model, the modified model had a significantly better fit. The

modified model decreased the chi-square value of 50.24 (52.48-2.24) from the original model.

With the degree of freedom difference of 7 (9-2), the chi-square difference was statistically

significant (p < .001).


Table 5.33

Proportional factor score weights for the indicators within the construct ‘cultural capital’

v20 v19 v18 v17 Sum



The scale score for the construct ‘cultural capital’ then can be calculated by the

following equation:

v20 × .18 + v19 × .29 + v18 × .35 + v17 × .19

Reliability and validity of the construct ‘cultural capital’ were further checked by four

measures of reliability, namely SMC, construct reliability, variance extracted, and coefficient

198

H, as well as two measures of validity, namely construct validity and convergent validity.

These measures are reported below.


cutoff value of .30 (Jöreskog & Sörbom, 1996) and the preferred value of .50 (Jöreskog &

Sörbom, 1996). Therefore, the construct ‘cultural capital’ accounted for a reasonable level of

the variance of its indicators. The reliability of all the indicators of the construct ‘cultural

capital’ represented by the SMC measures was good. The construct reliability for ‘cultural

capital’ was .66, which was greater than the cutoff value of .50 (Fornell & Larcker, 1981).

The variance of the indicators captured by the construct ‘cultural capital’ was greater than

that due to the measurement errors. This indicated that the construct reliability was acceptable.

The construct ‘cultural capital’ accounted for 74.02% of the overall variation in the four

indicators. This was well above the cutoff of 50% (Fornell & Larcker, 1981). Therefore, the

reliability of the indicators of the construct ‘cultural capital’ represented by the variance

extracted was good. The coefficient H value for the construct ‘cultural capital’ was .89, which

was well above the cutoff value of .70 (G. R. Hancock & Mueller, 2001). This represented a

high reliability.

The modified measurement model for the construct ‘cultural capital’ had a good model

fit, which indicated the unidimensionality of the four indicators. This supported the claim for

its construct validity. As stated earlier, the unstandardised regression weight of all the

indicators of the construct ‘cultural capital’ corresponded to a significant critical ratio.

Therefore, all the indicators demonstrated a good convergent validity.

199

Social capital

Four indicators were used to operationalise the construct ‘social capital’. These indicators

were reflective of the size and strength of the social network consisting of members of

Chinese descent. This section discusses whether the four indicators fit into the measurement

model for the construct ‘social capital’.

The correlation matrix demonstrated that all the indicators had a statistically significant


the cutoff range between .30 and .90 (Field, 2009). This suggested the basis for EFA.

Table 5.34

Correlation matrix for the indicators within the construct ‘social capital’

Correlation v23 v24 v25 v26

v23 1.00 .63 .68 .59

v24 .63 1.00 .59 .54

v25 .68 .59 1.00 .55

v26 .59 .54 .55 1.00

Sig. v23 v24 v25 v26

v23 .000 .000 .000

v24 .000 .000 .000

v25 .000 .000 .000

v26 .000 .000 .000


the cutoff value of .50 (Kaiser, 1974), which demonstrated the EFA should yield distinct and



Table 5.35

KMO and Bartlett's test for the indicators within the construct ‘social capital’



df 6

Sig. .000

200

An investigation of communalities indicated that a reasonable level of variance in the


construct ‘social capital’. The extraction in all cases explained more variance in an indicator

than the initial model (see Table 5.36). This indicated the appropriateness of the solution for

the current factor extraction.

Table 5.36

Communalities for the indicators within the construct ‘social capital’

Initial Extraction

v23 .57 .72

v24 .48 .57

v25 .52 .63

v26 .42 .50



total variance of the four indicators (see Table 5.37). Consequently, only one factor was


Table 5.37

Total variance of the construct ‘social capital’ explained by its indicators



1 2.80 69.99 69.99 2.41 60.25 60.25

2 .47 11.84 81.83

3 .41 10.32 92.16

4 .31 7.84 100.00


201

Figure 5.15. Scree plot of the ‘social capital’ scale



value of |2| (Field, 2009). The value of Mardia’s coefficient was 4.75, which was higher than

the suggested value of 3.00 (Yuan et al., 2002). This indicated that the multivariate normality

was violated. Therefore, bootstrapping techniques were applied when conducting CFA.

Table 5.38

Assessment of normality for the indicators within the construct ‘social capital’


v26 1.00 7.00 -0.05 -0.33 -1.10 -3.40

v25 1.00 7.00 0.23 1.40 -0.90 -2.80

v24 1.00 7.00 0.12 0.75 -1.10 -3.40

v23 1.00 7.00 -0.07 -0.45 -0.93 -2.88


A single-factor measurement model for the construct ‘social capital’ was specified as a


Figure 5.16.

202

Figure 5.16. Measurement model for the construct ‘social capital’

The model presented a non-significant chi-square value of .32 (p = .851). The values of

all the baseline comparisons fit indices (NFI, RFI, IFI, TLI, and CFI) were above the cutoff

value of .90 (Bentler, 1990) and the value of RMSEA (< .001) was well below the cutoff

value of .08 (R. Ho, 2006). These indicated a good model fit (see Table 5.39). The

unstandardised regression weight of each indicator corresponded to a significant critical ratio.

All the values of the standardised regression weight (see Figure 5.16) were above the cutoff

value of .50 (Hair et al., 2006) and equal to or above the preferred value of .70 (Hair et al.,

2006). This indicated that all four corresponding items were significant reflective indicators

of the respective construct ‘social capital’. The model converged.

203

Table 5.39

Model fit indices for the construct ‘social capital’

CMIN


Default model 8 .32 2 .851 .16 Saturated model 10 .000 0


Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 1.00 1.00 1.00 1.01 1.00



RMSEA




Table 5.40

Proportional factor score weights for the indicators within the construct ‘social capital’

v26 v25 v24 v23 Sum


Proportional factor score weight .16 .26 .20 .37 .99

The scale score for the construct ‘social capital’ then can be calculated by the following

equation:

v26 × .16 + v25 × .26 + v24 × .20 + v23 × .37

Reliability and validity of the construct ‘social capital’ were further checked by four



These measures are reported below.


cutoff value of .30 (Jöreskog & Sörbom, 1996) and equal to or above the preferred value

204

of .50 (Jöreskog & Sörbom, 1996). Therefore, the construct ‘social capital’ accounted for a

reasonable level of the variance of its indicators. The reliability of all the indicators of the

construct ‘social capital’ represented by the SMC measures was good. The construct

reliability for ‘social capital’ was .61, which was greater than the cutoff value of .50 (Fornell

& Larcker, 1981). The variance of the indicators captured by the construct ‘social capital’ was

greater than that due to the measurement errors. This indicated that the construct reliability

was acceptable. The construct ‘social capital’ accounted for 69.99% of the overall variation in

the four indicators. This was well above the cutoff of 50% (Fornell & Larcker, 1981).

Therefore, the reliability of the indicators of the construct ‘social capital’ represented by the

variance extracted was good. The coefficient H value for the construct ‘social capital’ was .87,

which was well above the cutoff value of .70 (G. R. Hancock & Mueller, 2001). This

represented a high reliability.

The measurement model for the construct ‘social capital’ had a good model fit, which



of the construct ‘social capital’ corresponded to a significant critical ratio. Therefore, all the

indicators demonstrated a good convergent validity.

Symbolic capital

Originally, five indicators were used to operationalise the construct ‘symbolic capital’. These

indicators were reflective of the degree to which financial wealth, Chinese cultural

dispositions, and Chinese social networks were recognised or valued. This section discusses

whether the five indicators fit into the measurement model for the construct ‘symbolic capital’

and reports on the modification of the measurement model to improve the model fit.

205

The correlation matrix demonstrated that all indicators had a statistically significant


the cutoff range between .30 and .90 (Field, 2009). This suggested the basis for EFA.

Table 5.41

Correlation matrix for the indicators within the construct ‘symbolic capital’

Correlation v30 v31 v32 v33 v34

v30 1.00 .53 .59 .55 .53

v31 .53 1.00 .72 .66 .63

v32 .59 .72 1.00 .69 .62

v33 .55 .66 .69 1.00 .75

v34 .53 .63 .62 .75 1.00

Sig. v30 v31 v32 v33 v34

v30 .000 .000 .000 .000

v31 .000 .000 .000 .000

v32 .000 .000 .000 .000

v33 .000 .000 .000 .000

v34 .000 .000 .000 .000


the cutoff value of .50 (Kaiser, 1974), which indicated the EFA should yield distinct and

reliable factors. Bartlett’s test of sphericity confirmed the adequacy of magnitude of the


Table 5.42

KMO and Bartlett's test for the indicators within the construct ‘symbolic capital’




df 10

Sig. .000

An examination of communalities indicated that a reasonable level of variation in the


construct ‘symbolic capital’. The extraction in all cases explained more variance in an

206

indicator than the initial model (see Table 5.43). This indicated the appropriateness of the

current factor extraction.

Table 5.43

Communalities for the indicators within the construct ‘symbolic capital’

Initial Extraction

v30 .41 .45

v31 .59 .65

v32 .63 .69

v33 .66 .73

v34 .60 .64


There was only one factor with an eigenvalue above 1. Consequently, only one factor

was extracted. This factor accounted for 70.35% of the total variance of the five indicators

(see Table 5.44). The scree plot also supported a one-factor solution (see Figure 5.17).

Table 5.44

Total variance of the construct ‘symbolic capital’ explained by its indicators



1 3.52 70.35 70.35 3.16 63.25 63.25

2 .53 10.58 80.93

3 .44 8.73 89.65

4 .29 5.71 95.36

5 .23 4.64 100.00


Figure 5.17. Scree plot of the ‘symbolic capital’ scale

207



value of |2| (Field, 2009). Mardia’s coefficient value of 11.30 was much greater than the

cutoff value of 3.00 (Yuan et al., 2002). This indicated significant deviation from multivariate

normality. Therefore, bootstrapping techniques were applied when conducting CFA.

Table 5.45

Assessment of normality for the indicators within the construct ‘symbolic capital’


v34 1.00 7.00 -0.06 -0.40 -0.77 -2.37

v33 1.00 7.00 -0.12 -0.73 -0.91 -2.81

v32 1.00 7.00 0.19 1.15 -1.00 -3.11

v31 1.00 7.00 0.05 0.33 -1.08 -3.36

v30 1.00 7.00 0.31 1.89 -0.95 -2.93


A single-factor measurement model for the construct ‘symbolic capital’ was specified

as a latent variable with five indicators. The model with standardised parameters is illustrated

in Figure 5.18.

Figure 5.18. Measurement model for the construct ‘symbolic capital’

208

Model fit indices are demonstrated in Table 5.46. Although all the baseline

comparisons indices were above the cutoff value of .90 (Bentler, 1990), the significant

chi-square value of 23.98 and the RMSEA value of .13, which was above the cutoff value

of .08 (R. Ho, 2006), implied that the model fit could be improved.

Table 5.46

Model fit indices for the construct ‘symbolic capital’

CMIN




Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 0.96 0.93 0.97 0.94 0.97



RMSEA


Default model 0.13 0.08 0.18 0.01



modification index (see Table 5.48) indicated that v33 and v34 were of concern. These two

indicators were associated with a high value of standardised residual covariances, a high

value of covariances between indicator error terms, and a high value of regression weights

between indicators (as highlighted in Table 5.47 and 5.48). The online questionnaire items

corresponding to these two indicators were revisited. The item v33 reflected the symbolic

capital produced by the participants’ Chinese heritage while v34 reflected the symbolic

capital produced by the participants’ popularity among their peers. This popularity may or

may not be attributed to participants’ Chinese heritage or CHL. Therefore, v33 was more

meaningful in the current context. The item v34 was tentatively excluded from the model.

209

Table 5.47

Standardised residual covariances for the indicators within the construct ‘symbolic capital’

v34 v33 v32 v31 v30

v34 0.00

v33 0.73 0.00

v32 -0.63 -0.20 0.00

v31 -0.26 -0.34 0.66 0.00

v30 -0.09 -0.28 0.48 -0.07 0.00

Table 5.48

Modification indices for the construct ‘symbolic capital’


M.I. Par change

e33 <--> e34 12.69 .27

e32 <--> e34 7.54 -.23 e31 <--> e32 8.49 .26

Before testing the fit of the modified model, multivariate normality was checked (see

Table 5.49). Mardia’s coefficient was 6.43, which was greater than the cutoff value of 3.00

(Yuan et al., 2002). This implied the violation of the multivariate normality. Therefore,

bootstrapping was applied when conducting CFA.

Table 5.49

Assessment of normality for the indicators within the construct ‘symbolic capital’ (modified)


v33 1.00 7.00 -0.12 -0.73 -0.91 -2.81

v32 1.00 7.00 0.19 1.15 -1.00 -3.11

v31 1.00 7.00 0.05 0.33 -1.08 -3.36

v30 1.00 7.00 0.31 1.89 -0.95 -2.93


A modified measurement model for the construct ‘symbolic capital’ was specified as a


Figure 5.19.

M.I. Par change

v33 <--- v34 4.02 .08

210

Figure 5.19. Modified measurement model for the construct ‘symbolic capital’

The model presented a non-significant chi-square value of .82 (p = .665). The values of

all the baseline comparisons fit indices (NFI, RFI, IFI, TLI, and CFI) were above the

cutoff .90 (Bentler, 1990) and the value of RMSEA (< .001) was well below the cutoff .08 (R.

Ho, 2006). These were indicative of a good model fit (see Table 5.50). The unstandardised


of the standardised regression weight (see Figure 5.19) were above the cutoff value of .50

(Hair et al., 2006) and most values of the standardised regression weight were above the

preferred value of .70 (Hair et al., 2006). The only exception was associated with the

standardised regression weight of v30. However, the value of .67 was only marginally

below .70. Therefore, all the four corresponding items were significant reflective indicators of

the respective construct ‘symbolic capital’. The model converged.

211

Table 5.50

Model fit indices for the indicators within the construct ‘symbolic capital’

CMIN


Default model 8 .82 2 .665 .41 Saturated model 10 .00 0


Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 1.00 1.00 1.00 1.01 1.00



RMSEA




modified model decreased the chi-square value by 23.16 (23.98-.82) from the original model.

The chi-square difference was a statistically significant one (p < .001) with a degree of

freedom difference of 3 (5-2).


Table 5.51

Proportional factor score weights for the indicators within the construct ‘symbolic capital’

v33 v32 v31 v30 Sum



The scale score for the construct ‘symbolic capital’ then can be calculated by the

following equation:

v33 × .24 + v32 × .38 + v31 × .26 + v30 × .13

Reliability and validity of the construct ‘symbolic capital’ were further checked by four



212


cutoff value of .30 (Jöreskog & Sörbom, 1996) and most of them were above the preferred

value of .50 (Jöreskog & Sörbom, 1996). The only exception, the value of .45, was

considered not too far away from the cutoff .50. Therefore, the construct ‘symbolic capital’

accounted for a reasonable level of the variance of its indicators. The reliability of all the

indicators of the construct ‘symbolic capital’ represented by the SMC measures was good.

The construct reliability for ‘symbolic capital’ was .63, which was greater than the cutoff

value of .50 (Fornell & Larcker, 1981). The variance of the indicators captured by the

construct ‘symbolic capital’ was greater than that due to the measurement errors. This

indicated that the construct reliability was acceptable. The construct ‘symbolic capital’

accounted for 72.09% of the overall variation in the four indicators. This was well above the

cutoff of 50% (Fornell & Larcker, 1981). Therefore, the reliability of the indicators of the

construct ‘symbolic capital’ represented by the variance extracted was good. The coefficient

H value for the construct ‘symbolic capital’ was .89, which was well above the cutoff value


The modified measurement model for the construct ‘symbolic capital’ had a good

model fit, which indicated the unidimensionality of the four indicators. This supported the

claim for its construct validity. As stated earlier, the unstandardised regression weight of all

the indicators of the construct ‘symbolic capital’ corresponded to a significant critical ratio.


CHL proficiency

Originally, eight indicators were used to measure ‘CHL proficiency’. These indicators were

reflective of four CHL skills, namely listening, speaking, reading, and writing. Each skill was

213

measured by two indicators. This section discusses whether the eight indicators fit into the

measurement model of ‘CHL proficiency’ and reports on the modification of the

measurement model to improve the model fit.

The correlation matrix demonstrated that all the indicators had a statistically significant

medium to strong correlation with other indicators (p < .001 as shown in Table 5.52), varying

within the cutoff range between .30 and .90 (Field, 2009). This suggested the basis for EFA.

Table 5.52

Correlation matrix for the indicators within the construct ‘CHL proficiency’

Correlation v35 v36 v37 v38 39 v40 v41 v42

v35 1.00 .89 .79 .79 .76 .69 .67 .67

v36 .89 1.00 .77 .86 .77 .79 .68 .73

v37 .79 .77 1.00 .88 .82 .72 .75 .69

v38 .79 .86 .88 1.00 .82 .81 .73 .76

v39 .76 .77 .82 .82 1.00 .84 .82 .82

v40 .69 .79 .72 .81 .84 1.00 .74 .85

v41 .67 .68 .75 .73 .82 .74 1.00 .83

v42 .67 .73 .69 .76 .82 .85 .83 1.00

Sig. v35 v36 v37 v38 39 v40 v41 v42

v35 .000 .000 .000 .000 .000 .000 .000

v36 .000 .000 .000 .000 .000 .000 .000

v37 .000 .000 .000 .000 .000 .000 .000

v38 .000 .000 .000 .000 .000 .000 .000

v39 .000 .000 .000 .000 .000 .000 .000

v40 .000 .000 .000 .000 .000 .000 .000

v41 .000 .000 .000 .000 .000 .000 .000

v42 .000 .000 .000 .000 .000 .000 .000


the cutoff value of .50 (Kaiser, 1974), which indicated that the EFA should yield distinct and



214

Table 5.53

KMO and Bartlett's test for the indicators within the construct ‘CHL proficiency’




df 28

Sig. .000

An examination of communalities indicated that a good level of variation in the


construct ‘CHL proficiency’. Of concern was that the extraction of all cases explained less

variance in an indicator than the initial model (see Table 5.54). This was an indication that

there might be an issue in the model fit. A closer examination of this issue will be discussed

subsequently.

Table 5.54

Communalities for the indicators within the construct ‘CHL proficiency’

Initial Extraction

v35 .84 .73

v36 .88 .79

v37 .83 .78

v38 .87 .85

v39 .85 .84

v40 .83 .78

v41 .77 .70

v42 .82 .74

Extraction method: Maximum likelihood.


total variance of the eight indicators (see Table 5.55). Thus, only one factor was extracted.

An examination of the scree plot also supported a one-factor solution (see Figure 5.20).

215

Table 5.55

Total variance of the construct ‘CHL proficiency’ explained by its indicators



1 6.44 80.52 80.52 6.22 77.78 77.78

2 .55 6.87 87.38

3 .32 4.01 91.39

4 .26 3.27 94.66

5 .16 2.02 96.68

6 .11 1.35 98.03

7 .09 1.14 99.17

8 .07 .83 100.00


Figure 5.20. Scree plot of the ‘CHL proficiency’ scale


univariate normal distribution by displaying the critical ratio of skewness, kurtosis, or both

beyond the cutoff value of |2| (Field, 2009). The value of Mardia’s coefficient was 33.52,

which was much higher than the cutoff value of 3.00 (Yuan et al., 2002). This indicated

significant deviation from multivariate normality. Therefore, bootstrapping techniques were

applied when conducting CFA.

216

Table 5.56

Assessment of normality for the indicators within the construct ‘CHL proficiency’


v42 1.00 7.00 0.65 4.01 -0.76 -2.35

v41 1.00 7.00 0.26 1.62 -1.10 -3.40

v40 1.00 7.00 0.58 3.60 -0.94 -2.92

v39 1.00 7.00 -0.03 -0.20 -1.34 -4.16

v38 1.00 7.00 0.12 0.75 -1.30 -4.04

v37 1.00 7.00 -0.14 -0.89 -1.26 -3.91

v36 1.00 7.00 -0.25 -1.52 -1.27 -3.93

v35 1.00 7.00 -0.57 -3.55 -0.91 -2.81


A single-factor measurement model for the construct ‘CHL proficiency’ was specified

as a latent variable with eight indicators. The model with standardised parameters is

illustrated in Figure 5.21.

Figure 5.21. Measurement model for the construct ‘CHL proficiency’

The chi-square value of 357.78 corresponded to a significant p (<.001) value. All the

baseline comparisons fit indices were below the cutoff value of .90 (Bentler, 1990). The value

of RMSEA (.27) was well above the cutoff value of .08 (R. Ho, 2006). These were all

indicative of a poor model fit (see Table 5.57). This was not surprising because EFA has

already hinted that there was a model fit problem: The extraction of all cases explained less

variance in an indicator than the initial model.

217

Table 5.57

Model fit indices for the construct ‘CHL proficiency’

CMIN




Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 0.85 0.79 0.86 0.80 0.86



RMSEA


Default model 0.27 0.25 0.30 .00 Independence model 0.61 0.59 0.63 .00


modification index (see Table 5.59) identified some problems. The two highest scores of

standardised residual covariances, covariances between error terms, and regressions weights

were all associated with v35 and v36, and v41 and v42. Moreover, v39 and v40 were of

concern because these two indicators were also associated with high standardised residual

covariances, error terms covariances, and/or regression weights. The questionnaire items

corresponding to these seemingly problematic indicators were revisited. Indicators v35 and

v36 were used to measure CHL listening proficiency; v41 and v42 were used to measure

CHL writing proficiency; and v39 and v40 were used to measure CHL speaking and reading

proficiencies respectively. Each skill was measured by two indicators. A solution to improve

the model fit was then made by deleting one indicator from the measures for each skill. As a

result, CHL proficiency was still measured by participants’ self-reporting proficiency of

listening, speaking, reading, and writing, but each CHL skill was measured by only one

indicator. Indicators v36, v39, v40, and v42 were tentatively removed from the model.

218

Table 5.58

Standardised residual covariances for the indicators within the construct ‘CHL proficiency’

v42 v41 v40 v39 v38 v37 v36 v35

v42 0.00

v41 1.37 0.00

v40 1.05 -0.01 0.00

v39 0.37 0.65 0.38 0.00

v38 -0.38 -0.58 -0.05 -0.27 0.00

v37 -0.87 0.09 -0.75 0.12 0.70 0.00

v36 -0.41 -0.81 0.00 -0.58 0.40 -0.16 0.00

v35 -0.83 -0.49 -0.73 -0.27 0.01 0.39 1.65 0.00

Table 5.59

Modification indices for the construct ‘CHL proficiency’


M.I. Par change M.I. Par change e41<-->e42 43.57 0.49 v42<---v41 12.13 0.12

e40<-->e42 38.27 0.41 v42<---v40 7.37 0.09

e39<-->e42 7.02 0.16 v42<---v37 5.07 -0.08

e39<-->e41 17.97 0.28 v42<---v35 4.56 -0.08

e39<-->e40 9.15 0.18 v41<---v42 10.31 0.12

e38<-->e42 8.11 -0.17 v40<---v42 9.10 0.10

e38<-->e41 15.69 -0.25 v40<---v37 4.82 -0.07

e38<-->e39 7.65 -0.14 v40<---v35 4.44 -0.08

e37<-->e42 26.02 -0.34 v39<---v41 5.05 0.07

e37<-->e40 24.67 -0.32 v39<---v36 4.24 -0.06

e37<-->e38 34.58 0.33 v38<---v41 4.41 -0.06

e36<-->e42 6.01 -0.17 v38<---v37 6.84 0.08

e36<-->e41 20.02 -0.32 v37<---v42 6.19 -0.09

e36<-->e39 22.60 -0.28 v37<---v40 4.77 -0.07

e36<-->e38 12.06 0.20 v37<---v38 4.26 0.07

e35<-->e42 17.99 -0.28 v36<---v41 5.59 -0.08

e35<-->e41 5.23 -0.16 v36<---v35 23.40 0.18

e35<-->e40 17.46 -0.26 v35<---v42 4.26 -0.07

e35<-->e37 4.80 0.14 v35<---v36 16.99 0.13

e35<-->e36 91.97 0.61

Before testing the fit of the modified model, multivariate normality was checked again

because four indicators had been removed from the model. Mardia’s coefficient was 4.36,

which was greater than the cutoff value of 3.00 (Yuan et al., 2002). This implied the violation

of the multivariate normality (see Table 5.60). Therefore, bootstrapping was applied when

conducting CFA.

219

Table 5.60

Assessment of normality for the indicators within the construct ‘CHL proficiency’ (modified)


v41 1.00 7.00 0.26 1.62 -1.10 -3.40

v38 1.00 7.00 0.12 0.75 -1.30 -4.04

v37 1.00 7.00 -0.14 -0.89 -1.26 -3.91

v35 1.00 7.00 -0.57 -3.55 -0.91 -2.81


A modified measurement model for the construct ‘CHL proficiency’ was specified as a


Figure 5.22.

Figure 5.22. Modified measurement model for the construct ‘CHL proficiency’

Model fit was improved. The chi-square value of 1.17 corresponded to a

non-significant p value of .557. All the baseline comparisons indices (NFI, RFI, IFI, TLI, and

CFI) were above the cutoff value of .90 (Bentler, 1990). The RMSEA value (< .001) was well

below the cutoff value of .08 (R. Ho, 2006). All of these were indicative of a good model fit

(see Table 5.61). The unstandardised regression weight of each indicator corresponded to a

significant critical ratio. All the values of the standardised regression weight (see Figure 5.22)

were above the cutoff value of .50 (Hair et al., 2006) and the preferred value of .70 (Hair et

220

al., 2006). Therefore, all the four corresponding items were significant reflective indicators of

the respective construct ‘CHL proficiency’. The model converged.

Table 5.61

Model fit indices for the construct ‘CHL proficiency’ (modified)

CMIN


Default model 8 1.17 2 .557 .59 Saturated model 10 .00 0


Model NFI

Delta1 RFI

rho1 IFI

Delta2 TLI

rho2 CFI

Default model 1.00 1.00 1.00 1.00 1.00



RMSEA


Default model 0.00 0.00 0.11 0.71



modified model decreased the chi-square value of 356.61 (357.78-1.17) from the original

model. This value was statistically significant (p < .001) with a difference in degrees of

freedom of 18 (20-2).


Table 5.62

Proportional factor score weights for the indicators within the construct ‘CHL proficiency’

v41 v38 v37 v35 Sum

Factor score weight .11 .34 .41 .16 1.02


The scale score for the construct ‘CHL proficiency’ then can be calculated by the

following equation:

v41 × .11 + v38 × .33 + v37 × .40 + v35 × .16

221

Reliability and validity of the construct ‘CHL proficiency’ were further checked by

four measures of reliability, namely SMC, construct reliability, variance extracted, and


validity.



Sörbom, 1996). Therefore, the construct ‘CHL proficiency’ accounted for a reasonable level

of the variance of its indicators. The reliability of all the indicators of the construct ‘CHL

proficiency’ represented by the SMC measures was good. The construct reliability for ‘CHL

proficiency’ was .77, which was greater than the cutoff value of .50 (Fornell & Larcker, 1981).

The variance of the indicators captured by the construct ‘CHL proficiency’ was greater than

that due to the measurement errors. This indicated that the construct reliability was good. The

construct ‘CHL proficiency’ accounted for 82.56% of the overall variation in the four

indicators. This was well above the cutoff of 50% (Fornell & Larcker, 1981). Therefore, the

reliability of the indicators of the construct ‘CHL proficiency’ represented by the variance

extracted was good. The coefficient H value for the construct ‘CHL proficiency’ was .95,

which was well above the cutoff value of .70 (G. R. Hancock & Mueller, 2001). This

represented a high reliability.

The modified measurement model for the construct ‘CHL proficiency’ had a good

model fit, which indicated the unidimensionality of the four indicators. This supported the

claim for its construct validity. As stated earlier, the unstandardised regression weight of all

the indicators of the construct ‘CHL proficiency’ corresponded to a significant critical ratio.


222

Section summary

In this section, the six theoretical constructs, namely ‘Chineseness’, ‘economic capital’,

‘cultural capital’, ‘social capital’, ‘symbolic capital’, and ‘CHL proficiency’, were specified

as latent variables with their corresponding indicators. The initial EFA and the subsequent

CFA improved the model fit to a reasonably good level. These constructs and their respective

indicators included in the (modified) measurement models are summarised in Table 5.63.

Table 5.63

Item sets and their corresponding constructs in the (modified) measurement models

Construct Indicators Item description

Chineseness

(modified)

v2 Handwriting tells you a lot about a person’s character. To what extent do you


v3 Children must be taught to paint, dance, or play musical instruments. To what


v4 Academic education is the most important thing in school. To what extent do

you agree with this statement?


title plus surname rather than their first name. To what extent do you agree


v9 Confucianism values the virtue of frugality. Confucius said that frugality is a

common character of people with virtue. He also said that frugality is very

important in family life. In general, how much degree would you agree with

these values?

Economic

capital

v11 Money is not an issue if I wish to travel to China. To what extent do you


v12 I can afford to study in Chinese language schools as many hours as I wish. To

what extent do you agree with this statement?

v13 I can afford to employ private tutors to teach me Chinese if I wish. To what


v14 Money is never an issue if I want to buy Chinese language learning materials,

such as textbooks, dictionaries, and tapes. To what extent do you agree with

this statement?

Cultural

capital

(modified)


listening to the radio, reading newspapers, or surfing online. To what extent



concerts, if they feature Chinese culture. To what extent do you agree with

this statement?

v19 I read a lot of books about China. To what extent do you agree with this

statement?


learning musical instruments, calligraphy, or painting. To what extent do you


Social

capital

v23 Most of my friends are of Chinese descent. To what extent do you agree with

this statement?

223


include members of Chinese descent. To what extent do you agree with this

statement?

v25 I tend to mix exclusively with Chinese social groups. To what extent do you


v26 I think it is necessary to learn Chinese language in order to socialise with my

Chinese peers. To what extent do you agree with this statement?

Symbolic

capital

(modified)

v30 The wealth of my family is well-known in Chinese communities. To what


v31 My Chinese language competency is well regarded. To what extent do you


v32 Learning Chinese has increased my status in Chinese communities. To what


v33 People value my Chinese heritage. To what extent do you agree with this

statement?

CHL

proficiency

(modified)

v35 I can easily understand my family members and friends when they talk to me

in Chinese. To what extent do you agree with this statement?

v37 I can easily handle complex situations in Chinese, such as banking, arguing,

purchasing a house or a car. To what extent do you agree with this statement?

v38 I can read Chinese language textbooks easily. To what extent do you agree


v41 I can always write Chinese characters and Chinese words correctly. To what


5.2.2 Hypothesis testing: SEM

In the previous section (5.2.1), modified measurement models for the theoretical constructs

and their respective indicators were specified by CFA. As such, CFA provides the foundation

for hypothesis testing. In this section, the primary investigation will shift to the examination

of the relationships between the theoretical constructs. Drawing insights from the theoretical

framework, the structural models can specify the nature and magnitude of the relationships

between constructs. Therefore, the structural models function to test the hypotheses, and

further, test whether the structural models accord with the theoretical framework. The testing

of Hypotheses One and Two follows.

224

H01. The habitus of ‘Chineseness’ of young Chinese Australian adults has no


The relationship between ‘Chineseness’ and ‘CHL proficiency’ was specified in the structural

model, with the standardised parameters demonstrated in Figure 5.23.

Figure 5.23. Structural model for the relationship between ‘Chineseness’ and ‘CHL proficiency’

The model had a significant chi-square value of 57.92. However, the ratio between the

chi-square value and the degrees of freedom (57.92/26 = 2.23) was not very high; all the

baseline comparison fit indices were above the cutoff value of .90 (Bentler, 1990); and the

RMSEA value of .07 was below the cutoff value of .08 (R. Ho, 2006). It can be argued that

the model had a reasonably good fit given its complexity (see Table 5.64). The

unstandardised regression weight of each indicator corresponded to a significant critical ratio.

All the values of the standardised regression weight were above the cutoff value of .50 (Hair

et al., 2006) and equal to or above the preferred value of .70 (Hair et al., 2006). The model

converged.

225

Table 5.64

Model fit indices for the structural model: ‘Chineseness’ and ‘CHL proficiency’

CMIN


Default model 19 57.92 26 .000 2.23




Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI

Default model .96 .95 .98 .97 .98


Independence model .00 .00 .00 .00 .00

RMSEA


Default model .07 .05 .10 .06

Independence model .41 .40 .43 .00

The results indicated that the construct ‘Chineseness’ was significantly and positively

related to the construct ‘CHL proficiency’ (r = .71, see Figure 5.23). The null hypothesis H01

can be rejected. ‘Chineseness’ of young Chinese Australian adults had a statistically

significant strong positive impact on their CHL proficiency. When young Chinese Australian

adults had a stronger sense of ‘Chineseness’, they tended to have a higher level of CHL

proficiency. About 50% of the variance of ‘CHL proficiency’ was accounted for by the

variance of ‘Chineseness’ (see Figure 5.23). The remaining 50% could not be explained by

the model and was thus attributed to other factors.

In line with Bourdieu’s capital metaphor, Garnham and Williams (1993) argue that the

investment of all forms of capital leads to a given agent’s choice of cultivating a language. As

such, young Chinese Australian adults’ investment of various forms of capital may explain

part of the variance in their CHL proficiency. Accordingly, a null hypothesis follows.

226

H02. The investment of capital of young Chinese Australian adults has no statistically

significant impact on their CHL proficiency.

According to Bourdieu (1986, 1991), capital has four forms, namely economic capital,

cultural capital, social capital, and symbolic capital. The theoretical construct ‘capital’ was

then treated as a latent variable with four indicators, with each indicator representing one

form of capital. Each capital form was treated as an ‘item parcel’. This strategy of ‘item

parcelling’ combines indicators into sets of several composite variables by either summing or

averaging the indicator score of the combined indicators (Hair et al., 2006). As illustrated in

the earlier sections, the scale score for the four composite variables, namely ‘economic

capital’, ‘cultural capital’, ‘social capital’, and ‘symbolic capital’, can be computed as

continuous variables by multiplying the individual’s raw score on each indicator by the

proportionally weighted factor score of each indicator and summing. The formulae for

computation are summarised in Table 5.65.

Table 5.65

Formulae for computation of the four composite variables

Economic capital v14 × .21 + v13 × .31 + v12 × .36 + v11 × .12

Cultural capital v20 × .18 + v19 × .29 + v18 × .35 + v17 × .19

Social capital v26 × .16 + v25 × .26 + v24 × .20 + v23 × .37

Symbolic capital v33 × .24 + v32 × .38 + v31 × .26 + v30 × .13

There are several methodological considerations regarding the rationale for item

parcelling. First of all, indicator number was of concern. The theoretical construct ‘capital’

was reflected by 16 indicators with each capital form reflected by four reliable and valid

indicators. Applications involving more than 15 reflective indicators can call for item

parcelling (Hair et al., 2006). Secondly, the 16 indicators displayed high internal consistency

reliability with a Cronbach’s α value of .95. Cronbach’s α above the cutoff value of .90

suggested the basis for item parcelling (Hair et al., 2006). Thirdly, the parcels, or the

227

composite variables contained groups of indicators with the most conceptual similarity. This

is true because each parcel represented a form of capital with its four reflective indicators.

The previous reliability test, EFA, and CFA have confirmed that the reflective indicators

within their corresponding parcels shared a conceptual basis. Moreover, the four parcels

suggested unidimensionality because EFA run on these four parcels demonstrated that they all

loaded on one factor. This indicated that the four capital forms were represented by one

construct, that is, ‘capital’. The EFA results are discussed as follows.

The correlation matrix demonstrated that all capital forms had a statistically significant

medium correlation with other capital forms (p < .001 as shown in Table 5.66), varying

within the cutoff range between .30 and .90 (Field, 2009). This suggested the basis for

conducting EFA.

Table 5.66

Correlation matrix for the indicators within the construct ‘capital’

Correlation Economic Cultural Social Symbolic

Economic 1.00 .64 .57 .65

Cultural .64 1.00 .72 .78

Social .57 .72 1.00 .74

Symbolic .65 .78 .74 1.00

Sig. Economic Cultural Social Symbolic

Economic .000 .000 .000

Cultural .000 .000 .000

Social .000 .000 .000

Symbolic .000 .000 .000


the cutoff value of .50 (Kaiser, 1974), which indicated that the EFA should yield distinct and



228

Table 5.67

KMO and Bartlett's test for the indicators within the construct ‘capital’




df 6

Sig. .000

As highlighted in Table 5.68, a good level of variation in the four forms of capital

(between 51% and 81%) was explained by the construct ‘capital’. The extraction of all cases

explained greater variance in a capital form than the initial model.

Table 5.68

Communalities for the indicators within the construct ‘capital’

Initial Extraction

Economic .47 .51

Cultural .67 .76

Social .61 .68

Symbolic .70 .81

Extraction method: Maximum likelihood.


total variance of the four forms of capital (see Table 5.69). As a result, only one factor was


Table 5.69

Total variance of the construct ‘capital’ explained by its indicators



1 3.05 76.32 76.32 2.76 68.96 68.96

2 .45 11.35 87.67

3 .28 6.96 94.63

4 .21 5.37 100.00


229

Figure 5.24. Scree plot of the ‘capital’ scale

A measurement model for the construct ‘capital’ was evaluated by CFA. As shown in

Table 5.70, the data violated the multivariate normal distribution assumption because

Mardia’s coefficient was 4.75, which was greater than the cutoff value of 3.00 (Yuan et al.,

2002). Bootstrapping techniques were applied when conducting CFA.

Table 5.70

Assessment of normality for the indicators within the construct ‘capital’


Symbolic capital 1.00 7.00 0.14 0.86 -1.01 -3.11

Social capital 1.00 7.00 -0.01 -0.05 -0.98 -3.05

Cultural capital 1.00 7.00 0.22 1.35 -0.97 -2.99

Economic capital 1.00 7.00 0.01 0.08 -1.01 -3.11


Item parcelling has the potential to improve model fit simply because it reduces the

complexity of the model, and models with fewer indicators have the potential for better fit

(Landis, Beal, & Tesluk, 2000). This is particularly true for the item parcelling in the current

study. As shown in Table 5.71, the model achieved a good fit. Chi-square value (.97)

corresponded to a non-significance level (p = .616). All the baseline comparisons model fit

indices were above the cutoff value of .90 (Bentler, 1990). RMSEA value (< .001) was well

below the cutoff value of .08 (R. Ho, 2006).

230

Table 5.71

Model fit indices for the construct ‘capital’

CMIN


Default model 8 .97 2 .616 .49




Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI

Default model 1.00 1.00 1.00 1.01 1.00



RMSEA




The standardised parameters are demonstrated in Figure 5.25. The unstandardised


of the standardised regression weight were above the cutoff value of .50 (Hair et al., 2006)

and the preferred value of .70 (Hair et al., 2006). The model converged.

Figure 5.25. Measurement model for the construct ‘capital’

231

Table 5.72 demonstrates the proportional factor score weight of each form of capital.

Table 5.72

Proportional factor score weights for the indicators within the construct ‘capital’

Economi

c capital

Cultural

capital

Social

capital

Symbolic

capital Sum



The scale score for the construct ‘capital’ can then be computed by the following

equation:

economic capital × .11 + cultural capital × .29 + social capital × .22 + symbolic capital × .38

Four measures of reliability, namely SMC, construct reliability, variance extracted, and


validity, are reported below.



Sörbom, 1996). Therefore, the construct ‘capital’ accounted for a reasonable level of the

variance of its indicators, namely four forms of capital. The reliability of all the indicators of

the construct ‘capital’ represented by the SMC measures was good. The construct reliability

for ‘capital’ was .69, which was greater than the cutoff value of .50 (Fornell & Larcker, 1981).

The variance of four forms of capital captured by the construct ‘capital’ was greater than that

due to the measurement errors. This indicated that the construct reliability was acceptable.

The construct ‘capital’ accounted for 76.32% of the overall variation in four forms of capital.

This was well above the cutoff of 50% (Fornell & Larcker, 1981). Therefore, the reliability of

the indicators of the construct ‘capital’ represented by the variance extracted was good. The

232

coefficient H value for the construct ‘capital’ was .91, which was well above the cutoff value


The measurement model for the construct ‘capital’ had a good model fit, which



of the construct ‘capital’ corresponded to a significant critical ratio. Therefore, all the

indicators demonstrated a good convergent validity.

The relationship between ‘capital’ and ‘CHL proficiency’ was then specified in the

structural model, with the standardised parameters demonstrated in Figure 5.26.

Figure 5.26. Structural model for the relationship between ‘capital’ and ‘CHL proficiency’

Although the model had a significant chi-square value of 66.30, the ratio of chi-square

value and degree of freedom (3.50) was not very high. Although the RMSEA value of .10

was above the cutoff value of .08 (R. Ho, 2006), all the baseline comparison fit indices were

above the cutoff value of .90 (Bentler, 1990). Particularly, CFI and TLI, the two most widely

referenced indices (Hair et al., 2006), had relatively high values of .97 and .96 respectively.

233

Therefore, it can be argued that the model had a reasonably good fit given its complexity (see

Table 5.73). The unstandardised regression weight of each indicator corresponded to a

significant critical ratio. All the values of the standardised regression weight (see Figure 5.26)

were above the cutoff value of .50 (Hair et al., 2006) and most of the standardised regression

weights were above the preferred value of .70 (Hair et al., 2006). The only exception was the

standardised regression weight associated with ‘economic capital’. However, the value of .69

was only marginally lower than the preferred value of .70. Therefore, the model was

considered to be converged.

Table 5.73

Model fit indices for the structural model: ‘Capital’ and ‘CHL proficiency’

CMIN


Default model 17 66.30 19 .000 3.50




Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI

Default model .96 .94 .97 .96 .97



RMSEA




The results indicated that the construct ‘capital’ was significantly and positively related

to the construct ‘CHL proficiency’ (r = .86, see Figure 5.26). The null hypothesis H02 can be

rejected. The investment of capital of young Chinese Australian adults had a statistically

significant strong positive impact on their CHL proficiency. When young Chinese Australian

adults put more investment of different forms of capital into CHL learning, they tended to

achieve a higher level of CHL proficiency. ‘Capital’ can therefore explain part of the variance

of ‘CHL proficiency’ that could not be explained by ‘Chineseness’.

234

What is also of interest is that the variance of ‘capital’ was reflected in the variance of

different forms of capital at different levels. Compared to other capital forms, symbolic

capital had the most variance caused by the variance of capital (.93, see Figure 5.26). This is

because symbolic capital is perhaps the most valuable form of accumulation in a society

(Bourdieu, 1977b). It will be recalled that economic, cultural, and social capital all had

stronger correlations with symbolic capital than their correlations with other forms of capital

(as shown in Table 5.66). This is because participants’ investment of economic, cultural, and

social capital, when perceived and recognised as legitimate (Bourdieu, 1985b, 1989b, 1991,

2000), can all exist and act through the form of symbolic capital.

According to Bourdieu (1991), people make choices about different aspects of a

particular language according to the amount of different resources (capital) as well as the

dispositions (habitus) that they have in a given field. The theoretical framework developed in

Chapter Three extended this Bourdieusian stance to people’s choices about different

languages according to their habitus in a field and their capital valued by that field. The data

confirmed this theoretical framework and justified the understandings that both participants’

investment of ‘capital’ and habitus of ‘Chineseness’ contributed to their ‘CHL proficiency’.

5.2.3 Habitus and capital working together: Regression

In the previous section (5.2.2), it was concluded that participants’ ‘Chineseness’ (habitus) and

their various resources (capital) made statistically significant positive contributions to their

CHL proficiency, when the relationship between ‘Chineseness’ and ‘CHL proficiency’ and

the relationship between ‘capital’ and ‘CHL proficiency’ were examined separately. This

section extended the scope and depth of the findings from the quantitative data analysis by

investigating the co-contribution of ‘habitus’ and ‘capital’ to CHL proficiency. Multiple

235

regression was used to model the relationship. ‘CHL proficiency’ was treated as the

dependent variable and ‘Chineseness’ and ‘capital’ were treated as independent variables. A

dependent variable refers to the variable whose variance is the presumed effect of, or

response to, the variance of the variables that have been manipulated; while an independent

variable, or a predictor variable, refers to the manipulated variable that is the presumed cause

of the dependent variable (Field, 2009; Hair et al., 2006).

As illustrated in the earlier sections, the scale score for ‘CHL proficiency’, ‘capital’,

and ‘Chineseness’ can be computed as continuous variables by multiplying the individual’s

raw score on each indicator by the proportionally weighted factor score of each indicator and

summing. The formulae for computation are summarised in Table 5.74.

Table 5.74

Formulae for computation of the constructs ‘Chineseness’, ‘capital’, and ‘CHL proficiency’

Chineseness v9 × .18 + v7 × .21 + v4 × .21 + v3 × .16 + v2 × .23

Capital economic capital × .11 + cultural capital × .29 + social capital × .22 +

symbolic capital × .38

CHL proficiency v41 × .11 + v38 × .33 + v37 × .40 + v35 × .16

As shown in Table 5.75, any pair of variables had a significantly medium positive

correlation. Multicollinearity was not suggested at this stage because there was no correlation

above .90 (Field, 2009).

Table 5.75

Correlation matrix for the variables

CHL proficiency Chineseness Capital

Pearson Correlation

CHL proficiency 1.00 .64 .79

Chineseness .64 1.00 .76

capital .79 .76 1.00

Sig. (1-tailed)

CHL proficiency . .000 .000

Chineseness .000 . .000

capital .000 .000 .

236

The way in which independent variables are entered into the model can have a great

impact on the values of the regression coefficients. There are three methods for entering

independent variables into a model. In the hierarchical blockwise entry method, independent

variables should be entered into the model in order of their importance in predicting the

dependent variable. According to Bourdieu, habitus and capital contribute to certain language

choices in a given field, with neither habitus nor capital primary or dominant. Therefore, the

hierarchical blockwise entry was considered inappropriate for the current study. In the

stepwise entry method, the order in which independent variables are entered into the model is

based on a purely mathematical criterion. This takes many important methodological

decisions out of the hands of the researcher. Therefore, stepwise entry method is best avoided

except for exploratory model building (Field, 2009). Unlike the first two methods, forced

entry is a method in which independent variables are forced into the model simultaneously.

This method is believed to be the only appropriate method for theory testing (Field, 2009;

Studenmund & Cassidy, 1987). As such, entering ‘Chineseness’ and ‘capital’ at the same time

into the model can test whether the data fit Bourdieu’s theory. Therefore, the forced entry

method was selected.

As shown in Table 5.76, 62% of the variation of ‘CHL proficiency’ was explained by

the variation of ‘Chineseness’ and ‘capital’. For any two observations, the residual terms

should be uncorrelated or independent (Field, 2009). This assumption can be tested by

Durbin-Watson test. The test statistics can vary between 0 and 4 with a value of 2 meaning

that the residuals are uncorrelated (Durbin & Watson, 1950, 1951). As a rule of thumb, values

less than 1 or greater than 3 should raise concern (Field, 2009). The Durbin-Watson statistics

was 1.97, which was very close to 2 (see Table 5.76). This indicated that the assumption was

adequately met.

237

Table 5.76

Model summary

Model R R

square

Adjusted R

square

Std. Error of

the estimate

R square

change F change df1 df2

Sig. F

change

Durbin-

Watson

1 .79a .62 .62 1.14 .62 187.67 2 227 .000 1.97

a. Predictors: (Constant), capital, Chineseness

b. Dependent variable: CHL proficiency

ANOVA results were demonstrated in Table 5.77. It indicated that the model was a

significant fit of the data overall. ‘Chineseness’ and ‘capital’ together made a statistically

significant contribution to ‘CHL proficiency’ (F = 187.67, p < .001).

Table 5.77

ANOVA results

Model Sum of squares df Mean square F Sig.

1 Regression 485.77 2 242.87 187.67 .000a

Residual 293.79 227 1.29

Total 779.56 229

a. Predictors: (Constant), capital, Chineseness b. Dependent variable: CHL proficiency

As shown in Table 5.78, when ‘Chineseness’ and ‘capital’ worked together, ‘capital’

was a significant contributor to ‘CHL proficiency’ when ‘Chineseness’ was held constant (t =

11.44, p < .001); while ‘Chineseness’ was not a significant contributor to ‘CHL proficiency’

when ‘capital’ was held constant (t = 1.40, p = .163). This issue was examined from a

statistical perspective.

Table 5.78

Model parameters

Model

Unstandardised

coefficients

Standardised

coefficients t Sig.

Collinearity

statistics

B Std. Error Beta Tolerance VIF

1 (Constant) .24 .24 1.02 .310

Chineseness .12 .08 .09 1.40 .163 .42 2.39

capital .92 .08 .72 11.44 .000 .42 2.39

Dependent variable: CHL proficiency

238

The Variance Inflation Factor (VIF) and the tolerance statistics can be used as measures

for multicollinearity. The largest VIF greater than 10, the average VIF substantially greater

than 1, and/or the tolerance less than .20 are indicative of multicollinearity (Field, 2009). For

the current model, VIF values of 2.39 were well below the cutoff of 10 and the tolerance

statistics of .42 were above .20 (see Table 5.78). Multicollinearity was not diagnosed at this

stage.

However, a closer examination of the collinearity diagnostics identified a problem. The

variance proportions for each independent variable should be distributed across different

dimensions. For the current model, ‘Chineseness’ and ‘capital’ had most of their variance

loading onto the same dimension, with 95% of the variance of ‘Chineseness’ and 79% of the

variance of ‘capital’ on dimension 3 (see Table 5.79). Multicollinearity between ‘Chineseness’

and ‘capital’ was diagnosed. The correlation matrix was revisited (see Table 5.75). The

correlation between ‘Chineseness’ and ‘capital’ (r = .76, p < .001) was stronger than the

correlation between ‘Chineseness’ and ‘CHL proficiency’ (r = .64, p < .001). Since the

correlation between the two independent variables (‘Chineseness’ and capital) was stronger

than that between the independent variable (‘Chineseness’) and the dependent variable (CHL

proficiency), multicollinearity between ‘Chineseness’ and ‘capital’ was diagnosed from a

statistical perspective. This will be discussed from a Bourdieusian theoretical perspective in

Chapter Seven.

Table 5.79

Collinearity diagnostics

Dimension Eigenvalue Condition index Variance proportions

(Constant) Chineseness Capital

1 2.90 1.00 .01 .01 .01

2 .07 6.41 .92 .05 .20

3 .03 10.46 .07 .95 .79


239

The multicollinearity and independence of residuals have been checked earlier. At the

final stage in this regression analysis, the assumptions of normality and homoscedasticity (the

residuals at each level of the independent variable have similar variances) were checked. The

distribution of the standardised residuals of ‘CHL proficiency’ was roughly normal with a

nearly bell-shaped curve, although there was a slight left-skewness. The normal probability

plot also demonstrated a roughly normal distribution. Although the points of the observed

residuals distributed up and down along the normal distribution line, there was no substantial

deviation from the straight line (see Figure 5.27).

Figure 5.27. Histogram and P-P plots of the standardised residuals of the ‘CHL proficiency’

The scatter plot of the standardised residuals against the standardised predicted values

as well as that of the studentised residuals against the standardised predicted values

demonstrated a random array of dots evenly dispersed around zero. This indicated that the

assumptions of normality and linearity were met (see Figure 5.28).

240

Figure 5.28. Scatter plots of the standardised and the studentised residuals of ‘CHL proficiency’

against the standardised predicted values of ‘CHL proficiency’

The partial plots were the scatter plots of the residuals of ‘CHL proficiency’ and those

of ‘capital’ and ‘Chineseness’, when ‘CHL proficiency’ was regressed separately on ‘capital’

and ‘Chineseness’ (see Figure 5.29). The plots demonstrated a strong positive relationship

between ‘CHL proficiency’ and ‘capital’ and between ‘CHL proficiency’ and ‘Chineseness’.

There were no obvious outliers on the plots. Although the relationship between ‘CHL

proficiency’ and ‘Chineseness’ looked less linear than that between ‘CHL proficiency’ and

‘capital’, the clouds of dots were evenly spaced around the gradient line in both graphs.

These were indicative of homoscedasticity.

Figure 5.29. Partial plots of the residuals of ‘CHL proficiency’ against ‘capital’ and ‘Chineseness’

241

Despite the glitch of multicollinearity between ‘capital’ and ‘Chineseness’, the

assumptions of independence of residuals, normality, and homoscedasticity have been met. It

can probably be assumed that the model appeared in most senses to be both accurate for the

sample and generalisable to the population.

5.3 Chapter summary

This chapter reported the strategies of the quantitative data analysis in detail. At the first stage

of the pilot study, face validity was improved by adjusting the wording of some items

according to the feedback from five participants. The second stage of the pilot study worked

with the data produced by 38 complete cases. The inter-item reliability of each theoretical

construct was checked in turn. Taking account of both statistical and theoretical

considerations, problematic items were excluded from the model. The overall inter-item

reliability was considered to be improved.

In the main study, data were produced by 230 complete cases. EFA was first applied to

each theoretical construct. Results from EFA suggested the basis for CFA. Each theoretical

construct was then specified in a measurement model by CFA as a latent variable with its

respective indicators. Both statistical and theoretical considerations were taken into account

when excluding problematic indicators from the model to improve the model fit. Under the

modified measurement model, an equation to calculate the scale score of each theoretical

construct was offered. Four measures of reliability, namely SMC, construct reliability,

variance extracted, and coefficient H, as well as two measures of validity, namely construct

validity and convergent validity, were also checked to confirm the reliability and validity of

the modified measurement models.

242

Relationships between these theoretical constructs were specified in the structural

models according to the first part of the theoretical framework (Figure 4.1). The structural

models served the hypothesis testing. The results from SEM rejected both null hypotheses

and led to the conclusion that both ‘Chineseness’ (r = .71, p < .001) and ‘capital’ (r = .86, p

< .001) had statistically significant positive contributions to ‘CHL proficiency’. The analyses

demonstrated that ‘Chineseness’ (habitus) of young Chinese Australian adults and various

resources (capital) available to them explained their CHL proficiency. This was interpreted

through Bourdieu’s argument that language choice can be attributed to agents’ dispositions

(habitus) and various resources (capital) available to them.

In the multiple regression analysis, ‘CHL proficiency’ was regressed on ‘Chineseness’

and ‘capital’. ‘Chineseness’ and ‘capital’ accounted for 62% of the variance of ‘CHL

proficiency’. Statistically, there was a problem of multicollinearity between ‘Chineseness’

and ‘capital’. Theoretically, this multicollinearity aligns with Bourdieu’s (1973; Bourdieu,

1985a) argument that habitus can be understood as a form of capital when valued and

recognised in a given field. The connection between participants’ ‘Chineseness’ and cultural

capital may explain the fact that the contribution made by ‘Chineseness’ was actually

embedded in the contribution made by ‘capital’. This will be further discussed in the final

chapter.

This chapter has reported on the investigation of the first research question. If

‘Chineseness’ qua habitus and accessible resources as ‘capital’ can produce CHL proficiency,

it will be meaningful to go one step further and ask: Does investment in CHL have any

returns? What are the returns on this investment? This leads to the second research question:

How do young Chinese Australian adults understand their CHL learning in relation to

(potential) profits produced by this linguistic capital in given fields? This will be explored

through the qualitative phase reported in the next chapter.

243

Chapter 6: Qualitative Data Analysis and Findings

This chapter presents the analysis of the interview data and reports the findings from the

qualitative phase of the research. The focus of this chapter addresses RQ2: How do young

Chinese Australian adults understand their CHL learning in relation to (potential) profits

produced by this linguistic capital in given fields? The examination of this research question

occurs within the second part of the theoretical framework summarised in Figure 4.4 and

explores the meanings that participants attached to CHL in terms of its profit in capital and

habitus (re)production within particular fields. This chapter is organised in five sections.

Section 6.1 explains the rationale behind the purposive selection of interview participants.

Section 6.2 draws portraits of the five participants and outlines their key characteristics

relevant to this study. Section 6.3 details the interview data analysis and presents the findings.

Section 6.4 considers how the findings from the qualitative phase talk back to the findings

from the initial quantitative phase. Section 6.5 concludes this chapter with a brief summary.

6.1 How were the five participants selected?

Nineteen participants indicated at the end of the online questionnaire that they were willing to

participate in an interview and left their contact details for arranging a time and place of the

interview. Their respondent numbers in the online questionnaire were listed as follows: 30, 46,

49, 59, 66, 77, 79, 87, 89, 91, 92, 120, 129, 134, 139, 141, 225, 227, and 228. These

respondents became candidates for the interview.

As stated in the description of the design of the interview in Chapter Four, Patton’s

(1990) extreme sampling was applied at the first stage to illuminate dimensions behind the

244

average cases. Outliers in the initial quantitative phase were considered as extreme cases for

the purposes of the qualitative phase. To identify these outliers, ‘CHL proficiency’ was

regressed on ‘Chineseness’ and ‘capital’ respectively. Outliers outside ±2 standard deviations

were highlighted. This is because only about 5% of the cases will fall outside ±2 standard

deviations in a normal distribution. Table 6.1 demonstrated the outliers in the regression

relationship between ‘CHL proficiency’ and ‘Chineseness’ as well as in that between ‘CHL

proficiency’ and ‘capital’.

Table 6.1

Case-wise diagnostics

Chineseness-CHL proficiency

Respondent no. Std. residual CHL proficiency Predicted value Residual

5 2.067 7.00 4.0542 2.94583

9 -2.314 1.27 4.5683 -3.29831

29 2.238 6.78 3.5898 3.19021

33 2.100 5.72 2.7274 2.99263

67 2.337 6.49 3.1586 3.33142

83 -2.800 1.00 4.9912 -3.99122

88 -2.826 1.71 5.7375 -4.02755

114 -2.197 2.15 5.2815 -3.13146

122 -2.599 1.32 5.0244 -3.70439

130 -2.269 1.16 4.3942 -3.23416

134 -2.174 1.32 4.4190 -3.09904

142 2.055 7.00 4.0708 2.92925

218 -2.406 2.15 5.5800 -3.42999

228 2.299 7.00 3.7225 3.27753

Capital-CHL proficiency

Respondent no. Std. residual CHL proficiency Predicted value Residual

5 2.82 7.00 3.78 3.22

27 2.42 6.27 3.51 2.76

42 2.94 6.78 3.43 3.35

44 2.03 6.89 4.58 2.31

62 2.31 6.67 4.04 2.63

63 -2.44 1.99 4.77 -2.78

79 2.46 6.56 3.75 2.81

86 -2.05 1.59 3.93 -2.34

88 -2.73 1.71 4.82 -3.11

89 -2.28 2.76 5.36 -2.60

129 3.48 6.34 2.37 3.97

140 2.79 6.60 3.42 3.18

142 3.59 7.00 2.91 4.09

230 2.62 7.00 4.01 2.99


245

These outliers were extreme cases because their ‘CHL proficiency’ could not be well

predicted by ‘Chineseness’ and/or ‘capital’. Among these extreme cases, five candidates (79,

89, 129, 134, and 228) had indicated their interest in participating in an interview. Invitations

were sent to these candidates. Only one respondent (assigned the pseudonym, Adam)

accepted the interview invitation.

At the second stage, Patton’s (1990) maximum variation sampling strategy was applied.

The maximum variation in CHL proficiency was the main concern. The scale score of ‘CHL

proficiency’ of interview candidates that were not extreme cases (30, 46, 49, 59, 66, 77, 87,

91, 92, 120, 139, 141, 225, and 227) was revisited. Their scores presented a wide spread,

from 2.21 (respondent 46) to 6.89 (respondent 77). To capture the maximum variation in their

CHL proficiency, interview invitations were sent to all of these candidates. Four respondents

(Bob, who chose the pseudonym; and Crystal, Dianna, and En-ning, who were assigned the

pseudonyms) agreed to participate in the interview. The scores of their CHL proficiency were

2.21, 2.57, 6.27, and 2.89 respectively. Some of their demographic features also demonstrated

a wide variation in terms of their birthplace, age at which they moved to Australia if born

overseas, and years of formal CHL learning. Taking account of their difference in CHL

proficiency and demographic features, “maximum variation” was considered to be garnered.

Table 6.2 presents an overview of the four respondents, who offered maximum variation (Bob,

Crystal, Dianna, and En-ning), as well as the respondent (Adam), who was identified as an

extreme case.

246

Table 6.2

Overview of participants

Respondent 89 46 87 139 225

Pseudonym Adam Bob Crystal Dianna En-ning

Gender Male Male Female Female Female

Age in years 28 18 18 21 23

Birthplace Indonesia Hong Kong Australia Taiwan Australia

Age of immigration if

born overseas 12 9 months N/A* 13 N/A*

Generation 1st 1

st 2

nd 1

st 2

nd

Resident city Sydney Brisbane Sydney Brisbane Canberra

Citizenship Australian Australian Australian Australian Australian

Language usage at

home

English and

Indonesian with

a few Mandarin

words

English and

Cantonese English

Hokkien and

Mandarin English

Years of formal CHL

learning None 1 5 13 10

CHL proficiency 2.76 2.21 2.57 6.27 2.89

Number of interviews 3 1 1 2 2

N/A*: Born in Australia

6.2 Who were the five participants?

Each participant was interviewed about life experiences associated with their CHL learning

and their Chinese heritage background (see interview schedule in Appendix 4). This section

briefly introduces the five participants and outlines their key characteristics related to this

study. Further thematic analysis of their interview data will be presented in the next section.

Adam: “I just need to catch up!”

Twenty-eight-year-old Adam was born in Indonesia. He moved to live in Australia at the age

of 12. He also lived in Singapore for one year at the age of 21 on an exchange program.

Adam was considered an extreme case because the statistical models could not predict his

CHL proficiency given his investment of various forms of capital. As such, he constituted an

‘outlier’ in the initial quantitative phase. He seemed to be very aware of this, as he agreed, “I

247

think my case is a little bit unique.” This uniqueness may be attributed to his life trajectory.

Adam was born in Jakarta, Indonesia, which he thought was “one of the most challenging

aspects to learning Mandarin”. To understand the challenge that Adam confronted as a CHLL

when growing up in Jakarta, it is useful to provide some background information.

Discrimination against Chinese Indonesians is evident throughout the history of

Indonesia, although more recent government policies have attempted to redress this.

Resentment of Chinese economic success grew in the 1950s as native Indonesian merchants

felt they could not remain competitive. In response, the government approach was to

propagate a stereotype that conglomerates owned by Chinese Indonesians were corrupt. As a

result, the word ‘Chinese’ became synonymous with corruption in the perception of most

Indonesians (Purdey, 2006). The situation deteriorated when General Suharto came into

power in 1967. His approach was to forcibly assimilate Chinese Indonesians, taking

advantage of Chinese economic success whilst eliminating their perceived economic

dominance (Purdey, 2006). In order to do this, he sought a depoliticised system where

formation of a cohesive ethnic Chinese identity was no longer allowed (Purdey, 2006).

Expressions of Chinese culture through language, religion, and traditional festivals were

banned. Chinese Indonesians had to relinquish their Chinese names and ‘happily’ adopt

Indonesian-sounding names. Anti-Chinese sentiment gathered intensity in the 1990s when

major riots broke out in many Indonesian cities. Property and businesses owned by Chinese

Indonesians were targeted by mobs. Many women were sexually assaulted and numerous

people died (Purdey, 2006) . Following these events, large numbers of Chinese Indonesians

fled the country. In the late 1990s, the Asian Financial Crisis had dire consequences for the

Indonesian economy and shook Suharto’s regime. After nation-wide demonstrations against

the presidency organised by university students, Suharto’s 31 years of power came to an end.

248

Subsequent governments launched campaigns to abolish anti-Chinese regulations and rebuild

the confidence of Chinese Indonesians. Although Chinese Indonesians have gradually gained

political and social freedoms as a result of policy reform efforts, anti-Chineseness is still an

issue in Indonesian. It was against this backdrop that Adam was born and brought up before

moving to Australia at the age of 12. He noted, “Because of the challenge, I’ve got a lot of

catching-up to do.”

Bob: “Chinese is part of me.”

Eighteen-year-old Bob was born in Hong Kong and moved to live in Australia when he was

nine months old. He had returned from a two-week family trip in Hong Kong shortly before

his interview. Bob explained, “I just went there with my dad. I haven’t been there for 15 years.

So I went there and we saw our family.”

Bob was the only participant who did not use a single Chinese word during the

interview. Bob’s self-reported CHL proficiency measured in the online questionnaire was the

lowest of the five participants (as shown in Table 6.2). This may be the reason behind his

scant usage of CHL. Even if he did not have many opportunities to use his CHL in his

“younger years”, he intentionally chose Chinese when both Chinese and French lessons were

available in his primary school. In response to the question about the reasons behind this

effort, Bob said, “Because of my Chinese heritage. French would be good but I think it’s just

nothing part of me. But Chinese is part of me. I just think it’s a good opportunity to do that.”

Although he came to live in Australia when he was only nine months old, his embodied

dispositions, or in his own words “part of me”, was durable and transposable. He felt very

comfortable with it. As he said, “You can’t pass Bruce Lee and his movies. He made me feel

happy about my heritage when I was watching these movies and stuff.” He had developed a

249

cultural attachment to the long history of Chinese Kung Fu that contributed to his degree of

comfort with his Chinese heritage.

Crystal: Chinese entertainment stuff is “really funny”.

Eighteen-year-old Crystal was born in Australia. She was a second-generation Chinese

Australian. She said she enjoyed Chinese TV shows. In her own words, “It’s really funny (to

watch Chinese TV shows)!” When asked whether she was trying to improve her Chinese, she

said, “Yes, that’s part of the reason why I watch Chinese TV shows.” For Crystal, watching

Chinese TV shows and learning Chinese seemed to be mutually reinforcing. As she indicated,

watching Chinese TV shows was one way to improve her Chinese proficiency. In return, her

Chinese, to a certain degree, helped her access a wider range of Chinese entertainment. She

reported typing Chinese and searching Chinese websites for Chinese TV shows, which were

not available in English websites, as she said, “You don’t have this kind of thing in Australia.”

When asked whether her CHL benefited her social life in Australia, she said, “I can find my

own entertainment so I can also talk to other people about entertainment. When I go to

Karaoke, I can sing Chinese songs. Yes, I guess it helps me.” Crystal can thus share her

knowledge of Chinese entertainment in a Chinese-speaking community in Australia, where

Chinese popular culture is of common interest. In this situation, CHL proficiency helped her

not only access a wider range of Chinese entertainment, but also build social networks.

Dianna: “I felt how important (it was) to speak my own home language!”

Twenty-one-year-old Dianna was born and brought up in Taiwan. She moved to live in

Australia at the age of 13. When asked to recall her school life in Taiwan, she said she

250

“probably forgot everything”. Interestingly, when asked whether her CHL had ever made her

feel special, she immediately recalled the following experience. Once at lunchtime in high

school, her white Australian classmates were “curious” and asked her and her Chinese friend

whether they could speak Chinese. Dianna said she could because she used Chinese at home.

She then said a few words in Chinese. Her Chinese friend felt “embarrassed” because she was

not able to speak Chinese at all. Dianna recalled, “At that time I felt how important (it was) to

speak my own home language.” She could clearly remember what her white Australian

friends said at that moment, “If you are Asian, you should be able to speak your home

language. If you can’t, it’s sort of a shame.” Dianna considered her CHL a source of “pride”

to her and “part of” her “Chinese identity”. This perception was deeply rooted in herself,

which can be understood as constituting a durable and transposable disposition. As she added,

“I still need to keep my home language going no matter where I am. So I will pass it on to my

children as well.”

En-ning: “Learning Chinese is definitely something I am going to pursue in the rest of

my life.”

Twenty-three-year-old En-ning was born in Australia. She was a second-generation Chinese

Australian. At the time of the interview, she was attending Chinese programs at a prestigious

university in China. She reported that she did not see any point in learning Chinese when she

was a child, growing up in a “very white area”. This implied that she did not have many

opportunities to use Chinese at that time and she was not interested in learning the language

at all. However, she said she did not know what “came over” her in Year Nine when she

decided to learn Chinese in school. From then on, she started to realise that learning Chinese

was “an incredible personal experience” for her and she was trying to do “as much as” she

251

could in Chinese. In return, En-ning profited from her CHL learning. Her CHL proficiency

helped her win a prestigious scholarship at a leading Australian university and a national

scholarship awarded by the Australian Federal Government. She confessed, “Of course I

don’t think I would have got the scholarships if I didn’t have at least some Chinese language

background.” En-ning spent a gap year in China, which “totally changed” her life. She was

able to “explore the other side”, or “the Chinese side” of herself. She said, “I feel a lot

happier, especially in my gap year when I had the opportunity to explore that side of who I

am.” Learning Chinese was so “amazing and incredible” for En-ning that she said, “Learning

Chinese is definitely something that I am going to pursue in the rest of my life.”

6.3 What did CHL mean to the five participants?

Different reasons and various meanings behind CHL learning emerged from the above

glimpse of the five participants. When legitimised in given fields, participants’ CHL

competence could serve as linguistic capital. The exchange value of this linguistic capital can

produce benefits or profits in these fields. These benefits and profits can be understood as a

return on investment in CHL learning. The quantity and quality of production was dependent

upon the structure of the participants’ lived social worlds at a given time. As such, CHL was

reported to have rich and varied meanings for all participants. These meanings emerging from

the interview data resonated with Bourdieu’s theorisation of habitus, capital, and language

choice within a given field. The data are reported, interpreted, and analysed thematically in

the following sections.

252

6.3.1 Potential production of economic capital through CHL learning

When asked whether their Chinese language could contribute to job opportunities in future,

all participants indicated that one of the benefits from learning Chinese was the expectation

of more and better job opportunities. Dianna shared a story about one of her Chinese friends

who got a job as a translator in a company because of his Chinese proficiency. Dianna

seemed to cast no doubt on the value of Chinese competence in the labour market. She

described always telling her friend’s story to her brother and saying, “You will be thankful

when you get a job because of your ability to speak two languages.” She explained, “When

you are looking for a job, people tend to consider someone who can speak two languages or

more than two. You will find yourself getting that priority.” By saying this, she indicated her

belief that her Chinese competence would be valued in such a field where people considered

both Chinese and English important. When valued in a future job market, her CHL

competence will serve as linguistic capital by which she can “get that priority”.

This view was shared by all participants. Adam, who was doing research in air

pollution, hoped that his Chinese competence would open up opportunities to find work to do

with air pollution in China. Bob, who was studying paramedical science, reported that his

Chinese competence might be helpful if he wanted to get a job in Hong Kong as a paramedic.

Crystal, who wanted to possibly work in Hong Kong one day, considered that her Chinese

competence might function as an instrumental advantage for her. En-ning, who changed her

major from arts and law to Asian studies and law, considered her Chinese competence useful

if she wanted to “pursue Australia-China relations or work as a lawyer” in her future career.

In summary, all participants reported if valued in a future labour market, or a given

field in Bourdieu’s sense, their CHL competence would become a valuable asset for them.

These future labour markets where Chinese is valued will become extra fields open to these

253

participants. This accords with the findings indicated by previous studies (Francis et al., 2009;

Weger-Guntharp, 2006; Wong & Xiao, 2010) that CHL competence is explicitly associated

with extra job opportunities, and ultimately convertible into economic capital.

6.3.2 Production of cultural capital through CHL learning

When asked to explain their Chinese learning experiences and practices, Adam said that he

bought some Chinese textbooks and worked through them by himself; Crystal reported that

she bought some Chinese TV series and watched them as a way of learning Chinese; and

Dianna indicated that she collected some Chinese books for reading and family discussion.

These Chinese textbooks, Chinese TV series, and Chinese literature are material cultural

goods that can be understood as objectified cultural capital produced through CHL learning.

When asked to talk about language and culture, Adam explained, “I have a bit of a

philosophical aspect in learning a language or any languages because it provides a wider

window to look into the culture and see how people react to things.” This philosophy made

sense when he reported that he started to understand local people’s particular ways of doing

things during his Chinese language study tour in Shanghai. These understandings gained

through CHL learning became an embodied asset. As Bourdieu (1991) explained, embodied

cultural capital would include the know-how capacity that people bring with them when

moving across different social spaces. By virtue of these embodied resources, when Adam

accompanied his sister to visit Shanghai, he became the person “who showed her around”,

told her “how unique people in Shanghai are”, and “trained her how to cross the road safely”.

Apart from his capture of the local ways of doing things, Adam also reported gaining

some Chinese cultural knowledge through his Chinese learning. He explained:

254

We can use the word ‘networking’, but it’s not as great as the Mandarin word, (I am

sorry for my tones), ‘关系’. In China, ‘关系’ is a lot deeper concept. It’s more about

knowing the right people and somehow having a working relation. It’s like a ‘you

scratch my back and I scratch your back’ thing.

This excerpt demonstrated his understanding of the special meaning of the Chinese word ‘关

系’ , equivalent to the English word ‘networking’, within a Chinese cultural context. He

gained this Chinese cultural knowledge through his CHL learning. Later he added:

I am always fascinated with the old Chinese sayings, which still ring true to this day. It

just has a deeper meaning when you say it in Mandarin or even in classical Chinese

instead of reading the English translated one.

He gave a few examples of the old Chinese sayings and said, “Remember the wise sayings

from our forefathers: 一山还有一山高 (there is not a mountain which is the tallest); 天外

有天 (there is always another heaven beyond this one).” These sayings indicate that no one

can be the best and people should be humble. It was apparent that Adam appreciated the

meaning behind these sayings. He referenced these sayings when trying to argue that he

would not feel too proud of his Chinese competence, though he was happy with his progress

in Chinese. Through CHL learning, Adam reported that he had gained Chinese cultural

knowledge and embedded this knowledge into his mind and body, from which his

behavioural patterns, actions, and perceptions emerged.

Bob also reported gaining some Chinese cultural knowledge through his Chinese

learning. When asked to talk about language and culture, he reported that his mum taught him

255

something about “the formation of Chinese characters” and “how different characters had

changed over the years.” It was his mum’s inculcation through CHL practice that generated

this aspect of his Chinese cultural knowledge. Similarly, Crystal reported that she really

enjoyed Chinese entertainment, especially Chinese TV series, and she attributed this taste to

her upbringing, as she explained:

I grew up watching that. When I was very little, my dad introduced me to these series.

My parents never formally taught me Chinese really. But they show me TV shows. We

used to watch together. That’s what I did.

As Bourdieu (1973) argues, the reproduction of cultural capital emphasises the importance of

the early familial environment for the learning and accumulation of cultural capital. By this

he meant that reproduction of cultural capital is not established in a vacuum but is

intergenerationally sponsored. This intergenerational influence urges each agent to reproduce

the cultural capital of his or her group, augmenting it if possible. Through CHL learning,

Bob’s mum and Crystal’s dad were able to pass on the Chinese cultural knowledge to them as

the reproduction of embodied cultural capital. When asked whether her Chinese language

benefited her, Crystal said:

Yes, I guess so. It exposes me to more entertainment stuff. Before, when I was little, I

used to watch series but my dad had to give me the series to watch. Now I can go

online and type in Chinese to find my own series to watch.

256

By saying “more”, Crystal indicated that she would not have been able to access Chinese

entertainment resources without her CHL learning experience and CHL competence. When

she was little, her dad had to mediate access to these resources for her because her Chinese

was too limited. When she grew up, she was able to use her Chinese to search Chinese

websites for “more entertainment stuff”. Moreover, both Crystal and En-ning reported that

learning Chinese enabled them to sing Chinese songs. For Adam, Bob, Crystal, and En-ning,

their Chinese cultural capability generated through learning CHL can be understood as

embodied cultural capital that helped them access a wider range of Chinese cultural

resources.

In addition to objectified and embodied cultural capital acquired through CHL learning,

CHL competence was also reported to produce institutionalised cultural capital. All

participants except Adam reported that they had undertaken formal Chinese learning in

schools or at universities.

When I was probably in Grade Three or Grade Four, my parents sent me to a Chinese

school to learn Mandarin… In Grade Seven…Chinese was the thing in the curriculum

(of my school). We got to learn Chinese. (Bob)

I think I took my first formal Chinese lesson when I was in Year Three at my primary

school… I did two years of Chinese at high school in Year Seven and Year Eight…

From Year Nine to Year 11, I went to my community Chinese school, run by Chinese

people… I am going to do my first Chinese course this coming semester (at my

university). (Crystal)

257

I went to one (Chinese community school) in Sunnybank for a few months. (Dianna)

I went to Cleveland Street Saturday Chinese School until…possibly Year Six. And

then...my school started offering Chinese. And then I did it in Year Nine… I did

Chinese Five, Chinese Six, and Chinese Seven in my first semester and the second

semester and the second year first semester (at my university)… And then I went for an

exchange in Santa Barbara in California at this time last year for a semester. I took like

three Chinese language classes as part of my exchange requirements. (En-ning)

Because the Chinese courses were part of participants’ (except Adam) curriculum,

participants (except Adam) received credits towards the programs that they were enrolled in.

These credits can be understood as institutionalised cultural capital.

6.3.3 Production of social capital through CHL learning

When asked about their language choices at home, at work, with friends, and in their social

lives, all participants indicated that Chinese was useful and beneficial in at least one of these

settings. First of all, they all indicated that their Chinese competence helped strengthen their

family ties. By their accounts, this was particularly important when some family members did

not have, or did not have enough, English competence. Dianna reported that Chinese usage at

home was the only way to communicate with her parents because they did not have any

English. Crystal and En-ning reported that Chinese usage helped them communicate “more”,

particularly with their grandparents:

258

I would like to communicate with my family a bit more. I really like my more distant

family (grandparents), so I would like to be able to talk to them more when I visit them

and tell them about my life. (Crystal)

It was very hard to communicate with them (grandparents) before I properly learnt

Chinese. I guess it’s another part of why learning Chinese is a personal thing because I

want to better speak to my grandparents. (En-ning)

In the excerpts above, Crystal and En-ning used the words “more” and “better”. By this they

meant something they did not have or did not have enough of before. The connotation

embedded in the words “more” and “better” resonates with the capital metaphor because

capital can accumulate and generate ‘added value’. For these participants, the added value

generated through CHL learning was better intergenerational communication, by which their

family ties could be strengthened.

Secondly, CHL competence could reportedly build friendship networks. All

participants talked of speaking Chinese with their friends, either in Australia or in China, or

even in other parts of the world. When asked about their language choices with friends,

Crystal said, “I also have some mainland China friends whom I like to talk to in Chinese

sometimes.” In this respect, CHL is important for Crystal to maintain and strengthen the ties

with her friends in the Chinese Mainland. When asked the same question, Dianna explained:

I do have friends from Hong Kong, or China. We do speak Chinese to each other, but

not in class to be polite. If you meet someone with an Asian background who can speak

Chinese, you always feel comfortable to speak Chinese to them.

259

Later she added, “I tend to be with Asians (Chinese) more, so we can speak our home

language, Mandarin Chinese. Sometimes we can speak about things that we can’t describe in

English. We are comfortable that way.” Her friendship could be strengthened through this

more comfortable way of communication in Chinese, especially when there was something

that they could not “describe” in English. In such a situation, CHL had a unique and

instrumental function to facilitate communication between Dianna and her Chinese friends,

which they could not achieve through any other language. En-ning also reported how her

Chinese competence benefited her friendships:

I think of how many diverse and amazing experiences I have been able to have as a

result of coming to China and learning Chinese. For example, I have memories of

overseas Chinese from all over the world hanging out in a dorm room and sharing

stories about our experiences as overseas Chinese…Chatting to a bunch of Chinese

girls on a train, you know anything that is relevant to their lives. I think if I can’t speak

some Chinese, I would not have access to it.

Moreover, CHL competence by the participants’ accounts benefited their daily lives.

All participants described experiences when their Chinese competence helped them socialise

with other people. Adam used to mix with other Chinese-speaking people in a Chinese mall

when he was in Singapore. When asked in the second interview why he wanted to ‘catch up’

with his Chinese, Adam added:

I want to catch up with my Chinese language so that I can converse with anybody easier.

I think I am sort of on my way there. For example, I just helped a couple of my friends

260

with their graduation and most of them are Malaysian. Even a year ago, in the previous

graduation season, I wouldn’t even think of being able to talk with any of their parents

in Chinese. This year at least I could say a few things.

As a return on his efforts in CHL learning, Adam was able to socialise in a more active way.

Bob and Crystal reported the experience at work of using Chinese to help people who could

not speak English. Crystal was also very happy that she was able to use Chinese in China

when shopping. Dianna taught small children simple Chinese keywords or sang Chinese

songs with them over the Chinese New Year. En-ning was studying in China when the

interviews with her were conducted. She indicated that her Chinese language competence

was indispensable because she had to use Chinese in her daily life to socialise with other

Chinese people.

In contrast, a lack of CHL proficiency was reported to hinder the establishment and

development of social networks. When asked whether it was a struggle to speak Chinese, Bob

said:

Definitely I was keen to socialise with people but it’s hard because I don’t know what

to say. If I knew what to say, I would definitely say (something). Yes in a way, it was a

struggle. But I was willing to try and make that effort to say something.

On the one hand, a lack of CHL proficiency precluded Bob from socialising with

Chinese-speaking people. Without certain level of CHL proficiency, it was difficult for Bob

to establish social networks with Chinese-speaking people. On the other hand, Bob described

trying his best to practice his CHL in order to better enable socialising.

261

In brief, CHL competence has maintained, extended, or improved the social networks

for all participants in their Chinese-speaking social world, in terms of tightening family ties,

developing new friendships or improving their existing ones, and socialising with other

people. This is consistent with the existing literature. As indicated in previous studies

(Francis et al., 2009; Wong & Xiao, 2010; J. Zhang, 2009), CHL as a means of

communication has a functional significance in terms of establishing valuable networks

within fields where Chinese is the medium of social relations.

6.3.4 Production of symbolic capital through CHL learning

CHL competence reportedly became symbolic capital when recognised or valued by others.

Adam recalled his feeling that local people understood him when he was in Shanghai:

I felt really great! It was just like a little kid managing to do something different for the

first time. I had thought I would have never done it because of all the difficulties that I

had. For them to be able to reply to me: “I understand what you said”, it looked like I

managed to climb up a pretty high mountain and I would be able to do it. The feeling

was like excitement and wanting to do more along the way…Yes, it really made me

think, “I can do this. Let’s learn more.”

This “excitement” and motivation to “learn more” may be attributed to his experience that his

Chinese language was understood, accepted, and legitimised by the local Shanghai people.

Other participants reported similar experiences of recognition and affirmation when their

Chinese language was recognised or valued by others. When asked what other people thought

of his Chinese ability, Bob recalled, “My aunty was pretty nice. She said that I am good for

262

someone who has been brought up in Australia.” When asked about her reasons for learning

Chinese, Crystal said that her friends’ parents made a fuss of her when she talked to them in

Chinese. She recalled, “They kind of like me more.” Dianna had the experience that she was

praised by her parents’ friends for her CHL competence. She recalled what her parents’

friends said to her, “Oh, it’s really good you can still speak Chinese here”. She also recalled

the experience that she was asked by her white Australian classmates in high school to speak

Chinese. When her classmate said, “I only speak English. You can speak two languages”,

Dianna reported that, “I was really, really proud of myself. I have two languages. I don’t feel

any shame at all.” Because her Chinese proficiency was recognised by others, she did not

“feel any shame at all” for being able to speak Chinese. Instead, it was a source of pride for

her. En-ning recalled, “In terms of my grandparents, they are very proud and happy that I am

pursuing this language.” When asked whether learning Chinese has ever helped her win any

qualifications or awards, En-ning was “pretty sure” that “almost everything” she was able to

win was “partially because of” her Chinese learning efforts. These included a prestigious

scholarship at a leading Australian university and a national scholarship awarded by the

Australian Federal Government. As she explained, “Of course I don’t think I would have got

the scholarships if I didn’t have at least some Chinese language background.” Because of the

symbolic value attached to her Chinese language, En-ning said, “I am very aware how lucky I

am to be an Australian with a Chinese heritage.”

For these participants, CHL competence became linguistic capital with legitimised and

symbolic value, which was then convertible into symbolic capital. As indicated in previous

studies (Francis et al., 2009; J. Zhang, 2009), CHL competence helped CHLLs gain praise,

respect, pride, status, and honour.

263

6.3.5 Making sense of the habitus of ‘Chineseness’ through CHL learning

As argued in Chapter Three, habitus is rooted in cultural history, or “previous state” in

Bourdieu’s term (2000, p. 161). For some participants, the origins of their ‘Chineseness’

reportedly carried deep meaning and relevance through their Chinese learning. En-ning took

the opportunity when she was learning Chinese in China to explore her family origin. She

said, “This 春节 (Chinese New Year) I found my grandparents’ 老家 (hometown) in the

south. I spent two weeks there.” It should be noted that the Chinese word ‘老家’ has a

different meaning from the English word ‘hometown’. ‘老家’ is the ‘root’ of a clan, a native

place where one’s ancestors originated and were buried. It signifies not only the physical

place itself but also the genealogical connections and the deep attachments to the land,

customs, and compatriots, forged through generations of shared ancestry, history, culture, and

language. En-ning seemed to attribute this trip to explore her cultural roots to her growing

Chinese competence. As she explained, “I wouldn’t have managed if I didn’t have at least

some Mandarin. Even my parents have never been (there). It was incredible.” En-ning’s

account is in line with the literature that the heritage background nurtures the HL learning

efforts and reinforces HLLs’ identity (Carreira, 2004).

Adam thought that Chinese language resonated with ‘Chineseness’, as he explained,

“For some people, they speak Chinese so that they can claim their ‘Chineseness’”. When

asked about the driving force behind all his efforts in learning Chinese, Adam replied:

The driving force is that I couldn’t do it because of the political reasons. I could keep

saying I was the victim of the political situation back then, but now I have the chance

and I will not let the opportunity pass. I am sort of trying to discover my heritage again.

It was denied but now I have the chance and I am going to grab it.

264

He concluded his first interview with the following comments:

I am not a typical Chinese because most of the time, let’s say Westerners, maybe use

that word, they probably learn Chinese because they have the opportunity and they

have always been able to do it. Well, in my case, I was just a little bit behind because of

the policy I had back then. But of course I can’t say, “Hey, it got banned. I got left

behind. There is no point of learning.” No, that’s not the case. I just need to catch up.

That’s all.

Adam was trying to explain his uniqueness here. In his perception, “Westerners” can “have

the opportunity” to learn Chinese and “they have always been able to do it”. In contrast, he

did not have the opportunity and he was not able to do it while growing up in Jakarta. Here he

might use the terms “Westerners” and “typical Chinese” to refer to white Australians and

Chinese Australians respectively, who were brought up with the opportunity of learning

Chinese. In fact, he did indicate earlier in the interview that the multicultural society offered

Australians the opportunity to learn Asian languages because these languages were “just

around them”. Compared with these “Westerners” and “typical Chinese”, he felt left behind

in terms of Chinese learning. This can explain his lower Chinese proficiency reported in the

quantitative phase. This was also the reason that he wanted to “catch up”. When asked what

exactly he wanted to catch up, Adam replied, “I want to catch up with my Chinese language.”

By this “catch-up”, he explained that he wanted to “discover” his “heritage again”. By saying

“again”, he indicated that he sought to recapture his claim to ‘Chineseness’ that “was denied”

in the past due to the difficult political situation. He added this comment at the end of the

third interview:

265

For me, Mandarin is a tool to enforce the Chinese identity…It’s to satisfy my curiosity

about my heritage and how history, culture, etc. have shaped who I am today as a part

of my own ‘who do you think you are’ project.

It is worth nothing that Adam was referring to the TV genealogy program ‘Who Do You

Think You Are?’ It is an Australian television documentary series, following the BBC series

of the same name. Each episode profiles a celebrity tracing his/her family roots. Adam

referred to this program due to its promotion of genealogy, which resonated with his

“curiosity” about his ‘Chineseness’.

Other participants also claimed that Chinese language carried deeply rooted meanings

for them. When asked why he chose to learn Chinese in school, Bob said, “Chinese is part of

me.” When asked what Chinese language meant to her, Dianna replied, “It’s also part of my

identity, my Chinese identity. I was born into this culture, this colour, and this language, so

it’s part of my identity, my Chinese identity.” When asked, “Have you ever thought about

why they (your parents) pushed you to learn Chinese”, En-ning replied:

Yes, I have thought about it. I guess they want me to have a connection with that side of

who I am…They also gave us (me and my brother) opportunity to explore the other

side, the Chinese side of us when we were very young.

En-ning’s parents hoped to reproduce a sense of ‘Chineseness’ in their children through

learning CHL. En-ning seemed very engaged in this reproduction of ‘Chineseness’. She spent

her gap year in China learning Chinese, which “totally changed” her life. She said, “I feel a

lot happier, especially in my gap year when I had the opportunity to explore that side of who I

266

am.” When asked at the end of the interview to add some more comments to her Chinese

language learning and Chinese background, she said:

I am so glad that I discovered this so early in my life. I do meet a lot of overseas

Chinese who came to China much older, looking for something. But I am glad I came

early…I think if we have the opportunity to explore this other side of who we are or

might be, that can be very empowering, which in turn can very much affect our life

experiences.

In short, participants’ CHL learning and practices resonated with a constitutive dimension and

sense of themselves. As such, CHL learning and practices helped recapture or reinforce a

habitus of ‘Chineseness’ for the participants, which was embraced as “part of” themselves or

“the other side of” themselves. This is in line with the existing studies where CHLLs reported

that Chinese was a central part of themselves (Comanaru & Noels, 2009).

All participants indicated that they intended to reproduce ‘Chineseness’ in the future

through intergenerational Chinese learning and practice. When asked, “If you were to have

children, would you like to encourage them to learn another language other than English?” all

participants agreed that they would like their children to learn Chinese.

I would love to see them learn Mandarin. Yes, I would love to see them learn

Mandarin…It’s definitely biased of course. (Adam)

I wouldn’t force them. They can decide what they want to know, like where their

grandparents’ heritage was from. But I won’t force them. If they choose (Chinese) and

267

they are interested, by all means I will support them and try to coach them…They can

learn where their dad is from. (Bob)

I would like my kids to learn Chinese…so that we don’t lose it as time goes on. We are

Chinese family and we should speak Chinese. (Crystal)

I need to keep my home language going no matter where I am. So I will pass on to my

children as well. (Dianna)

I would encourage them to learn Chinese…I personally would like them to learn

Chinese because of the personal connections they would have. (En-ning)

Dianna expressed the strongest sense of the maintenance of her CHL and definitely intended

to pass it on to her children. This disposition may be attributed to her upbringing in Taiwan,

where in general Chinese language and culture are more valued than in the places where the

other interview participants were brought up. This conforms to Bourdieu’s contention that

‘field’ generates ‘habitus’. Adam’s “biased” disposition demonstrated his eagerness to claim

his ‘Chineseness’ and transfer it to his children through CHL learning because their current

lived world, Australia, does not ban their CHL and ‘Chineseness’. Bob, Crystal, and En-ning

also demonstrated a certain degree of desire to maintain the ‘Chineseness’ in their next

generation through CHL learning because they have all enjoyed the benefits of CHL to some

extent.

In brief, the habitus of ‘Chineseness’ was explicitly linked to past roots, present

moments, and future anticipations. This habitus of ‘Chineseness’, as a system of internalised

268

cognitive and motivating structures (Bourdieu, 1977b, 1990), is produced by structures of

past and present cultural and social environment and will be reproduced in future CHL

learning through its generativity. When CHL is legitimised and imbued with symbolic value,

the loyalty to CHL may persist over many generations of Chinese Australians who relearn

CHL that symbolises their ethnic identity, and in Bourdieu’s sense, their habitus. This CHL

learning is a strategy to reproduce the habitus of ‘Chineseness’ in the next generation and

even generations further removed. A habitus of ‘Chineseness’, to borrow a Bourdieusian

metaphor, contains the genetic information which both allows and disposes successive

generations to reproduce the world that they inherit from their previous generation (Bourdieu

& Passeron, 1990). In this respect, this durable and transposable ‘Chineseness’ is transmitted

across generations and over historical time through CHL learning.

6.3.6 Capital in different fields

As argued in Chapter Three, capital only accrues value in a particular field. The interview

data illustrated this well. Adam has lived in three different countries, where Chinese language

did not carry the same value. He was born in Jakarta, Indonesia. When he grew up there,

learning Chinese was almost impossible. He recalled and revealed this experience as follows:

During the Suharto era they had a bit of a policy where nobody could learn any other

language other than English or Indonesian. You can say that the Mandarin language was

banned…Anything with Chinese writing wasn’t allowed. Interestingly, if you go to

Indonesia around that period of time, filling the customs card, like what we fill when

entering into Australia, there was actually a part where it says whether you have

anything that has Chinese writing or something like that.

269

During the Suharto era in Indonesia, English and Indonesian were positioned at the top of the

language hierarchy and therefore were considered legitimate and valuable linguistic capital

within this particular field. Commitment to learning these two languages was an astute

investment, given the return on which language learners could have access to resources, such

as better education, more job opportunities, and a better life. In contrast, Chinese language

was illegal and would have even accrued a penalty. Investment in Chinese would not have

any positive public returns. This was “one of the most challenging aspects to learn Mandarin”

for Adam. The situation was very different when Adam lived in another field, Singapore,

where Chinese was widely used. When asked about his Chinese practices in Singapore,

Adam recalled:

What happens is that they have more textbooks that teach little kids Mandarin. I just

bought them and I actually went through them myself from the first grade to the second

grade…I learned quite a lot of Mandarin in Singapore because more people can speak

Mandarin. I was just looking at the little kids around to see what words they used. If the

little kids use the words, I can bet ten bucks that it’s going to be one of those few words

that I need to know.

In the second interview, Adam added, “It looks like they (Singaporeans) put a lot of

importance on our ability to speak Chinese.” In his perception, Chinese language seemed to

carry much value in Singapore. This can explain why he said that the reason to do the

exchange in Singapore was to “try to learn a bit more Mandarin along the way”. This one

year in Singapore helped his Chinese, as he added, “When I was in Singapore, I definitely

learned more Mandarin. I stayed there for one year and I took my time learning it… At least

270

the writing and the listening definitely improved when I was in Singapore.” However, Adam

had not pursued any formal Chinese learning in Australia. This may be because the social

fields he lived in did not accord similar value to Chinese. When asked whether many students

chose to learn Chinese and whether Chinese was popular among them, he recalled:

I would say nearly none. Nobody talked about Chinese at all. Maybe because the school

did not offer it, nobody thought about it at all. Even if I asked other people if they were

interested in Chinese, they would say “no, what for?”

When asked about his language choices in Australia, Adam explained his language choices at

home, at work, with his friends, and in his social life. At home, he spoke mainly Indonesian

with his parents who “put more effort into (his) English because it is pretty much the working

language of the world”. He spoke half English and half Indonesian with his sister. At his work,

he believed that his Chinese competence may bring him opportunities in his future career, but

he added, “I wouldn’t say it will be a direct benefit. For example, because I am a researcher,

probably…especially in science, everybody just uses English.” He spoke mostly English with

his friends, who tended to say, “Yeah, Mandarin is important but not that essential.” In his

social life, he did not have many chances to use Chinese because “everybody here can speak

English anyway”. For Adam, Chinese was not a valuable resource in Australia.

According to Adam’s experiences, Chinese accrued different values in Indonesia,

Singapore, and Australia. Likewise, Chinese was valued differently in Hong Kong and

Australia. Because of the colonial history, English was once the only official language in

Hong Kong. Since the ‘handover’, English has kept its value due to this history and its global

utility while Chinese has gained more legitimised value due to the current political situation.

271

Both English and Chinese serve as official languages and linguistic capital in Hong Kong.

They have both been positioned at the top of the language hierarchy. Bob recalled his

experience during his recent Hong Kong trip:

My dad’s sister said, “It’s very important you should learn and you should try to

improve your speaking, so you can communicate well”…They are quite supportive.

When they knew that I was trying to say something, they would try to help me out to

get the point cross.

The words of Bob’s aunty implied that Chinese accrues value in Hong Kong because “it’s

very important”. Unlike Hong Kong, Bob perceives that Australia is not a place where

Chinese is valued:

Well, when I was there (Hong Kong), I thought it (Chinese) was a quite important thing.

But since, you know, I have been here (Australia) for so long, I felt in my younger

years it wasn’t quite a benefit for me because I don’t use that language.

This experience was shared by En-ning. She reported that she “grew up in a very white area”

and indicated she did not have many chances to use Chinese when she was small. When

asked about his language choices in Australia, Bob explained that English was his dominant

language at home, at work, and in his social life. However, for Bob, Chinese did have its

value in other fields. For example, when asked why he chose Chinese in primary school,

apart from the reason that Chinese was “part of” him, Bob replied, “I think a few of my

friends also did that”, suggesting that learning Chinese was a popular choice among his

272

friends. In order to better maintain his membership in this friend network, or this particular

field in Bourdieu’s sense, where Chinese had a degree of legitimate value, Bob pushed

himself to invest in his CHL learning.

Similarly, although English was the dominant language for Crystal, her Chinese

competence was sometimes valued in certain fields. She said, “I have some Chinese friends.

Sometimes I think it is good when I go to their house and I talk to their parents in Chinese,

they kind of like me more.” For Dianna, her Chinese accrued value when she translated for

her parents in given situations. When asked why she considered her Chinese competence a

source of “pride”, she responded, “I guess every time I bring my parents to places like

immigration, post office, and Brisbane City Council, where they need translation, I am the

translator. I feel proud of being able to translate for my parents.” In these situations, Chinese

language became important capital for Dianna.

In contrast to the linguistic fields in Indonesia, Singapore, Hong Kong, and Australia,

the linguistic field in the Chinese Mainland positions Chinese at the top of the language

hierarchy of all the 129 languages in China (Sun, Hu, & Huang, 2007). Within the Chinese

language, Mandarin, the largest group of Chinese dialects in terms of population and

geographical distribution, is considered the most valuable linguistic capital, the possession of

which can determine people’s position in a given field. Different geographical regions have

their own legitimised Mandarin accents. Adam reported an example for this. When he was in

Shanghai, speaking to local people with the wrong tones, the locals said, “Hang on, how

come you can’t speak Mandarin! Where are you from?” Likewise, En-ning recalled a similar

story during her study in Beijing. “I am definitely conscious sometimes like I’ve been to a 食

堂 (dining hall) with my friends and people will be looking at me because they are all talking

Chinese but they can tell I have a very funny accent.” In such a social field, the quantity and

273

quality of the pertinent linguistic capital, that is, a particular accent of Mandarin, can

distinguish the ‘in-group’ from the ‘out-group’. It was the absence of this linguistic capital

that caused Adam and En-ning to feel excluded from a legitimate Chinese cultural citizenship.

This cultural citizenship can be understood as a set of “cultural practices and beliefs” and

“hegemonic forms that establish the criteria of belonging within a national population and

territory” (Ong et al., 1996, p. 738). In this case, the different accent of Adam and En-ning

did not fit the “hegemonic forms” within the Chinese cultural and linguistic fields and

therefore failed to satisfy the criteria of being a legitimate Chinese cultural citizen.

Drawing insights from these data, different political states and social places, such as

family, friendship network, workplace, and school, function as fields in which CHL

competence is deemed necessary or unnecessary, and simultaneously, as fields in which the

value of CHL competence is determined. The more legitimate CHL competence is in a given

field, the more necessary and profitable it is to be competent in it, and the more damaging

and costly it is to be incompetent. For Bourdieu (1989a), the social “sites” are “markets

which, by their positive or negative sanctions, evaluate performance, reinforcing what is

acceptable, discouraging what is not, condemning valueless dispositions to extinction” (p. 85).

In other words, CHL competence never accrues absolute, universal, or guaranteed value.

Instead, its value conforms to and depends on certain logic, interest, and rules of exchange in

a market system. The competence of CHL as linguistic capital realises its exchange value

through its usage in a particular way and in a given field.

6.3.7 Catering to the field of forces

In Confucian social worlds, standards of excellence and successful learning may be

determined both by the individual learners and by “significant others, the family, the group,

274

or the society as a whole” (K. S. Yang, 1986, p. 114). This view was shared by the current

study. What emerged from the data was that participants’ CHL practices were shaped by the

forces within particular fields. These forces established rules of the fields that structured

participants’ agency and shaped their practices in line with these rules. For example, within

the field of family, participants reported the ‘rule’ of parents’ forcing their children to learn

Chinese. Bob recalled:

When I was probably in Grade Three or Grade Four, my parents sent me to a Chinese

school to learn Mandarin. I wasn’t very interested in that. I said, “I am not interested in

this. There is no point sending me to do this.”

En-ning recalled similar experiences:

Basically until I was in Year Nine, I hated learning Chinese with a passion. I think it’s

probably the thing I hated the most in my life. It was definitely my parents forcing me

and my brother to go. It was not my own choice. From my memory, I used to ask to

quit all the time.

Likewise, Dianna shared a friend’s story:

One of my friends actually told me that…he was born here (in Australia) and his mum

forced him to learn Chinese when he was small. He hated it. He didn’t want to go and

he thought it was the hardest language ever.

275

By these accounts, Bob, En-ning, and Dianna’s friend were not willing to learn Chinese when

they were small. Instead, they were commanded and pressured by their parents to do so. This

resonates with previous studies (Lao, 2004; Weger-Guntharp, 2006; J. Zhang, 2009): CHLLs,

when small, had to obey the parental power, which functioned as the rule in the field of

family.

Apart from parents’ forcing children to learn Chinese, En-ning reported a rule within

the field of the Chinese language classroom. She recalled:

I think there are often different expectations and standards to which Chinese

Australians are held. I’ve felt this throughout my experience of learning Chinese and

just this week, I had one of my Chinese teachers tell me that she holds me to a higher

standard, expects more of me, etc…I personally think that these different standards are

what drive me to keep learning. That my teachers often expect more of me gives me a

lot of motivation.

Teacher’s expectations and standards were reportedly imposed on En-ning, because of the

Confucian rule that teacher-student relations are strongly hierarchical (Biggs & Watkins,

1996). Playing by these rules within the field of classroom, willingly or unwillingly, En-ning

tended to put more effort into learning Chinese.

In addition, the stereotypical perception that looking Chinese meant being able to speak

Chinese was reportedly another driving force behind participants’ commitment to learning

Chinese. By their interview accounts, Adam, En-ning, and Dianna’s friend were all

confronted with the question: “Why can’t you speak Chinese while you look Chinese?” This

rule produced embarrassment, discomfort, and contradictions within themselves. As such,

276

they were in a disadvantaged place within social fields where this stereotype functioned as a

positioning rule. These social fields then became a site of struggle for them and CHL

competence became more valuable capital for them in order to play within the rules of these

fields.

6.4 Are the findings consistent?

The previous section reported the main themes that emerged from the interview data. CHL

reportedly had rich and varied meanings for all participants: CHL had the potential to

produce economic capital; participants gained a wider range of cultural, social, and symbolic

capital, and reinforced or recaptured their habitus of ‘Chineseness’ through CHL learning; at

the same time, these different forms of capital accrued differential value in different fields;

and participants’ CHL learning was shaped by the structures of these fields. In addition, the

interview data supported the findings from the initial quantitative phase, which concluded

that participants’ capture of habitus of ‘Chineseness’ and their investment of various forms of

capital in CHL practices contribute to their CHL proficiency. In Bourdieu’s sense, the

production of language competence demanded capital and habitus. In the subsequent

qualitative phase, learning CHL was reportedly both conscious and unconscious practice for

the interview participants. The production of CHL competence through participants’

intentional and unintentional learning will be discussed in this section.

6.4.1 Capital and CHL proficiency

It will be recalled that the initial quantitative phase justified the positive contribution of

various forms of capital to CHL proficiency. In the subsequent qualitative phase, the

277

interview data indicated that CHL competence was reportedly the return on participants’

investment of economic, cultural, social, and symbolic capital.

Firstly, the investment of economic capital helped improve participants’ CHL

proficiency. Adam bought some Chinese textbooks in Singapore and went through them by

himself. The money spent on textbooks seemed worthwhile. This helped him learn some

simplified Chinese characters because the characters he had learnt were the traditional ones.

He recalled his first response to reading these textbooks: “At first it’s a bit odd. I thought

hang on, what is it? Oh, hang on, it’s the simplified character.” Besides the money spent on

Chinese textbooks, he paid for the air tickets for the Shanghai study tour. He said, “The only

thing (I had to pay) was the air ticket” because the local expenses of the study tour were

covered by the program. The money spent on this trip seemed worthwhile too. As he recalled,

“The one that really, really made me learn Mandarin was that 2010 cultural tour…It was such

a great experience…2010, as you can see, was a big eye open-up.” Like Adam, Crystal also

spent some money learning Chinese. She bought Chinese TV shows and watched them to

improve her Chinese. When the interview was conducted, En-ning was doing a fee-paying

program at a leading Chinese University in Beijing. When asked, “Do you think it’s worth it”,

she replied, “Yes, it’s worth it.” Given the intensive training, En-ning expected to improve

her Chinese proficiency. As she confessed, “I am very comfortable saying that 95% of the

Chinese I have learnt has been in China.” For Adam, Crystal, and En-ning, the money spent

on Chinese learning can be understood as an investment of economic capital eventually

convertible into their CHL proficiency.

Secondly, the investment of cultural capital reportedly contributed to participants’ CHL

learning. When asked, “Are you still trying to improve your Chinese language”, Crystal

replied, “Yes, that’s part of the reason why I watch Chinese TV shows.” As stated earlier,

278

Crystal either bought or downloaded Chinese TV shows. These Chinese TV shows, or

cultural goods, demanded time and Chinese cultural knowledge to consume. By watching

these TV shows or consuming these cultural goods, she improved her CHL proficiency.

Similarly, Dianna and her family sometimes read Chinese books together at home. When

asked about her informal Chinese learning, she replied, “Sometimes at home, we have kind of

family discussion. We read a book together in Chinese and we discuss in Chinese.” She

considered this family reading and discussion as informal Chinese learning that may improve

her Chinese language. For Crystal and Dianna, the consumption of these Chinese cultural

goods can be understood as the investment of cultural capital, eventually convertible into

CHL proficiency.

In addition, all the participants interviewed indicated that their parents and family

members were very supportive of their CHL learning and encouraged them to learn Chinese

when they were small. These family commitments were valuable resources in these

participants’ CHL practice and could be understood as social capital that contributed to their

CHL proficiency.

Resonating with the findings from the quantitative phase, symbolic capital was also

beneficial for these participants in their Chinese practice. When asked about the driving force

behind his efforts in learning Chinese, Adam said:

The driving force, in a way, may be “jealousy”. I don’t know whether that’s the right

word to say. For example, my friends who are Malaysians or Singaporeans can speak in

Mandarin or in any other dialects, like Cantonese. I thought, “Hang on, if they can do it,

surely I can do it, right?”

279

This exemplifies a ‘face’ issue, as explained in Chapter Three. Adam attached symbolic value

to his ‘face’. This might be the reason why he felt jealous of his friends who could speak

Chinese. In this situation, it was a matter of ‘face’ that drove him to learn CHL.

In Bourdieu’s sense, the desire for symbolic capital was contributing to his CHL learning.

The initial quantitative phase and the subsequent qualitative phase revealed a two-way

relationship between capital and CHL proficiency. This conforms to the empirical findings in

the existing literature. As explained in Chapter Two, empirical studies have not explicitly

used the concept of capital to investigate CHL learning, but Bourdieu’s capital metaphor

would shed additional light on these empirical findings. The opportunities for CHLLs to be

exposed to CHL are determined by the quantity and quality of various resources available to

CHLLs (Lao, 2004; Luo & Wiseman, 2000; J. Zhang, 2009). The more opportunities to use

CHL in a variety of economic, cultural, and social contexts, the more vital CHL is. These

opportunities will shape present and future actions with regard to CHL learning. As a return

on this CHL learning, CHLLs may acquire wider access to more resources, e.g. more

employment opportunities and better career development (Francis et al., 2009; A. Hancock,

2006; Lao, 2004; Weger-Guntharp, 2006; Wong & Xiao, 2010); more exam credentials or

knowledge of China and Chinese culture (Francis et al., 2009; J. Zhang, 2009); better

communication with Chinese family members (Francis et al., 2009; Lao, 2004) and Chinese

friends (J. Zhang, 2009); deeper involvement in Chinese communities (Lao, 2004); as well as

the enhancement of reputation and the accumulation of awards (J. Zhang, 2009). In short,

CHLLs will capitalise on their CHL by investing various forms of capital. As a return on their

investment, their CHL, when valued and recognised as linguistic capital, can be converted

back into various forms of capital. It is through the two-way interaction between CHL

learning and capital (re)production that CHL proficiency realises its exchange value.

280

6.4.2 ‘Chineseness’ and CHL proficiency

Another resonance with the findings from the quantitative phase was the role played by

‘Chineseness’ qua habitus in CHL practice. As embodied dispositions, habitus is rooted in the

agents’ past. It underpins agents’ present perceptions and actions, and produces agents’

practices. Language choice as a form of practice can be quite unconscious sometimes. When

asked about the reasons for learning Chinese, Adam said, “I tried to learn Mandarin because I

just wanted to learn it”. He added, “I don’t know how it happened but it just happened. It’s a

good weird thing that had happened.” When asked the same question, Crystal said, “I like it

because I want to.” She paused then added, “I feel I just want to know it.” En-ning said, “I

don’t know what came over me but I decided I wanted to learn it in school.” For Adam,

Crystal, and En-ning, they did not have to know what drove them to learn CHL consciously.

Their body knew. When Bob was asked why he chose to learn Chinese in school, he said it

was “because of my Chinese heritage”. By this account, it was his Chinese heritage rooted in

his body that generated his intention for learning Chinese. As Bourdieu (1977b) argues,

agents’ actions are the product of habitus of which the agents may have no conscious mastery

because habitus always exceeds conscious intentions. As such, the habitus of ‘Chineseness’

produced these participants’ CHL practices, without either explicit reason or signifying intent,

to be none the less sensible and reasonable, and to be immediately intelligible and foreseeable,

and hence taken for granted. This ‘Chineseness’ was the immanent law laid down in these

participants by their earliest upbringing and “previous state” (Bourdieu, 2000, p. 161), which

were the preconditions for their CHL practices.

Being a constitutive dimension of ‘Chineseness’, looking Chinese is one of the bodily

attributes that belong to all participants interviewed. Chinese appearance of yellow skin and

black eyes and hair can be understood as contributing to the durable and transposable habitus

281

of ‘Chineseness’ that the participants have to carry with them all through their lives, willingly

or unwillingly. This visibility of looking Chinese is a biological fact for these participants, an

element of habitus that cannot be erased or made over (Luke, 2009). This biological essence

of ‘Chineseness’, to which the participants were involuntarily and inextricably attached,

connected them through their blood and flesh to their Chinese ethnicity, and linked them to

their CHL learning. As Dianna said, “I was born into this culture, this colour, and this

language.” Similarly, Crystal said, “We are Chinese family and we should speak Chinese.”

Dianna had an experience deeply inscribed in her memory that her friends in high school told

her: “If you are Asian, you should be able to speak your home language. If you can’t, it’s sort

of shame.” This point was shared by En-ning. She said, “How difficult it is for a lot of

overseas Chinese and how shameful it is! They look Chinese but they don’t speak Chinese.”

Adam explained his similar experience:

As far as I know, all my grandparents are Chinese. Of course, I will look like Chinese

anyway. If I try to speak Mandarin, they don’t say, “Wow, you can speak Mandarin.”

They tend to go, “How come you can’t speak Mandarin?” …So I guess people around

me tend to say, “Oh, if you look like a Chinese, you can speak Mandarin.”

He also recalled his experience when he lived in Singapore:

Even in Singapore, the taxi drivers sometimes looked at me and said …what are the

words…I tried to copy what they said in Mandarin…“你是华人么 (Are you Chinese)?”

If I said yes, they said “你说你是华人，为什么你不可以讲华语 (You said you are

Chinese, why you can’t speak Chinese)?”

282

For Adam, these experiences constructed the assumption that looking Chinese meant being

able to speak Chinese. It was this stereotypical perception deeply rooted in his mind that

largely drove his conscious commitment to Chinese learning. He tried to learn Mandarin in

Year 12 but the school did not offer it. He was very keen to learn Chinese so he decided to

learn Japanese Kanji in school, which was “the closest thing” to Chinese. As he explained,

“They have the Kanji which is supposed to be similar to 汉字 (Chinese character). So I

thought that’s ok. I will just do it.” He reported that Japanese Kanji became the first Chinese

characters that he was exposed to. In daily life, he tended to “train” and “force” himself to

read Chinese news and watch Chinese TV, and tried to “expose” himself “as much as possible”

to Chinese. For Adam, it might not be the return on the investment of various forms of capital

but his embodied dispositions of ‘Chineseness’ that drove him to pursue CHL. This was

reflected in the statistical models in the initial quantitative phase whereby his CHL

proficiency was predicted by his habitus of ‘Chineseness’ instead of various forms of capital

that he invested in CHL practice. En-ning also considered this habitus of ‘Chineseness’ an

internal driving force behind learning CHL. She reported, “The difference is between a white

person learning Chinese and the overseas Chinese learning Chinese.” When asked to explain

this, En-ning added, “For Chinese Australians learning Chinese, learning the language

immediately raises issues of identity, belonging, culture, and history. I think this means they

are often highly motivated because of the numerous factors driving them to learn.” For

En-ning, ‘Chineseness’ seemed to be associated with Chinese ethnic identity, a sense of

belonging to Chinese cultural history and heritage. She considered these dimensions within

‘Chineseness’ to be the factors driving Chinese Australians to learn CHL.

In brief, these participants’ ‘Chineseness’ as their habitus operated at a level that was

simultaneously conscious and unconscious. They did make language choices in strategic

283

ways, and try to use the rules of different fields to their advantage, but at the same time they

were influenced, or almost driven, by values and expectations that they derived from their

habitus. Though they might be conscious of learning Chinese as their HL strategically, they

might not be aware that their motives, goals, and aspirations were generated through their

habitus of ‘Chineseness’.

6.4.3 Quantitative coda

In this thesis, the findings from the initial quantitative phase and the subsequent qualitative

phase have supported the claim of the mutually constitutive effects between ‘Chineseness’

and CHL proficiency, and between ‘capital’ and CHL proficiency. These findings conformed

to Bourdieu’s theorisation that habitus and capital make sense of people’s language choices

and practices in particular fields. These findings also prompted the question of whether

‘Chineseness’ and ‘capital’ were related to other social practices of the participants, such as

their language usage at home, years of formal CHL learning, frequency of visits to China, and

self-perceptions of who they were. These issues emerging from both the quantitative phase

and the qualitative phase are discussed below.

Spearman’s Correlation Test and Kendall’s Correlation Test were conducted. ‘Years of

formal CHL learning’ and ‘frequency of visits to China’ were variables measured at interval

level. Correlation between ‘Chineseness’ and these two variables were examined by

Spearman’s Correlation Test. Language usage at home and self-labelling were variables

measured at ordinal level, ranked in three levels and five levels respectively. Consequently,

there were many scores with the same ranking. As such, correlation between ‘Chineseness’

and these two variables were examined by Kendall’s Correlation Test. Results demonstrated

that ‘Chineseness’ had a statistically significant small to medium positive correlation with

284

language usage at home (τ = .44, p < .001), years of formal CHL learning (r = .43, p < .001),

frequency of visits to China (r = .48, p < .001), and self-labelling (τ = .43, p < .001). Likewise,

the relationship between different forms of capital and these social practices was also

examined. The results demonstrated that different forms of capital had a statistically

significant small to medium positive correlation with these social practices. The correlation

matrix is reported in Table 6.3.

Table 6.3

Correlation between forms of capital and social practices

Language usage

at home

Years of formal

CHL learning

Frequency of

visits to China Self-labelling

Economic capital τ = .31, p < .001 r = .44, p < .001 r = .51, p < .001 τ = .30, p < .001

Cultural capital τ = .39, p < .001 r = .46, p < .001 r = .47, p < .001 τ = .39, p < .001

Social capital τ = .43, p < .001 r = .46, p < .001 r = .48, p < .001 τ = .43, p < .001

Thus, the stronger participants’ ‘Chineseness’ was and the more value of capital participants

accrued, the more likely they were to use Chinese at home, to spend more years on formally

learning Chinese, to visit China more frequently, and to consider themselves more Chinese

than Australian. Likewise, when participants were more engaged in these social practices,

their habitus of ‘Chineseness’ was more likely to be reinforced or recaptured, and more

capital was to be produced or reproduced. In brief, there is a mutual relationship between

‘Chineseness’/capital and social practices associated with CHL learning.

It is important to highlight the correlation between ‘Chineseness’ and the frequency of

visits to China. In this respect, participants’ visits to China were linked to their ‘Chineseness’.

On the one hand, these physical returns, to some degree, were generated by their

‘Chineseness’. On the other hand, the physical returns have become a crucial process in

building and shaping their ‘Chineseness’. As such, China was not a place simply left behind,

but a concurrent and ongoing social space of cultural and historical attachment. The

285

qualitative data also resonated with this finding. During his Hong Kong trip, Bob tended to be

interested in his family roots. He said:

So maybe it’s just some specific parts that I found quite interesting to me. I found my

great grandmother was from a people called Hakka in Hong Kong, who were foreigners

but came to Hong Kong back in the early days. Maybe if I looked into that, it’s quite

interesting.

The Chinese characters for Hakka are 客家, which literally mean ‘guest families’. As the

name indicates, the Hakka people have a tradition of migration, and hence were considered

‘guests’ in the places where they have moved to. Hakka communities were thought to have

arrived in Hong Kong in the late 17th

century (Hase, 1995). Since they were not the original

residents of Hong Kong, Bob used the word “foreigners” to describe them. Although Bob

claimed that the two-week Hong Kong trip was not as exciting as he had expected, he thought

the discovery of his great grandmother’s origin was “quite interesting”. It will be recalled that

En-ning also reported her “incredible” trip to her grandparents’ hometown in China. These

data indicated that Bob and En-ning considered the discovery of the roots of their cultural

history meaningful for their present sense of self and belonging. As indicated in previous

research (Carreira, 2004), cultural heritage and historical background nurture and reinforce

ethnic identity.

It will be recalled that all the interviewees reported that their parents insisted that they

learn CHL when they were small and likewise, they themselves would encourage their next

generation to learn CHL as a way of maintaining the habitus of ‘Chineseness’. By this

inculcation, the habitus of ‘Chineseness’ becomes durable and transposable from generation

286

to generation. However, habitus is not immutable. This led to the investigation of

generational difference in habitus. Thus, the relationship between respondents’ ‘Chineseness’

and their generation was examined.

Two groups were firstly identified, namely 121 Australian-born participants (first

generation) and 109 non-Australian-born participants (later generation). Before comparing

these two groups, the statistical assumptions were checked. Normality was first checked by

Kolmogorov-Smirnov Test. The result demonstrated a non-significant (p = .200)

Kolmogorov-Smirnov score of .06 for both groups. The assumption of normality was

satisfied. The Levene’s Test for Equality of Variances demonstrated a non-significant (p

= .119) F value of 2.45, which satisfied the assumption of homogeneity of variance.

Satisfaction of these assumptions suggested the basis for t-test. Levels of ‘Chineseness’ of the

non-Australian-born group were shown to be on average higher than those of the

Australian-born group (t = 5.24, p < .001). This result presented a medium-sized effect (r

= .33). This genuine effect existing in relation to different birthplaces of Chinese Australians,

or social space in Bourdieu’s term, will be discussed in the final chapter.

Secondly, the Australian-born participants were further divided into second-, third-,

and fourth-generation and further removed. When generation was treated as a variable

measured at ordinal level, ranked in four levels, Kendall’s Correlation Test indicated a

statistically significant small to medium negative correlation between ‘Chineseness’ and

generation (τ = -.38, p < .001). In other words, ‘Chineseness’ of later generations was weaker

than that of earlier generations. These findings indicated that participants’ habitus of

‘Chineseness’ was not immutable. The habitus of ‘Chineseness’ associated with different

generations of Chinese Australians was shaped in relation to the passage of time. This will be

discussed in Chapter Seven.

287

6.5 Chapter summary

This chapter rendered a comprehensive analysis and interpretation of the data emerging from

the interviews. Findings relating to the meanings that participants attached to CHL were

interpreted through the theoretical framework. By their accounts, CHL competence can

produce new capital for all participants as a return on their investment. Firstly, all participants

indicated that CHL competence could create more and better job opportunities in future

labour markets. As such, their CHL competence has the potential to produce economic capital.

Secondly, participants’ CHL competence contributed to their capture of different forms of

cultural capital. Objectified cultural capital was produced through their CHL learning,

existing in the forms of purchased cultural goods, such as Chinese textbooks and books, as

well as Chinese TV shows. Embodied cultural capital, such as gaining Chinese cultural

knowledge, being able to sing Chinese songs, and being adapted to Chinese ways of doing

and thinking, was captured through participants’ CHL learning. Institutionalised cultural

capital, e.g. educational credits, was also produced through participants’ CHL learning.

Thirdly, CHL competence helped the participants build social capital in Chinese-speaking

worlds, including better communication with family members, more contacts within

friendship networks, and more engagement in social lives. Fourthly, CHL competence

benefited the participants in terms of access to recognised awards, honour, and pride, all of

which can be understood as forms of symbolic capital. In addition to the production of new

capital, CHL competence enabled the participants to claim their ‘Chineseness’. As such, CHL

competence was associated with participants’ habitus. In conclusion, CHL practice offered

meaningful benefits for these participants in the production or reproduction of various forms

of capital and the reinforcing or recapture of their habitus of ‘Chineseness’. Table 6.4

summarises the main findings from the interview data.

288

Tab

le 6.4

Interv

iew d

ata

sum

mary

Field

Hab

itus

Sym

bolic

capital

Social

capital

Cu

ltural

capital

Eco

nom

ic

capital

Con

cepts

Caterin

g to

the field

of fo

rces

Differen

t valu

es in d

ifferent situ

ations

Co

nscio

us actio

ns

Unco

nscio

us d

ispositio

ns

Em

bodied

disp

ositio

ns

Fam

ily ro

ots

Aw

ards

Reco

gnitio

n b

y p

eople

Socialisatio

n w

ith p

eople

Co

mm

unicatio

n w

ith frien

ds

Co

mm

unicatio

n w

ith fam

ily m

embers

Form

al edu

cation cred

its

Ch

inese w

ay o

f doin

g th

ings

Und

erstandin

g o

f Chin

ese cultu

re

Ch

inese cu

ltural g

oods

Job

oppo

rtunities

Ben

efits / Mea

nin

gs o

f CH

L

Th

e forces o

f fields th

at structu

re agen

cy an

d sh

ape ag

ents’

practices in

line w

ith th

e rules o

f fields

A so

cial place co

ntain

ing ag

ents stru

gglin

g fo

r po

sitions b

y

the co

ntro

l of reso

urces

Imm

anen

t pro

pen

sities linked

to cu

ltural h

istory

that d

rive

the actio

ns b

oth

inten

tionally

and u

nin

tentio

nally

Rep

utatio

ns fo

r com

peten

ce and

imag

es of resp

ectability

and

honourab

ility

Co

ntacts

and

gro

up

mem

bersh

ips

that

pro

vid

e actu

al or

po

tential su

pport an

d access to

valu

ed reso

urces

Institu

tionalised

cultu

ral capital: C

ultu

ral com

peten

ce with

a

conven

tional,

constan

t, leg

ally

guaran

teed

valu

e in

connectio

n to

certain in

stitutio

ns

Em

bodied

cultu

ral capital: L

asting

disp

ositio

ns o

f min

d an

d

body, ex

isting in

the fo

rm o

f schem

ata of p

erceptio

ns an

d

action

s

Ob

jectified cu

ltural cap

ital: Material fo

rms o

f cultu

ral goods

Co

nvertib

le into

mon

ey

Cod

ing criteria

√

√

√

√

√ √

√ √

√

√

√

Ad

am

√

√

√ √

√ √

√ √

√ √ √

Bo

b

√

√

√

√ √

√

√

√

√

√

√

√

√

Cry

stal

√

√

√ √ √

√

√

√

√ √

√

Dia

nn

a

√

√

√

√

√

√

√

√

√

√

√

√ √

En

-nin

g

289

Furthermore, the findings from the interview data conformed to the findings from the

initial quantitative phase. Data of the initial online questionnaire have revealed that both

participants’ habitus of ‘Chineseness’ and their invested capital can contribute to their CHL

proficiency. Similar findings emerged from the subsequent interview data. As summarised in

Table 6.5, money spent on CHL learning, parents’ and friends’ supports in CHL learning,

consumption of Chinese cultural goods in the process of CHL learning, face issues in CHL

learning, and conscious and unconscious dispositions of ‘Chineseness’ all contributed to CHL

proficiency. In Bourdieu’s sense, habitus and various forms of capital made sense of the

choice of CHL in given fields.

Table 6.5

Contributing factors to CHL proficiency

Concepts Contributing factors Adam Bob Crystal Dianna En-ning

Economic

capital

Money spent on CHL learning,

e.g. textbooks, tuition, visit to

China, and Chinese TV shows

√ √ √

Cultural

capital

Consumption of Chinese cultural

goods through CHL learning, e.g.

watching Chinese TV shows and

reading Chinese books

√ √

Social

capital

Family and friends’ supports in

CHL learning √ √ √ √ √

Symbolic

capital Recognition by people √

Habitus

Embodied dispositions √ √ √ √ √

Unconscious dispositions √ √ √

Conscious actions √ √ √ √ √

In addition, the findings led to new questions: Were participants’ ‘Chineseness’ and

‘capital’ related to a wider range of their practices? Was there a generational difference in

participants’ ‘Chineseness’? Further examination of the quantitative data offered answers to

these questions. This quantitative coda revealed that participants’ habitus of ‘Chineseness’

and their ‘capital’ were correlated to a wider range of their practices, such as language usage

at home, years of formal CHL learning, frequency of visits to China, and their

290

self-perceptions of who they were. The quantitative coda also discovered the generational

difference in participants’ ‘Chineseness’, with later generations having weaker levels of

‘Chineseness’.

In summary, participants’ CHL competence produced through various social practices

associated with CHL learning at given times across certain places was recognised as

linguistic capital with different value in different fields. Production or reproduction of various

forms of capital and reinforcement or recapture of the habitus of ‘Chineseness’ happened in a

process of exchange, an exchange of value and forms of resources. Participants’ various

forms of capital and their habitus of ‘Chineseness’ were convertible into their CHL

competence and reconvertible into themselves in identical or expanded forms within the field

of exchange. It was through this process that participants’ CHL competence, as linguistic

capital, realised its exchange value.

291

Chapter 7: Discussion and Conclusion

This chapter offers the discussion and conclusion of the research reported in this thesis.

Section 7.1 discusses the findings from the quantitative phase and the qualitative phase;

explains these findings by virtue of Bourdieu’s theoretical triad of habitus, capital, and field;

and talks back to the research questions. This section also discusses several important issues

emerging from the findings. These issues include: how habitus and capital related to other

social practices besides CHL learning; how generation impacted on ‘Chineseness’ across time

and space; how bodily emotion was generated through power relations; and how symbolic

power and symbolic violence were imposed within given fields. Section 7.2 highlights the

theoretical, methodological, and practical contributions of this study. Section 7.3 outlines the

limitations of the study. Section 7.4 suggests directions for future research. The last section

concludes the thesis with a succinct summary. It recalls the research journey over the life of

the thesis, talks back to the existing literature, and revisits the researcher’s subjectivity.

7.1 Discussions

Bourdieu’s (1989a, p. 101) equation “[(habitus) (capital)] + field = practice” summarises the

conceptually and empirically essential relationship amongst capital, habitus, and field, which

illuminates individuals’ practices. Specifically, Bourdieu (1991) theorises that people make

language choices as a form of social practice according to the amount of different resources

(capital) as well as the dispositions (habitus) they have within a given field. This theorisation

is the underpinning basis for the theoretical framework and the research design of this study.

Guided by Bourdieu’s sociology, the study investigated the entanglement amongst the habitus

292

of ‘Chineseness’ of young Chinese Australian adults, this group’s investment of capital in

CHL learning, and their CHL proficiency. Several important issues emerged from this

investigation. These issues are discussed below.

7.1.1 Habitus, capital, and practice in fields

In Chapter Five, SEM tested and explained that both ‘Chineseness’ and ‘capital’ made

statistically significant contributions to ‘CHL proficiency’. In the multiple regression model,

multicollinearity between ‘Chineseness’ and ‘capital’ was diagnosed from a statistical

perspective. This problem can be further illuminated from a theoretical perspective. As an

embodied property and “incorporated and quasi-postural disposition” (Bourdieu, 1985a, p.

13), habitus serves as a form of “capital” (Bourdieu, 1985a, p. 13), specifically a cultural

capital (Bourdieu, 1973). Accordingly, the habitus of “‘Chineseness’ becomes a form of

cultural capital” in particular contexts (A. Louie, 2004, p. 21; Ong, 1999, p. 6). In order to be

a form of capital, one’s habitus of ‘Chineseness’ must be valued and recognised within a

particular field. ‘Chineseness’ as a set of embodied dispositions associated with certain

Confucian values can be understood as cultural capital when these embodied dispositions of

‘Chineseness’ are accredited by the legitimate Confucian cultural tastes and inclinations.

Specifically, ‘Chineseness’, when valued, can be understood as embodied cultural capital

because it has been integrated within the lasting dispositions of mind and body, existing as

perceptions and behaviours of the participants. In this respect, ‘Chineseness’ was embedded

in the dimensions of cultural capital. Young Chinese Australian adults may develop their

‘Chineseness’ in relation to how much embodied cultural capital they have in a Chinese

cultural or social field. This relationship is very important because it reveals how the amount

of capital that they have can condition their dispositions of being, doing, and thinking.

293

When taken account of together, participants’ habitus of ‘Chineseness’ and their

various forms of capital became the overarching drivers behind the CHL practices that

produced CHL proficiency. Moreover, participants’ ‘Chineseness’ and capital produced other

social practices, such as language usage at home, years of formal CHL learning, frequency of

visits to China, and self-labelling, which in return reinforced ‘Chineseness’ and generated

various forms of capital. Data analysis in the quantitative coda demonstrated that language

usage at home, years of formal CHL learning, frequency of visits to China, and self-labelling

had a statistically significant small to medium positive correlation with ‘Chineseness’ and

capital. However, this ‘Chineseness’ and capital did not generate the same practices for every

Chinese Australian. Data analysis in the quantitative coda demonstrated a wide range of

variance in language usage at home (from English only to Chinese only, with different levels

of mixed usage in between); years of formal CHL learning (from none to over 15 years);

frequency of visiting China (from none to over 20 times); and self-labelling (from

considering themselves basically Australian to basically Chinese, with degrees of a blend of

both in between). As such, depending upon the stimuli and structure of the field, habitus and

capital will generate, but will not determine, outcomes (Bourdieu & Wacquant, 1992).

Interestingly, the social practices had the strongest correlation with the form of capital

to which they had most relevance. Language usage at home had the strongest correlation with

social capital in the Chinese community (r = .43, p < .001). When family connection became

a valued social capital, participants tended to use more CHL to maintain and strengthen their

family ties. Self-labelling also had the strongest correlation with social capital in the Chinese

community (r = .43, p < .001). When involvement in the Chinese community accrued more

value, participants tended to have stronger feelings of belonging to this community.

Consequently, they considered themselves more Chinese. Frequency of visits to China had

294

the strongest correlation with economic capital (r = .51, p < .001) because international travel

cost participants money. It also had a small to medium correlation with cultural capital (r

= .47, p < .001) and social capital (r = .48, p < .001) because participants were culturally and

socially connected to China, a place of origin, roots, and heritage. In brief, there is a two-way

relationship between ‘capital’ and participants’ various social practices associated with their

CHL learning. Participants invested various forms of capital in their CHL learning through

these social practices. As a return on their investment, their CHL proficiency improved.

When their CHL proficiency was recognised and legitimised as linguistic capital in certain

fields, this linguistic capital realised its exchange value, accrued added value, and produced

various forms of capital.

Of all the correlations discussed above, it is particularly important to highlight the

relationship between ‘Chineseness’ and visits to China, which fostered and maintained

participants’ ties to their original roots and played a crucial and ongoing role in informing

their notions of ‘Chineseness’. This finding echoed A. Louie’s (2004) study in which her

Chinese American participants reportedly “reinforced” (p. 110) their ‘Chineseness’ through

visits to their ancestral villages in China and one participant reported, “As soon as my feet

touched China, I became Chinese.” (p. 116) Literally, homeland can be interpreted as a

concrete place where intimate familial and social relations are established through the lived

experience of locality and community. For Chinese Australians, China is more than a

geographical place away from their home in Australia. It is also a homeland infused with

meanings and relevance that create a sense of ancestral origin, historical roots, and cultural

heritage. Traditionally, migration was a one-off process of movement from homeland to

hostland. At present, it is becoming easier than ever before for Chinese Australians to travel

back and forth between China and Australia. With regard to space, these movements function

295

as a form of deterritorialisation that enables co-existing homes in China and Australia. With

regard to time, such physical mobility allows Chinese Australians to continually move

between their past and their present.

In relation to space, the non-Australian-born group had emigrated from China or from

countries with Chinese population. They had spent their early years in social spaces with

greater access or engagement in Chinese language and culture than the Australian-born group,

suggesting a deeper understanding of and stronger sense of attachment to Chinese culture

than the Australian-born group. As Bourdieu (1991) acknowledged, cultural dispositions may

be acquired by prolonged exposure to a culture, or through the deliberate inculcation of a

culture. In line with this Bourdieusian perspective, ‘Chineseness’ can be understood as a set

of culturally learned dispositions acquired through experiences of daily life and structured

through activities within social fields that featured Chinese culture. The dispositions

associated with the habitus of ‘Chineseness’ of the non-Australian-born group were

inculcated by their Chinese cultural history prior to their immigration and then transposed

across different social spaces, from their birthplace to Australia. This could explain their

stronger expressions of ‘Chineseness’ than the Australian-born group (t = 5.24, p < .001, r

= .33).

In relation to time, the ‘Chineseness’ of later generations was significantly weaker than

that of more recent generations. The family of participants from later generations had longer

settlement periods in Australia than that of participants of more recent generations. Later

generations did not have as many Chinese cultural memories and experiences as the more

recent generations. Their physical and cultural ties to their Chinese roots were less intense.

As argued earlier, habitus is lasting but not immutable. With the passage of time, different

generations made gradual transitions from being migrants to becoming integrated members of

296

the host society as their lives gradually took root in a land away from the cultural home. They

tended to internalise the values and imperatives of the host society. Consequently, a habitus

of ‘Chineseness’ can be expected to gradually fade generation by generation (τ = -.38, p

< .001). This resonates with Luke’s (2009) position that hyphenated and diasporic identity

necessarily leads to imperfect intergenerational reproduction of embodied dispositions.

With the change of life worlds and the passage of time, different generations

experience different culturally and historically situated upbringings, which in turn can shape

a different habitus. Habitus is thus at the basis of strategies that tend to maintain these

differences, hence concurring in different practices (although not consciously or deliberately)

in reproducing the entire system of differences across time and space (Bourdieu, 1996). The

difference in participants’ ‘Chineseness’ thus led to different CHL practices, in terms of

different levels of effort put into learning CHL, formally or informally, willingly or

unwillingly, and consciously or unconsciously. Consequently, the difference in ‘Chineseness’

could be understood to produce the difference in CHL proficiency.

These findings echoed previous studies. First-generation CHLLs were more proficient

in CHL than their counterparts from later generations (Kiang, 2008). There was a significant

association between generation and feelings of being Chinese, as well as between

self-labelling and ethnic exploration (Rosenthal & Feldman, 1992). Over time and across

space, there was erosion of ethnic identification, behaviours, and knowledge but not of the

importance and evaluative components of ethnic dispositions (Rosenthal & Feldman, 1992).

In general, later generations do not maintain ‘Chineseness’ to the same level of

intensity as more recent generations because the habitus of ‘Chineseness’ is not immutable.

However, some sense of ‘Chineseness’ does persist within the Chinese ethnic group because

the intergenerational influences create a sense of maintaining ‘Chineseness’ that makes the

297

habitus of ‘Chineseness’ durable and transposable. In this respect, ‘Chineseness’ describes

the stronger or weaker allegiance of Chinese Australians to their ethnic group, with which

they have ancestral links. Although there is no necessity for continuation over time and

across space of exactly the same socialisation or cultural patterns, it is in a relation of

diversity within homogeneity that the shared habitus of the different members of the same

group are unified (Bourdieu, 1977b).

7.1.2 Field of forces

As argued in Chapter Three, a field is a structured social space, “a field of forces” (Bourdieu,

2011, p. 40). The forces within a field structure the power relations between agents who

dominate and those who are subordinate. The interview participants were conscious of their

subordinate or dominated positions in particular situations. Consequently, their feeling of

‘shame’, generated by experience of subordination, was repeatedly emergent from the

interview data. This is consonant with previous research in which participants considered not

being able to speak Chinese as a “disgrace”, “embarrassment”, and a reason for being

“ashamed” (Francis et al., 2009, p. 529). From a Bourdieusian perspective, this ‘disgrace’,

‘shame’, and ‘embarrassment’ can be understood as forms of bodily emotion, “the practical

recognition through which the dominated, often unwittingly, contribute to their own

domination by tacitly accepting, in advance, the limits imposed on them”, often taking the

forms of shame, timidity, anxiety, guilt, blushing, inarticulacy, clumsiness, and trembling

(Bourdieu, 2000, p. 169). These forms are ways of reluctantly submitting to the dominant

judgement, “sometimes in internal conflict and ‘self-division’, the subterranean complicity

that a body slipping away from the directives of conscious and will maintains with the

violence of the censures inherent in the social structures” (Bourdieu, 2000, pp. 169-170). It

298

will be recalled that in Chapter Six, En-ning reportedly felt ashamed for her “funny accent” in

a situation where Mandarin was the dominant and legitimate language. The lack of quantity

and quality of Mandarin, or linguistic capital in Bourdieu’s term, placed En-ning in a

subordinate position. En-ning, therefore, considered not being able to speak Chinese

“shameful”. Likewise, Adam apologised for his tones during the interview. In this case,

subordination emerged from hierarchies based on mastery of spoken Chinese. Adam gave

evidence of a sense of subordination through a possibly unconscious apology for his lower

Chinese proficiency, compared to the researcher’s native Mandarin proficiency. In these

examples, there were indeed moments where participants demonstrated the feeling of ‘shame’,

the bodily emotion of anticipating an experience of being dominated in a given field. In

contrast to Adam and En-ning’s experiences, Dianna was able to slip away from being

subordinate or dominated when her Chinese proficiency was recognised. In response to this

recognition, she did not feel “any shame at all” for being able to speak Chinese.

In relation to the forces within fields, participants also reported that their parents forced

them to learn CHL when they were small, their Chinese teachers set a higher standard for

them than their white peers, and people around them held the stereotype that looking Chinese

meant being able to speak Chinese. This is consistent with previous research that indicates

people with more power can construct “practical taxonomies of ways of seeing and doing in

the world” (Lane, 2000, p. 195). Parents’ insistence, teachers’ expectations, and people’s

assumptions were norms and conventions that granted parents, teachers, and those people

legitimate power and made them more powerful others to set the rules within particular fields.

These rules were believed, obeyed, and respected by the participants. Consequently, parents,

teachers, and people accrued symbolic capital, the possession of which constructs the

foundation of symbolic power (Bourdieu, 1989b). Parents, teachers, and people could wield

299

their symbolic power to ‘force’ their children to learn CHL, to set a higher standard for

CHLLs, and to expect Chinese-looking people to speak Chinese. Symbolic power is the

power to consecrate and reveal things with words (Bourdieu, 1989b). As such, parents,

teachers, and people may exert their symbolic power by persuasion, inculcation, and

discourse to legitimise CHL as a valuable resource in a particular social field. At the same

time, participants granted parents, teachers, and people the legitimacy and right to impose

their expectations. Bourdieu (1977a, p. 648) explained this power relation, “Those who speak

regard those who listen as worthy to listen and those who listen regard those who speak as

worthy to speak”.

This power relation can be understood as constituted within and by the rules and forces

within fields. As agents, the participants had practices that were structured by these forces of

the fields. Due to their agency, they tended to adjust to the structure of the field and submit to

these forces. Since these forces were imposed on them, their participation was quite often

reluctant. In this respect, these forces can be understood as symbolic violence.

The notion of symbolic violence is particularly relevant to the Chinese cultural and

social context. As discussed in Chapter Three, Confucianism has consecrated and legitimised

the power of parents and teachers by its core values of Zhong (忠, Loyalty), Xiao (孝, Filial

Piety), and Li (礼, Ritual or Propriety). These Confucian values position parents and teachers

at the top of the social hierarchy. In this respect, symbolic violence describes the tacit and

explicit modes of cultural and social domination by parents and teachers occurring within the

social lives of participants. Specifically, it accounts for this imposition of parents’ and

teachers’ will to urge CHL learning upon the less powerful participants, and the submission

to this imposition, willingly or unwillingly, by the participants. The symbolic violence that

imposes CHL learning on CHLLs was also revealed by existing literature: Parents forced

300

their children to learn CHL when their children were small (Weger-Guntharp, 2006; Wong &

Xiao, 2010); Chinese appearance was taken for granted as a sign of being able to speak

Chinese (Ang, 2001); some early generations of American Cantonese (A. Louie, 2004) and

British Cantonese (Watson, 1975) have the common sentiment that one could not be

considered truly Chinese without receiving an education in Chinese; and some Chinese

Canadians learn CHL because of various external “pressures” (Comanaru & Noels, 2009, p.

148).

Empirically, participants might consider themselves free individuals yet sometimes had

to base their everyday practices on the assumptions about dispositions, behaviour, and

attitudes of powerful others. Sociologically, their social practices (agency) could be explained

by social regularities (structure). These both raise fundamental questions that Bourdieu’s

notions of symbolic power and symbolic violence help to resolve and explain the explicit

rules dictating such practices. On the one hand, symbolic power and symbolic violence were

structures within fields that pushed participants to learn CHL. On the other hand, the external

symbolic power and symbolic violence were gradually internalised by the participants’

agency as their own habitus and capital to continue to pursue CHL as a linguistic capital.

7.2 Contributions of the study

There has been a recent surge of research interest in HLLs (Lynch, 2003). However,

compared to other HLLs, a relatively small body of HL research addresses issues particular to

CHLLs (Levesque, 2007; D. Li & Duff, 2008). Considering the size, vitality, and growth of

the CHL learning population, research with CHLLs warrants more attention (He, 2008; D. Li

& Duff, 2008). This mixed methods study, guided by Bourdieu’s theory of habitus, capital,

and field, has offered theoretical, methodological, and practical contributions.

301

7.3.1 Originality and contributions to new knowledge

HL research is an emerging field that has been gaining increasing attention (Lo-Philip, 2010).

The reciprocal and mutually constitutive effects between HL and ethnic identity have been

well documented in the literature (Cho, 2000; Cho et al., 1997; Fishman, 1977; Giles et al.,

1977; Giles & Johnson, 1987; Hurtado & Gurin, 1995; Smolicz, 1981), with studies

providing evidence about Chinese Americans/Canadians, Korean Americans/Canadians,

Japanese Americans/Canadians, Armenian-, Latin-, Asian-, Mexican-, and

Vietnamese-Americans, and Italian-, Portuguese-, Ukrainian-, and Jewish-Canadians.

Positive correlation between learners’ ethnic identity and their HL has been found by a large

number of social psychological studies in North America. However, there is scant

sociological research conducted outside North America on such topics. This study has

contributed theoretically to the literature by conceptualising and operationalising Chinese

Australians’ ‘Chineseness’ as their ethnic identity and constructing the essential link between

their ‘Chineseness’ and their CHL proficiency through Bourdieu’s sociological notion of

habitus.

There is an extensive tradition of studies on commitment to Second Language learning,

understood in psychological term as ‘motivation’ (Gardner, 1968, 1985; Gardner & Lambert,

1959, 1972). In contrast, the body of sociological work investigating commitment to Second

Language learning in general is smaller and more recent. As a sociological alternative to the

psychological concept of ‘motivation’, the concept of ‘investment’ (Norton, 1995) in Second

Language learning has offered a theoretical breakthrough (Canagarajah, 2006; Pavlenko,

2002). ‘Investment’ demands ‘capital’ (Bourdieu, 1986). To capture the complex process of

language learning and to offer an adequate investigation into learners’ trajectories, the notion

of capital has found wide purchase in the language education world. Following this route,

302

another theoretical contribution of this study was the use of Bourdieu’s sociological notion of

capital in a HL context to interpret Chinese Australians’ investment of available resources in

their CHL practice.

Habitus, capital, and field are the three fundamental thinking tools of Bourdieu’s

sociological approach. This study realised the full value of a Bourdieusian perspective by

constructing an inter-dependent and co-constructed triad of habitus, capital, and field to

examine language practice in a HL context. Moreover, this study extended Bourdieu’s

theorisation of language choice within a single language in a monolingual setting to language

choice among different languages in a multilingual context. A comprehensive theoretical

framework that can handle the interdependence of Chinese Australians’ ethnic identity, their

commitment to CHL learning, and their perceptions of CHL proficiency has been proposed.

7.3.2 Methodological contributions

In addition to the above theoretical contributions, the study also offered significant

methodological contributions. Much of the existing work attempting to apply Bourdieu’s

notions of capital, habitus, and field to people’s identities and commitment to language

learning has tended to be qualitative in approach (Connolly, 2011). However, there is

inevitably something missing by restricting the methodological focus to one mode of research

when operating Bourdieu’s theoretical package. In Bourdieu’s original works, such as

Distinction (Bourdieu, 1989a) and The State of Nobility (Bourdieu, 1996), data were also

quantitatively investigated by correspondence analysis. Though these data were mostly

analysed by a descriptive and exploratory multidimensional scaling technique, Bourdieu was

interested in tendency, prediction, and correlation based on these quantitative analysis. He

suggested (1990) that quantitative expressions of the distribution of capital in its different

303

forms among individuals can be developed and thus capital should be measurable. Likewise,

if habitus does represent the internalisation of broader social structures and does represent a

set of dispositions that are manifest, to some extent, in particular ways of thinking and

behaving, these wider patterns should be discernible and ultimately measurable at least to a

certain degree. Bourdieu’s concepts represent a methodological challenge in quantifying, but

given the increasing use of these concepts in sociology, such a methodology is needed

(Cockerham & Hinote, 2009). Following this route, a set of instruments was developed and

validated to quantify Bourdieu’s key concepts of capital and habitus within certain social

fields, and to investigate the research problems in this study.

Due to the technical limitations of the times, Bourdieu was not able to analyse the

quantitative data by the statistical models that have been developed today. The current study

has the benefit of statistical software and new modelling techniques to advance such lines of

enquiry. Moreover, Bourdieu’s key concepts can be investigated in a deeper manner by the

application of mixed methods design. Connolly (2011), informed by the work of Bourdieu,

suggests the need for greater use of quantitative methods in conjunction with in-depth

qualitative methods to further the understanding of the influence of identity in people’s lives.

In this respect, the current study made a methodological contribution to introduce a mixed

methods design that applies a Bourdieusian stance to the current research context.

7.3.3 Practical contributions

The pioneering research on HL dates back more than half a century, when Fishman (1964)

inquired into the field of minority language maintenance and shift. There were few echoes of

Fishman’s (1964) work until very recently when minority peoples’ HL and their ethnic

identities were recognised not only as assets to the individual minority people but also as

304

valuable resources to the social world around them (Brecht & Ingold, 2002; Campbell &

Rosenthal, 2000; He, 2006; Lynch, 2003; Peyton et al., 2001). Existing research argues

convincingly that developing HL speakers’ linguistic and cultural knowledge to advanced

levels is valuable not only for the HLLs themselves, their families, and communities, but also

for society more broadly (Fishman, 1991).

Firstly, understanding Chinese Australians’ heterogeneity and the potential drivers

behind CHL learning helps to better accommodate linguistically diverse learners in language

courses. As noted in Chapter One, there are a number of Confucius Institutes and Confucius

Classrooms established in Australia. There is now at least one Confucius Institute in all state

capitals of Australia. Confucius Classrooms in primary and secondary schools are also

proliferating. The courses offered by these Confucius Institutes/Classrooms are designed to

meet the demands of Foreign Language Learners of Chinese. However, many current,

prospective, and potential Chinese learners in Confucius Institutes/Classrooms are in fact

Chinese Australians who are CHLLs, as distinct from Foreign Language Learners of Chinese.

This highlights the importance of flexible teaching methods and pathways that cater to

Chinese learners’ varied learning backgrounds (Australian Government, 2012). By examining

Chinese Australians’ identities and their commitment to CHL learning, the current research

will be useful to Hanban and Australian institutions in partnership with Hanban, with an

interest in teaching and learning Chinese not only as a Foreign Language but also as a HL in

Australia. The research will help to develop Chinese language teacher training, Chinese

textbook and curriculum design, and Chinese language classroom teaching methodology to fit

the CHL context in Australia.

In addition, this study will appeal to the growing interest in Chinese Australians’

stories. The stories themselves are fascinating and absorbing because Chinese Australians

305

were never a uniform ‘heathen Chinee’ bent on flooding Australia with a low-level peasant

civilisation (Price, 1975, p. ix). Rather, they are a diverse group of people from different

areas, speaking different dialects and pursuing different aims and careers. However,

recognition of local Chinese community and heritage has largely focused on ‘gold-rush’

archaeology and immigration restrictions for much of the post-war era in Australia

(Couchman, Fitzgerald, & Macgregor, 2004). In this new millennium, there is a revitalisation

of interest in the life of Chinese Australians, among community-based amateurs, freelance

historians, and Chinese studies scholars from Australia and overseas. Moreover, many young

Chinese Australians are eager to rediscover the Chinese side of their family or personal

histories (Chan, 2004). This study will lead to a broader and deeper understanding of Chinese

Australians, with particular regard to their identity and commitment to CHL learning.

Lastly, the research is most timely. Upon the completion of the draft of the thesis, the

Australian Government released the Australia in the Asian Century White Paper in October

2012. The White Paper recognised the transformation of Asia into the economic powerhouse

of the world and stressed that Australia should seize the economic opportunities to work with

Asia. To this end, the White Paper urged more investment in Asian language education,

identifying Chinese as one of the priority languages. In recognising the economic, cultural,

and social needs to build a sound knowledge of Chinese language, the White Paper

encouraged Australians to become more Chinese literate. This will be pursued through school

education, university teaching and research, and industrial and community engagement. In

this respect, the White Paper will help shape Australia into a social field where Chinese

language has more legitimate value than at present. Resonating with the White Paper, the

research helps Australia and Australians to understand how Chinese language has become

valuable linguistic capital in contemporary Australia. In particular, the research reflects how

306

Chinese Australians negotiate their habitus of ‘Chineseness’ and capitalise on various

resources through learning Chinese. As a return on this investment, Chinese Australians will

have wider access to various forms of capital and will have a richer understanding of their

identity as being a Chinese in Australia. Consequently, Australia will benefit from this

powerful human resource that can serve the nation’s economic, cultural, social, and political

needs.

7.3 Limitations of the study

Given the practicality in research and the accessibility of resources available to research, this

study was constrained within a manageable scope. Consequently, the study, like all the other

studies, has limitations. These should be acknowledged.

The first limitation was associated with sampling. Snowball sampling was used in the

initial quantitative phase. It was argued in Chapter Four that snowball sampling was the

optimal sampling method for the current study, given the difficulty of identifying and

approaching the population under investigation. However, snowball sampling is not a

randomised method that samples the largest potential range of the demographic

characteristics of the population. This sampling method will probably net more participants

who are similar than those who are different. Consequently, it potentially reduces the

representativeness of the sample and ultimately may reduce the heterogeneity and

generalisability of the findings. These disadvantages existed in the current study.

Approximately 50% of the participants were from Brisbane because the snowball sampling

started in Brisbane. However, if the participants had been randomly sampled, Sydney should

have had the largest number of participants, followed by Melbourne and Brisbane. Another

problem with snowball sampling is that it makes it difficult to report the response rate.

307

Secondly, the pilot study in the quantitative phase was not able to check the validity of

the measurement due to limited sample size in the pilot stage. Although Cronbach’s alpha

suggested good internal consistency reliability of the measurement, this did not guarantee the

validity of the measurement. In this respect, the main study ran the risk of not being able to

achieve a reasonable level of validity, though validity of the measurement was confirmed in

CFA at a later stage. It is suggested that both reliability and validity of the measurement be

checked in the pilot phase when it is practically possible to get a sample size large enough

before major efforts are put into the main study.

Thirdly, the equations offered by SEM to calculate the scale score for the constructs

cannot be generalised to other samples. This is because the proportionally weighted factor

score of each indicator is sample-dependent in SEM. Unlike SEM, Rasch Modelling can

manage this problem. Therefore, Rasch Modelling is recommended in future research.

Next, although the explanatory mixed methods design added scope and depth to this

study, one pragmatic challenge was the longer time needed for data production and analysis

in both quantitative and qualitative ways. Good time and project management was essential to

overcome this challenge. Enhanced confidence in the trustworthiness and credibility of the

research findings, and the expanded depth and scope to the project with the application of the

mixed method design potentially offset the expanded time frame.

Another issue should also be acknowledged. The interviews in this study demanded

nuanced linguistic skill in English. This posed another challenge for the researcher, a native

Chinese speaker. When transcribing, the researcher found a number of issues that could have

been further probed at the interviews. In the interviews themselves, the researcher had to

concentrate on the literal meaning of the participants’ replies. Opportunities to explore the

deeper meanings behind participants’ replies were missed at times. Follow-up interviews

308

were done to explore these deeper meanings, which again extended the time length of the

project.

7.4 Future directions

This study has used Bourdieu’s signature notions of habitus, capital, and field to interpret

CHL practices of young Chinese Australian adults. The findings of this study supported the

claim that the habitus of ‘Chineseness’ and the investment of various forms of capital

produced practices associated with CHL, and in return, this habitus of ‘Chineseness’ was

reinforced or recaptured, and various forms of capital were generated through these CHL

practices. Besides the problems investigated in this study, there are several important and

absorbing issues emerging from this study to be addressed in future research.

Firstly, Chinese Australians never live in a closed system. Apart from habitus and

capital, there are many other sociological factors existing and emerging in social fields where

Chinese Australians interact as agents. These sociological factors, such as social class and

gender, together with habitus and capital, will have an impact on Chinese Australians’ CHL

practices. However, these sociological factors were not explored in this study. These

questions could be addressed in future research.

Secondly, the Chinese diaspora has a long history, dating back to the early 19th

century.

There are large populations of Chinese immigrants and their descendents living in different

social places, urban Australia, rural Australia, and other parts of the world. The participants in

the initial quantitative phase and the subsequent qualitative phase were sampled from urban

areas. Although an overwhelming proportion of Chinese Australians reside in urban areas, it

would be interesting to look at some cases in rural areas with perhaps less access to formal

CHL opportunities. In a global context, Chinese Australians only constitute a small body of

309

overseas Chinese. Are overseas Chinese elsewhere similar to, or different from, Chinese

Australians? It will be interesting to portray CHL practices of Chinese immigrants and their

descendents in a wider diasporic context.

Lastly, Bourdieu’s sociology is profound. His whole package of concepts and theories

will offer a substantial thinking tool when contemplating language dynamics in social,

cultural, and educational contexts. Apart from his signature notions of habitus, capital, and

field, his conceptualisation of ‘symbolic power’ and ‘symbolic violence’ would be

particularly enlightening when investigating the education and language socialisation of

indigenous people in a colonised or post-colonial context, such as the Aboriginal people in

Australia. These concepts would also be helpful when interpreting the ‘Mandarinisation’ of

the ethnic minority people, and Hong Kong and Macau people in contemporary and modern

China. In addition, these concepts could inform an examination of sociological reasons

behind the accent change of internal migrant workers in order to integrate into metropolitan

Chinese cities, such as Beijing, Shanghai, and Guangzhou. In brief, this thesis leads to

investigation of social problems of integration in a migration context, unity within a

diversified context, and the co-existence of homogeneity and heterogeneity in a globalised

context.

7.5 Overall conclusions of the study

This study started with review of the literature regarding HLLs’ identity in HL learning and

their commitment to HL learning. A rich body of literature has examined HLLs’ identity and

commitment from social psychological and poststructuralist perspectives. Related research

addressing CHLLs in particular was also reviewed. Social psychological studies have taken

an inside-out approach that tended to focus on CHLLs’ self-perceptions of their identity and

310

commitment. In contrast, poststructuralist investigations have taken an outside-in approach

that considered these issues as multiple, contradictory, ever changing, and socially

constructed through discourses. In contrast, Bourdieu’s three thinking tools of habitus, capital,

and field offer both an inside-out and an outside-in approach that can help make sense of

CHLLs’ language practices from a sociological perspective.

Beyond HL research conducted in North America, there is only a small body of

literature of relevant studies conducted in other parts of the world. Due to its idiosyncratic

colonised history, White Australia Policy, and multiculturalism, Australia is a complex social

setting for CHLLs. However, there is a scant sociological investigation taking account of both

CHLLs’ ethnic identity and their commitment to CHL learning in an Australian context.

Guided by Bourdieu’s theory, the point of departure of this study was to examine the triad of

relations amongst Chinese Australians’ ethnic identity qua habitus, their investment of

various resources as capital, and their CHL proficiency produced through language learning

practices within particular social fields.

The research design was guided by a particular philosophical position that combined a

positivist ontology, arguing that a real world exists independently of human beings’

knowledge, with a constructivist epistemology, arguing that understanding this real world is a

construction from human beings’ own perspectives and standpoint. Theoretically, Bourdieu’s

thinking through human beings’ agency and the social world’s structure fitted well in this

philosophical frame. Moreover, the methodological pluralism allowed this study to integrate

the initial quantitative investigation and the subsequent qualitative investigation to address

the research questions.

The findings from the initial quantitative phase justified Bourdieu’s theoretical

framework that agents make choices about their language practices according to the habitus

311

and capital that they have within particular fields. The analysis of the quantitative data

demonstrated that participants’ habitus of ‘Chineseness’ and their investment of various forms

of capital did contribute to their CHL proficiency. At the ontological level, Bourdieu’s

sociology can be understood as the underpinning mechanism of this emergent reality; it can

explain Chinese Australians’ CHL practices in the actual world; and the quantitative data

produced by the sampled Chinese Australians expressed this actual world in an empirical

form. However, the variance of ‘Chineseness’ and ‘capital’ only explained 62% of the

variance of ‘CHL proficiency’. The rest of the variance of ‘CHL proficiency’ was attributed

to other impact factors. At the epistemological level, habitus and capital in given fields were

the generative mechanisms that produced participants’ CHL proficiency. However, they could

not determine participants’ CHL proficiency because other social forces within fields also

contribute to this CHL proficiency.

The findings from the subsequent qualitative phase revealed the meanings associated

with CHL constructed from participants’ own perceptions. They reported developing their

habitus of ‘Chineseness’ and generating various forms of capital through CHL practice within

certain social fields. The meanings and reasoning behind CHL learning could only be

accessed through the subjective standpoint of the participants.

The findings from the quantitative phase and those from the qualitative phase

complemented and resonated with each other. The initial quantitative phase revealed that

participants’ habitus of ‘Chineseness’ and their capacity to invest in various forms of capital

generated their CHL practices that produced their CHL proficiency. In the subsequent

qualitative phase, the participants reported that they reinforced or recaptured their

‘Chineseness’ through CHL practice and they were rewarded with different forms of capital

as a return on their investment in CHL learning. As such, their CHL proficiency became

312

linguistic capital. Depending on the rules of different fields, this linguistic capital functions as

a medium to facilitate the exchange of value and forms of capital. In addition, some of the

findings from the qualitative phase conformed to those from the quantitative phase and

supported the claim that participants’ ‘Chineseness’ and their investment of capital

contributed to their CHL proficiency.

The quantitative coda demonstrated that ‘Chineseness’ and ‘capital’ were also related to

a wider range of practices associated with CHL learning. These practices included language

usage at home, years of formal CHL learning, frequency of visits to China, and self-labelling.

Furthermore, the findings from the quantitative coda also indicated the generational impacts

on the habitus of ‘Chineseness’. On the one hand, ‘Chineseness’ was durable and

transposable across space when the first-generation Chinese Australians migrated to Australia

and raised their descendents in Australia. On the other hand, ‘Chineseness’ was not

immutable across time because dispositions of this ‘Chineseness’ gradually faded away

generation after generation, accompanied by the gradual loss of CHL proficiency.

All the above findings enabled this study to talk back to the existing literature. On the

one hand, the findings of this study aligned with the existing literature that HLLs’ ethnic

identity and commitment contribute to their HL proficiency, and their HL can reinforce their

ethnic identity and their commitment to HL learning. On the other hand, the findings of this

study challenged the existing literature. The social psychological notion of motivation would

struggle to answer the questions: Why did highly motivated Adam hardly make any effort in

CHL learning when he was brought up in Indonesia? Why did En-ning, who was so

unmotivated to learn CHL when she was small, become very motivated afterwards and

decide to pursue CHL learning for her whole life? In addition, the social psychological

understanding of ethnic identity was not able to answer the question: Why did these

313

participants label themselves so differently while they were born into the same ethnic identity?

Likewise, a poststructuralist notion of investment would have trouble to explain why Bob did

not see any return on the investment of CHL learning but still committed to CHL learning

from time to time. The poststructuralist understanding of ethnic identity without a habitual

foundation is problematic: If ethnic identity is purely socially constructed, why did Adam,

growing up in an anti-Chinese social space, and En-ning, brought up in a very white area, still

keep their ‘Chineseness’, at least to a certain extent?

In contrast, Bourdieu’s triad of habitus, capital, and field offers a theoretical tool to

work with these social problems and extend social psychological and poststructuralist

understandings of HLLs’ commitment and ethnic identity. Learning Chinese did not accrue

any value in the social field in Indonesia and was not legitimised as capital, without which

Adam could hardly make any investment in CHL learning because investment demands

capital. Consequently, his CHL proficiency was low. When he moved to Australia, a place

where he considered CHL learning rewarding, he invested as much capital as possible in

CHL learning, to “catch up” in his own words. This also explained the fact that unmotivated

En-ning later became motivated to learn CHL. The social field where she was living at the

moment valued and recognised her CHL proficiency and consequently she had access to a

wider range of resources, economic, cultural, social, and symbolic, as a return on her

investment. The ethnic identity construction of Chinese Australians cannot be fully elaborated

by either the social psychological inside-out approach or the poststructuralist outside-in

approach alone. It was the habitus underpinning Adam and En-ning’s agency that made their

‘Chineseness’ durable and transposable but not immutable dispositions. It was also this

‘Chineseness’ that drove Bob to pursue CHL, consciously or unconsciously, while he did not

see any practical point of investing in CHL learning.

314

In conclusion, this study only touches a corner of sociology. Along this journey, the

researcher has been transformed from a scientist with a pure positivist ontology and

post-positivist epistemology into a scholar enlightened by the complexity and challenge of

philosophical and methodological pluralism. This different way of thinking will equip the

researcher to conduct future studies within the sociological realm. The thesis can be

concluded with a famous Chinese idiom “抛砖引玉”: Use the little to get to the big.

315

Appendices

Appendix 1: Information sheet for participation in the online survey

PARTICIPATE IN RESEARCH

Information for Prospective Participants

The following research activity has been reviewed via QUT arrangements for the conduct of research

involving human participation. If you choose to participate, you will be provided with more detailed

participant information, including who you can contact if you have any concerns.

Heritage Language for Young Chinese Australian Adults

Research Team Contacts

Guanglun Mu — PhD student Karen Dooley — supervisor

Faculty of Education Faculty of Education

Phone 0402463650 Phone 31393430

Email [email protected] Email [email protected]

Please contact the researcher team members to have any questions answered or if you require further

information about the project.

What is the purpose of the research?

The purpose of this project is to understand the interrelationship between Chinese Australians’ ‘Chineseness’,

their engagement in various Chinese cultural activities, and their Chinese Heritage Language.

Are you looking for people like me?

This study is looking for young Chinese Australian adults from cities, ranging in age from 18 to 35 years. You

may or may not be able to speak Chinese language. If you were born outside Australia, you will have to have

moved to Australia before the age of 13. You will be the potential participants if the above description fits

you.

What will you ask me to do?

Your participation will involve completing questions in an online survey, which may take you 15 minutes.

Are there any risks for me in taking part?

The theme of the study is related to your Chinese language learning experiences. There is no intent to elicit

sensitive or personal data. Your participation is the research is voluntary. You will be highly respected and

will be provided with detailed information about the study. You can withdraw from the study if you feel

uncomfortable during the survey.

Are there any benefits for me in taking part?

Though this project may not benefit you directly, you will help the researcher document and analyse the

language resources of young Chinese Australian adults.

I am interested – what should I do next?

If you would like to participate, please use the link below to start the online survey.

https://survey.qut.edu.au/survey/172445/5f10/ The researcher would be very grateful if you could complete the online survey at the end of September 2011.

Thank You! QUT Approval Number: 1100000165

https://outlook.qut.edu.au/OWA/redir.aspx?C=092a02bfc27f436eb4dd6da179fc875e&URL=https%3a%2f%2fsurvey.qut.edu.au%2fsurvey%2f172445%2f5f10%2f

316

Appendix 2: Information sheet for participation in the interview

PARTICIPANT INFORMATION FOR QUT RESEARCH

PROJECT



Guanglun (Michael) Mu – PhD Candidate Dr. Karen Dooley – Supervisor


Phone 0402463650 Phone 07 31383430

Email ([email protected]) Email ([email protected])

Description

This interview is being undertaken as part of PhD for Guanglun (Michael) Mu. The purpose of this interview

is to understand what Chinese Heritage Language means to young Chinese Australian adults. You will be

asked questions about your Chinese language usage and learning, as well as your attitudes towards Chinese

language. The interview will take about 30 minutes at a time and place convenient for you.

Participation

Your participation in this project is voluntary. You may stop the interview at anytime if you feel

uncomfortable with the questions being asked. Your decision not to participate will in no way impact upon

your current or future relationship with QUT.

Expected benefits

It is expected that this project will not directly benefit you. You will help the researcher document and

analyse the language resources of young Chinese Australian adults.

Risks

There are no risks beyond reflecting on your Chinese learning experiences associated with your participation

in this project.

QUT provides limited free counselling for research participants of QUT projects, who may experience

discomfort or distress as result of their participation in any research. Should you wish to access to this service

please contact the Clinic Receptionist of the QUT Psychology Clinic on 3138 0999.

Confidentiality

All comments and responses will be treated confidentially. All paper-based documents will be kept securely

in a locked filing cabinet in the researcher's office. E-files and recordings will be stored on a QUT

password-protected network drive. USB drives will only be used for data transfer. Only the researcher and

the supervisory team have access to the raw data.

You have indicated that you are willing to participate in a 30-min interview. You have left your name, phone

number or email address for the researcher to contact you to arrange a convenient time and place. This

information will not be used for any other purposes. The researcher will know your identity. However, an

identity code will be used on the transcripts and pseudonyms will be used in any reports. Your identity will

only be known and available to the researcher. They will not be disclosed. Any identification will be removed

from the transcripts. The code and pseudonyms will be stored separately from your raw data. Your comments

and responses will be recorded so that they can be transcribed. You will have a chance to check and amend

the transcript as you see fit. Only the researcher and the supervisory team have access to the transcripts and

audio recordings. They will be destroyed on completion of the project.

Consent to Participate

Your signature on the consent form indicates your agreement to participate.

Questions / further information about the project

Please contact the research team members named above to have any questions answered or if you require

further information about the project.

Concerns / complaints regarding the conduct of the project

QUT is committed to researcher integrity and the ethical conduct of research projects. However, if you do

have any concerns or complaints about the ethical conduct of the project you may contact the QUT Research

Ethics Unit on +61 7 3138 5123 or email [email protected]. The Research Ethics Unit is not

connected with the research project and can facilitate a resolution to your concern in an impartial manner.

Thank you for helping with this research project. Please keep this sheet for your information.

mailto:[email protected]

http://www.qut.edu.au/

317

Appendix 3: The consent form

CONSENT FORM FOR QUT RESEARCH PROJECT



Guanglun (Michael) Mu – PhD student Dr. Karen Dooley – Supervisor


Phone 0402463650 Phone 07 31383430

Email ([email protected]) Email ([email protected])

STATEMENT OF CONSENT

By signing below, you are indicating that you:

have read and understood the information document regarding this project

have had any questions answered to your satisfaction

understand that if you have any additional questions you can contact the research team

understand that you are free to withdraw at any time, without comment or penalty

understand that you can contact the Research Ethics Unit on +61 7 3138 5123 or email

[email protected] if you have concerns about the ethical conduct of the project

understand that the project will include audio recording and consent to this

agree to participate in the project

Name

Signature

Date / /

Please return this sheet to the investigator.

mailto:[email protected]

http://www.qut.edu.au/

318

Appendix 4: Interview schedule

FIRST INTERVIEW WITH INTERVIEWEE 89

Time and Date of the interview: 4pm Sydney time, 1 February, 2012

Place of the interview: Telephone interview in Room B349, Kelvin Grove

Check the audio recorders: Yes

I. Briefing

I have received the signed copy of the consent form. Thanks for that. I would also like to take

this opportunity to thank you for doing the online survey and thank you in advance for your

time to do the interview.

I will use an ID code on the transcripts and a pseudonym in any reports. Would you like to

choose your own pseudonym?

I want to assure you that there are no right or wrong answers. I have no particular agenda. I

am here to listen to and learn about your experiences.

Do you have any questions before we start the interview?

Now I will turn on the recorders.

II. Interview questions

CHL learning

First of all, I am interested in how you have gone about learning Chinese. Can you tell me

about any formal Chinese language study you have done, e.g. schools, universities,

community schools, and Confucius Institutes, etc. (probe for how long on each)?

Can you tell me about any informal Chinese learning you have pursued, e.g. with family

members, friends, independent study etc. (probe for how long on each)?

Are you still trying to improve your Chinese now? Why or why not?

Can you tell me some of your experiences or stories of how you practice your Chinese

intentionally?

CHL usage

Let’s move on to the next topic. I am interested in which language you use with whom, when

and where. Can you tell me about your language choices…

When speaking to your immediate family members / extended family members

When speaking to your (different) friends

At work

When accessing to media, e.g. internet, radio, TV, movies, music, newspaper, books,

magazines, etc.

What are your reasons behind these language choices and patterns? (Insert this question into

the above probes)

Attitudes towards CHL

Let’s talk about your attitudes towards Chinese language. Do you think your Chinese

language has benefited you in terms of …?

319

Communications with family members

Social life

Job opportunities

Chinese cultural knowledge

Credits, qualifications, awards, honours, etc.

Do you think your Chinese language will benefit you in the future? How and why?

How do you feel about (not) being able to speak Chinese?

Do you have any experiences that your Chinese language makes you feel special or different

to the others?

What do people think about your ability to use Chinese?

Family members

Friends

Colleagues

Other

If you were to have children, would you encourage them to learn another language other than

English? What is it? Why (or why not) you would encourage your children to learn Chinese?

Particular attributes

From your survey response, you are an interesting case because you have lived in several

different places. Can you tell me about your language choice when you were in Indonesia and

Singapore?

Can you tell me some of your experiences of learning Chinese when you were in Indonesia

and Singapore?

From your survey response, I know you have visited China for a few times. For what

purposes did you visit China? Can you share some of your cultural experiences in China?

How was your Chinese language helpful for your China trip?

III. Closing

Do you want to make any more comments about your Chinese language usage and learning,

as well as your attitudes towards Chinese language?

IV. Ending

Thank you again for your time. I will send you the transcript for your okay, and you are

welcome to change any wording or details to better reflect your opinions. It may take a

couple of weeks for me to get it to you.

SECOND INTERVIEW WIH INTERVIEWEE 89

Time and Date of the interview: 11am Sydney time, 26 March, 2012



I saw your facebook photos about a speech contest. Was it related to Chinese language at all?

320

I want to learn a bit more about your life in Singapore.

How old were you when you were there?

How long did you live there?

What did you do there?

You have said in the previous interview that you tend to “force” yourself to learn Chinese and

expose yourself to Chinese language “as much as possible”. How would you interpret the

driving force behind these efforts?

I am very interested in your experience of that study tour to Shanghai in 2010.

Why did you want to join in this trip?

Is it an expensive trip? Is it worth it?

Was there anyone encouraging you to do this? Where did you get this information?

You tried to learn Mandarin in Year 12 but the school did not offer it. Can you recall your

feelings at that time?

As far as I know, not many students in Australia are interested in learning Chinese.

Where did your ideas come from?

Was there anyone encouraging you to do this?

From the previous interview, I know you have had the experience that people supposed you

can speak Chinese or even supposed you are a native Chinese speaker.

What are your reflections or feelings towards this assumption?

Did this experience affect your interest in learning Chinese in any way?

When you visited China, what was your feeling when your Chinese was understood by native

Chinese speakers?

I also learned from the last interview that you had the challenge of not being able to study

Chinese when you grew up in Jakarta. However, you do want to catch it up. I would like to

know more about this part of your experience.

Did you parents ever say anything to you at that time regarding learning Chinese?

What exactly do you want to catch up?

Why you need to catch it up?

Some of my interview participants mentioned that it is a pride to be a Chinese and speak

Chinese, and it is a shame not to be able to speak Chinese. What do you think about it?

THIRD INTERVIEW WIH INTERVIEWEE 89

Time and Date of the interview: 1:30pm Sydney time, 4 April, 2012



When I read through the transcript of our second interview, I do have a quick question in

mind. At the end of the second interview, you indicated that being a Chinese and speaking

Chinese are more like “fit-in” for you. Can you please explain more? What exactly do you

mean by “fit-in”?

321

Bibliography

ACARA. (2011). The shape of the Australian curriculum: Languages. Sydney: Australiam

Curriculum, Assessment and Reporting Authority.

Ang, I. (2001). On not speaking Chinese: Living between Asia and the West. London:

Routledge.

Ang, I., & Stratton, J. (2001). Multilingualism in crisis: The new politics of race and national

identity in Australia. In I. Ang (Ed.), On not speaking Chinese: Living between Asia

and the West (pp. 95-111). London: Routledge.

Arnold, M. (1999). Literature & dogma: An essay towards a better apprehension of the Bible.

Cambridge: Chadwyck-Healey.

Au, T. K., Knightly, L. M., Jun, S. A., & Oh, J. S. (2002). Overhearing a language during

childhood. Psychological Science, 13(3), 238-243.

Australian Government. (2012). Australia in the Asian century white paper. Canberra:

Commonwealth of Australia.

Baker, C. (2003). Education as a site of language contact. Annual Review of Applied

Linguistics, 23, 95-112.

Baker, T. L. (1994). Doing social research (2nd ed.). New York: McGraw-Hill Inc.

Bankston, C. L., III, & Zhou, M. (1995). Effects of minority-language literacy on the

academic achievement of Vietnamese youths in New Orleans. Sociology of Education,

68(1), 1-17.

Barker, C. (2005). Cultural studies: Theory and practice. London: Sage.

Barnard, F. M. (1969). J. G. Herder on social and political culture. Cambridge: Cambridge

University Press.

Bartlett, L., & Holland, D. (2002). Theorising the space of literacy practices. Ways of

Knowing, 2(1), 10-22.

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin,

107(2), 238-246.

Bentler, P. M., & Chou, C.-P. (1987). Practical issues in structural modeling. sociological

Methods & Research, 16(1), 78-117.

Bernstein, B. (1973). On the classification and framing of educational knowledge. In R.

Brown (Ed.), Knowledge, education, and cultural change: Papers in the sociology of

education (pp. 363-392). London: Tavistock.

Bhatti, G. (2002). Asian children at home and at school: An ethnographic study. London:

Routledge.

Biggs, J. B. (1998). Learning from the Confucian heritage: So size doesn’t matter?

International Journal of Educational Research, 29(8), 723-738. doi:

http://dx.doi.org/10.1016/S0883-0355(98)00060-3

Biggs, J. B., & Watkins, D. A. (1996). The Chinese learner in retrospect. In D. A. Watkins & J.

B. Biggs (Eds.), The Chinese learner: Cultural, psychological, and contextual

influences (pp. 269-285). Hong Kong & Melbourne: CERC & ACER.

Binder, A. (1984). Restrictions on statistics imposed by method of measurement: Some reality,

some myth. Journal of Criminal Justice, 12(5), 467-481.

Block, D. (2002). Destabilized identities and cosmopolitanism across language and cultural

borders: Two case studies. Hong Kong Journal of Applied Linguistics, 7(2), 1-19.

Block, D. (2007). Second language identities. London; New York: Continuum.

Bourdieu, P. (1973). Cultural reproduction and social reproduction. In R. Brown (Ed.),

Knowledge, education, and cultural change: Papers in the sociology of education (pp.

71-112). London: Tavistock.

http://dx.doi.org/10.1016/S0883-0355(98)00060-3

322

Bourdieu, P. (1977a). The economics of linguistic exchanges. Social Science Information,

16(6), 645-668.

Bourdieu, P. (1977b). Outline of a theory of practice. Cambridge: Cambridge University

Press.

Bourdieu, P. (1981). Structures, strategies, and the habitus. In C. C. Lemert (Ed.), French

sociology: Rupture and renewal since 1968 (pp. 86-96). New York: Columbia

University Press.

Bourdieu, P. (1985a). The genesis of the concepts of Habitus and of Field. Sociocriticism,

2(2), 11-24.

Bourdieu, P. (1985b). The social space and the genesis of groups. Theory and Society, 14(6),

723-744. doi: 10.1007/bf00174048

Bourdieu, P. (1986). Forms of capital. In J. G. Richardson (Ed.), Handbook of theory and

research for the sociology of education (pp. 241-258). New York: Greenwood Press.

Bourdieu, P. (1987). In other words: Essays towards a reflexive sociology. Stanford, CA:

Stanford University Press.

Bourdieu, P. (1988). Homo academicus. Stanford, CA: Stanford University Press.

Bourdieu, P. (1989a). Distinction: A social critique of the judgment of taste. London:

Routledge & Kegan Paul Ltd.

Bourdieu, P. (1989b). Social space and symbolic power. Sociological Theory, 7(1), 14-25.

Bourdieu, P. (1990). The logic of practice. Cambridge: Polity Press.

Bourdieu, P. (1991). Language and symbolic power. Cambridge: Polity Press.

Bourdieu, P. (1993a). The field of cultural production. New York: Columbia University Press.

Bourdieu, P. (1993b). Sociology in question. London: Sage.

Bourdieu, P. (1994). In other words: Essays towards a reflexive sociology. Cambridge: Polity

Press.

Bourdieu, P. (1996). The state nobility: Elite schools in the field of power. Cambridge: Polity

Press.

Bourdieu, P. (1998). Practical reason: On the theory of action. Cambridge: Polity Press.

Bourdieu, P. (1999a). The order of things. In P. Bourdieu (Ed.), The weight of the world:

Social suffering in contemporary society (pp. 60-76). Cambridge: Polity Press.

Bourdieu, P. (1999b). Site effects. In P. Bourdieu (Ed.), The weight of the world: Social

suffering in contemporary society (pp. 123-129). Cambridge: Polity Press.

Bourdieu, P. (1999c). Understanding. In P. Bourdieu (Ed.), The weight of the world: Social

suffering in contemporary society (pp. 607-626). Cambridge: Polity Press.

Bourdieu, P. (2000). Pascalian meditations. Cambridge: Polity Press.

Bourdieu, P. (2005). Habitus. In J. Hillier & E. Rooksby (Eds.), Habitus: A sense of place (2

ed., pp. 43-49). Aldershot, Hants, England: Ashgate Publishing Limited.

Bourdieu, P. (2011). On television. Cambridge: Polity Press.

Bourdieu, P., & Passeron, J.-C. (1990). Reproduction in education, society and culture (2nd

ed.). London: Sage.

Bourdieu, P., & Wacquant, L. J. D. (1992). An invitation to reflexive sociology. Cambridge:

Polity Press.

Brecht, R. D., & Ingold, C. W. (2002). Tapping a national resource: Heritage Languages in

the United States. ERIC Digest (ED464515). Washington, DC: ERCIC Clearinghouse

on Languages and Linguistics.

Bucholtz, M., & Hall, K. (2009). Locating identity in language. In C. Llamas & D. Watt

(Eds.), Language and identity (pp. 9-17). Ddinburgh, UK: Edinburgh University

Press.

Burrell, G., & Morgan, G. (1979). Sociological paradigms and organizational analysis.

London: Heinemann.

323

Byrne, B. M. (2001). Structural Equation Modeling with AMOS: Basic concepts, applications,

and programming. London: Lawrence Erlbaum Associates, Inc.

Campbell, R. N., & Rosenthal, J. W. (2000). Heritage Languages. In J. W. Rosenthal (Ed.),

Handbook of undergraduate Second Language education (pp. 180-201). Mahwah, NJ:

Lawrence Erlbaum Associates, Inc.

Canagarajah, A. S. (2006). TESOL at forty: What are the issues? TESOL Quarterly, 40(1),

9-34.

Carreira, M. (2004). Seeking explanatory adequacy: A dual approach to understanding the

term "Heritage Language Learner". Heritage Language Journal, 2(1), 1-25.

Case, K., & Fair, R. (1996). Principles of economics (4th ed.). Upper Saddle River, NJ:

Prentice Hall International.

Castles, S., & Davidson, A. (2000). Citizenship and migration: Globalization and the politics

of belonging. London: Macmillan.

Cattell, R. B. (1966). The scree test for the numbers of factors. Multivariate Behavioral

Research, 1, 245-276.

Cavana, R. Y., Delahaye, B. L., & Sekaran, U. (2001). Applied business research: Qualitative

and quantitative methods. Milton, QLD: John Wiley & Sons.

Celik, S. (2007). You reap what you sow: Parental investment in language learning and its

reflection in a seven-year-old's world. Taiwan Journal of TESOL, 4(1), 99-114.

Chan, H. (2004). Becoming Australian: An afterword. In S. Couchman, J. Fitzgerald & P.

Macgregor (Eds.), After the rush: Regulation, participation, and Chinese

Communities in Australia 1860-1940 (pp. 237-244). Vic., Australia: Arena Printing

and Publishing Pty Ltd.

Chao, D.-L. (1997). Chinese for Chinese Americans: A case study. Journal of the Chinese

Language Teachers Association, 32(2), 1-13.

Chen, B. D. (1966, June 1). Sweep away all the evils, People’s Daily, p. 1.

Chinen, K., & Tucker, G. R. (2005). Heritage Language development: Understanding the

roles of ethnic identity and Saturday school participation. Heritage Language Journal,

3(1), 27-59.

Cho, G. (2000). The role of Heritage Language in social interactions and relationships:

Reflections from a language minority group. Bilingual Research Journal, 24(4),

369-384.

Cho, G., Cho, K.-S., & Tse, L. (1997). Why ethnic minorities want to develop their Heritage

Language: The case of Korean-Americans. Language, Culture and Curriculum, 10(2),

106-112.

Choi, C. Y. (1975). Chinese migration and settlement in Australia. Sydney: Sydney

University Press.

Clayre, A. (1984). The heart of the dragon. Glasgow, UK: William Collins Sons & Co. Ltd.

COAG. (1994). Asian languages and Australia's economic future - A report for the Council of

Australian Governments on a proposed national Asian languages/studies strategy for

Australian schools. Brisbane: Queensland Government Printer.

Cockerham, W. C. (2005). Health lifestyle theory and the convergence of agency and

structure. Journal of Health and Social Behavior, 46(1), 51-67.

Cockerham, W. C., & Hinote, B. P. (2009). Quantifying habitus: Future directions. In K.

Robson & C. Sanders (Eds.), Quantifying theory: Pierre Bourdieu (pp. 201-210).

Breinigsville, PA, USA: Springer Science + Business Media B. V.

Cohen, J. (1988). Statistical power analysis for the behavioural sciences (2nd ed.). New York:

Academic.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159.

Coltman, T., Devinney, T. M., Midgley, D. F., & Venaik, S. (2008). Formative versus

324

reflective measurement models: Two applications of formative measurement. Journal

of Business Research, 61(1), 1250-1262.

Comanaru, R., & Noels, K. A. (2009). Self-determination, motivation, and the learning of

Chinese as a Heritage Language. The Canadian Modern Language Review, 66(1),

131-158.

Commonwealth Advisory Committee. (1970). The teaching of Asian languages and cultures

in Australia. Canberra: Commonwealth of Australia.

Connolly, P. (2011). Using survey data to explore preschool children's ethnic awareness and

attitudes. Journal of Early Childhood Research, 9(2), 175-187.

Costa, P. I. D. (2010). From refugee to transformer: A Bourdieusian take on a Hmong

learner's trajectory. TESOL Quarterly, 44(3), 517-541.

Costello, C. Y. (2005). Professional identity crisis: Race, class, gender, and success at

professional schools. Nashville, TN: Vanderbilt University Press.

Couchman, S., Fitzgerald, J., & Macgregor, P. (2004). After the rush: Regulation,

participation and Chinese communities in Australia 1860-1940. Vic., Australia: Arena

Printing and Publishing Pty Ltd.

Cowan, M. (1963). Humanist without portfolio: An anthology of the writings of Wilhelm von

Humboldt. Detroit: Wayne State University Press.

Coyle, D. (2002). Sex, drugs & economics: An unconventional introduction to economics.

New York: Texere.

Creswell, J. W. (2008). Educational research: Planning, conducting, and evaluating

quantitative and qualitative research (3rd ed.). Upper Saddle River, N.J:

Pearson/Merrill Prentice Hall.

Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods

approaches. U.S.A.: SAGE Publications, Inc.

Creswell, J. W., & Clark, V. L. P. (2011). Designing and conducting mixed methods research.

U.S.A.: SAGE Publications, Inc.

Cummins, J. (1998). The teaching of international languages. In J. Edwards (Ed.), Language

in Canada (pp. 293-304). Cambridge: Cambridge University Press.

Cushman, J. W. (1984). A 'colonial casualty': The Chinese community in Australian

historiography. Asian Studies Association of Australia Review, 17(3), 100-113.

Danermark, B., Ekström, M., Jakobsen, L., & Karlsson, J. C. (2002). Explaining society:

Critical realism in the social sciences. London: Routledge.

De Graaf, N. D., & De Graaf, P. M. (2000). Parental cultural capital and educational

attainment in the Netherlands: A refinement of the cultural capital perspective.

Sociology of Education, 73(2), 92-111.

DEET. (1991). Australia's language: The Australian language and literacy policy.

Canberra: Australian Government Publishing Service.

Deng, Z. (2011). Confucianism, modernization and Chinese pedagogy: An introduction.

Journal of Curriculum Studies, 43(5), 561-568.

Department of Education. (1976). Report of the committee on the teaching of migrant

languages in schools. Canberra: Australian Government Publishing Service.

Diamond, J. B., Randolph, A., & Spillane, J. P. (2004). Teachers' expectations and sense of

responsibility for student learning: The importance of race, class, and organizational

habitus. Anthropology & Education Quarterly, 35(1), 75-98. doi:

10.1525/aeq.2004.35.1.75

DiMaggio, P. (1982). Cultural capital and school success: The impact of status culture

participation on the grades of U.S. high school students. American Sociological

Review, 47(2), 189-201.

Dörnyei, Z., & Skehan, P. (2003). Individual differences in second language learning. In C.

325

Doughty & M. H. Long (Eds.), Handbook of second language acquisition (pp.

589-630). Malden, MA: Blackwell.

Dovey, K. (2010). Becoming places: Urbanism, architecture, identity and power. London:

Routledge.

Draper, J., & Hicks, J. (2000). Where we've been; what we've learned. In J. B. Webb & B.

Miller (Eds.), Teaching Heritage Language Learners: Voices from the classroom (pp.

15-35). Yonkers, NY: ACTFL.

Duhachek, A., Coughlan, A. T., & Iacobucci, D. (2005). Results on the standard error of the

Coefficient Alpha Index of reliability. Marketing Science, 24(2), 294-301.

Dumais, S. A. (2002). Cultural capital, gender, and school success: The role of habitus.

Sociology of Education, 75(1), 44-68.

Durbin, J., & Watson, G. S. (1950). Testing for Serial Correlation in Least Squares

Regression, I. Biometrika, 37(3/4), 409-428.

Durbin, J., & Watson, G. S. (1951). Testing for Serial Correlation in Least Squares

Regression, II. Biometrika, 38(1/2), 159-179.

Durkheim, E. (1938). L'évolution pedagogique en France. Paris: Alcan.

Durkheim, E. (1964). The rules of sociological method. New York: Free Press.

Ellis, R. (1997). Second Language Acquisition. Oxford: Oxford University Press.

Erikson, E. (1968). Identity: Youth and crisis. New York: Norton.

Feuerverger, G. (1989). Jewish-Canadian ethnic identity and non-native language learning: A

social psychological study. Journal of Multilingual and Multicultural Development,

10(4), 327-357.

Feuerverger, G. (1991). University students' perceptions of Heritage Language learning and

ethnic identity maintenance. The Canadian Modern Language Review, 47(4),

660-677.

Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London: Sage.

Fishman, J. A. (1964). Language maintenance and language shift as a field of inquiry.

Linguistics, 9, 32-70.

Fishman, J. A. (1977). Language and ethnicity. In H. Giles (Ed.), Language, ethnicity and

intergroup relations (pp. 15-57). London: Academic Press.

Fishman, J. A. (1989). The historical and social contexts of an inquiry into language and

maintenance efforts. In J. A. Fishman (Ed.), Language and ethnicity in minority

sociolinguistic perspective (pp. 21-33). Clevedon: Multilingual Matters.

Fishman, J. A. (1991). Reversing language shift. Clevedon: Multilingual Matters.

Fishman, J. A. (2001). 300-plus years of Heritage Language education in the United States. In

J. K. Peyton, D. A. Ranard & S. McGinnis (Eds.), Heritage Languages in America:

Preserving a national resource (pp. 29-36). McHenry, IL: Centre for Applied

Linguistics / Delta Systems.

Fornell, C., & Larcker, D. F. (1981). Evaluating Structural Equation Models with

unobservable variables and measurement error. Journal of Marketing Research, 18,

39-50.

Fraenkel, J. R., & Wallen, N. E. (1996). How to design and evaluate research in education

(3rd ed.). New York: McGraw-Hill.

Francis, B., Archer, L., & Mau, A. (2009). Language as capital, or language as identity?

Chinese complementary school pupils' perspectives on the purposes and benefits of

complementary schools. British Educational Research Journal, 35(4), 519-538.

Galbally, F. (1978). Migrant services and programs: Report of the review of post-arrival

programs and services for migrants (Galbally Report). Canberra: Australian

Government Publishing Services.

Gardner, R. C. (1968). Attitudes and motivation: Their role in Second Language Acquisition.

326

TESOL Quarterly, 2(3), 141-150.

Gardner, R. C. (1985). Social psychology and language learning: The role of attitudes and

motivation. London: Edward Arnold.

Gardner, R. C., & Lambert, W. E. (1959). Motivational variables in Second Language

Acquisition. Canadian Journal of Psychology, 13(4), 266-272.

Gardner, R. C., & Lambert, W. E. (1972). Attitudes and motivation in Second Language

Learning. Rowley, MA: Newbury House.

Garnham, N., & Williams, R. (1993). Pierre Bourdieu and sociology of culture: An

introduction. In R. Collins, J. Curran, N. Garnham, P. Scannell, P. Schlesinger & C.

Sparks (Eds.), Media, culture & society: A critical reader (pp. 116-130). London:

Sage.

Gibbs, J. T., & Hines, A. M. (1992). Negotiating ethnic identity: Issues for Black-White

biracial adolescents. In M. P. Root (Ed.), Racially mixed people in America (pp.

223-238). Newbury Park, CA: Sage.

Giles, H., Bourhis, R., & Taylor, D. (1977). Toward a theory of language in ethnic group

relations. In H. Giles (Ed.), Language, ethnicity and intergroup relations (pp. 32-65).

London: Academic Press.

Giles, H., & Johnson, P. (1987). Ethnolinguistic identity theory. International Journal of the

Sociology of Language, 68, 69-99.

Gogolin, I. (2002). Linguistic and cultural diversity in Europe: A challenge for educational

research and practice. Eruopean Educational Research Journal, 1(1), 123-138.

Graham, J. M. (2006). Congeneric and (essentially) Tau-equivalent estimates of score

reliability. Educational and Psychological Measurement, 66(6), 930-944.

Green, S. B. (1991). How many subjects does it take to do a regression analysis? Multivariate

Behavioral Research, 26(3), 499-510.

Groves, R. M., Fowler, F. J., Couper, M. P., Lepkowski, J. M., Singer, E., & Tourangeau, R.

(2004). Survey methodology. New Jersey: John Wiley & Sons, Inc., Hoboken.

Guba, E. G., & Lincoln, Y. S. (1994). Competing paradigms in qualitative research. In N. K.

Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 105-117).

Thousand Oaks, CA: Sage.

Hair, J. F., Black, B. J., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate

data analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Hall, C. I. (1992). Please choose one. In M. P. Root (Ed.), Racially mixed people in America

(pp. 250-264). Newbury Park, CA: Sage.

Hancock, A. (2006). Attitudes and approaches to literacy in Scottish Chinese families.

Language and Education, 20(5), 355 - 373.

Hancock, G. R., & Mueller, R. O. (2001). Rethinking construct reliability within latent

variables systems. In R. Cudeck, S. D. Toit & D. Sörbom (Eds.), Structural Equation

Modeling: Present and future- A festschrift in honor of Karl Jöreskog (pp. 195-216).

Lincolnwood: Scientific Software International.

Haralambos, M., & Holborn, M. (2004). Sociology: Themes and perspectives (6th ed.).

London: Harper Collins Publishers Limited.

Hase, P. (1995). The alliance of ten: Settlement and politics in the Sha Tau Kok Area. In D.

Faure & H. F. Siu (Eds.), Down to earth: The territorial bond in South China (pp.

123-160). Standford, California: Standford University Press.

He, A. W. (2004). Identity construction in Chinese Heritage Language classes. Pragmatics,

14(2/3), 199-216.

He, A. W. (2005). Discipline, directives, and deletions: Grammar and interaction in Chinese

heritage language classes. In C. Holten & J. Frodesen (Eds.), The power of context in

language teaching and learning: A Festschrift for Marianne Celce-Murcia (pp.

327

115-126). Boston: Thomson Heinle.

He, A. W. (2006). Toward an identity theory of the development of Chinese as a Heritage

Language. Heritage Language Journal, 4(1), 1-28.

He, A. W. (2008). Chinese as a Heritage Language: An introduction. In A. W. He & Y. Xiao

(Eds.), Chinese as a Heritage Language: Fostering rooted world citizenry (pp. 1-12).

Hololulu: University of Hawai’i Press.

Hinton, L. (1999). Involuntary language loss among immigrants: Asian-American linguistic

autobiographies. (Report No. EDO-FL-99-10). Retrieved from CAL Center for

Applied Linguistics website: http://www.cal.org/resources/digest/involuntary.html.

Ho, D. Y. F., Xie, W., Liang, X., & Zeng, L. (2012). Filial piety and traditional Chinese

values: A study of high and mass cultures. PsyCh Journal, 1(1), 40-55. doi:

10.1002/pchj.6

Ho, R. (2006). Handbook of univariate and multivariate data analysis and interpretation with

SPSS. Boca Raton: Chapman & Hall/CRC.

Hodge, B., & O’Carroll, J. (2006). Borderwork in multicultural Australia. Sydney: Allen &

Unwin.

Holland, D., Lachicotte, W., Skinner, D., & Cain, D. (1998). Identity and agency in cultural

worlds. Cambridge, MA: Harvard University Press.

Holmes, D., Hughes, K., & Julian, R. (2007). Australian sociology: A changing society.

Frenchs Forest, N.S.W.: Pearson Education Australia.

Horvat, E. M., & Antonio, A. L. (1999). "Hey, those shoes are out of uniform": African

American girls in an elite high school and the importance of habitus. Anthropology &

Education Quarterly, 30(3), 317-342.

Hurtado, A., & Gurin, P. (1995). Ethnic identity and bilingualism attitudes. In A. M. Padilla

(Ed.), Hispanic psychology: Critical issues in theory and research (pp. 89-103).

Thousand Oaks, CA: Sage.

Imbens-Bailey, A. (1996). Ancestral language acquisition: Implications for aspects of ethnic

identity among Armenian American children and adolescents. Journal of Language

and Social Psychology, 15(4), 422-443.

Ingram, D. E. (2000, 18 August). Language policy and language education in Australia.

Paper presented at the Akita University, Japan.

Jackson, P. (2009). Pierre Bourdieu. In J. Edkins & N. Vaughan-Williams (Eds.), Critical

theorists and international relations (pp. 102-113). New York: Routledge.

Jenkins, R. (1992). Pierre Bourdieu. London and New York: Routledge.

Jones, P. (2004). What happened to Australia's Chinese between the World Wars? In S.

Couchman, J. Fitzgerald & P. Macgregor (Eds.), After the rush: Regulation,

participation, and Chinese communities in Australia 1860-1940 (pp. 217-236). Vic.,

Australia: Arena Printing and Publishing Pty Ltd.

Jones, P. (2005). Chinese-Australian journeys: Records on travel, migration and settlement,

1860-1975. Canberra: National Archives.

Jöreskog, K. G., & Sörbom, D. (1996). LISERAL 8 user's reference guide. Mooresville, IN:

Scientific Software International.

Kaiser, H. F. (1960). An application of electronic computers to factor analysis. Educational

and Psychological Measurement, 20(1), 141-151.

Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31-36.

Ke, C. (1998). Effects of language background on the learning of Chinese characters among

foreign languages students. Foreign Language Annals, 31(1), 91-102.

Kelley, K., & Lai, K. (2011). Accuracy in parameter estimation for the Root Mean Square

Error of Approximation: Sample size planning for narrow confidence intervals.

Multivariate Behavioral Research, 46(1), 1-32. doi: 10.1080/00273171.2011.543027

http://www.cal.org/resources/digest/involuntary.html

328

Kiang, L. (2008). Ethnic self-labeling in young American adults from Chinese backgrounds.

Journal of Youth and Adolescence, 37(1), 97-111.

Kim, K. H. (2005). The relation among fit indexes, power, and sample size in Structural

Equation Modeling. Structural Equation Modeling: A Multidisciplinary Journal,

12(3), 368-390. doi: 10.1207/s15328007sem1203_2

Kline, P. (1999). The handbook of psychological testing (2nd ed.). London: Routledge.

Kramsch, C. (2008). Pierre Bourdieu: A biographical memoir. In A. L. J. Albright (Ed.),

Bourdieu and literacy education (pp. 33-49). London: Routledge.

Kvale, S. (1996). Interviews: An introduction to qualitative research interviewing. London:

Sage.

Labov, W. (1972). Language in the inner city: Studies in the Black English vernacular.

Philadelphia: University of Pennsylvania Press.

Lai, K. L. (2008). An introduction to Chinese philosophy. Cambridge: Cambridge University

Press.

Landis, R. S., Beal, D. J., & Tesluk, P. E. (2000). A Comparison of approaches to forming

composite measures in Structural Equation Models. Organizational Research

Methods, 3(2), 186-207. doi: 10.1177/109442810032003

Lane, J. F. (2000). Pierre Bourdieu: A critical introduction. London: Pluto Press.

Lao, C. (2004). Parents' attitudes toward Chinese-English bilingual education and

Chinese-language use. Bilingual Research Journal, 28(1), 99-121.

Lee, J. S. (2002). The Korean language in America: The role of cultural identity in Heritage

Language learning. Language, Culture and Curriculum, 15(2), 117-133.

Lee, W. O. (1996). The cultural context for Chinese learners: Conceptions of learning in the

Confucian tradition. In D. A. Watkins & J. B. Biggs (Eds.), The Chinese Learner (pp.

45-67). Hong Kong: Comparative Education Research Centre.

Lehman, R. (1991). Statistics and research design in the behavioral sciences. Belmont,

California: Wadsworth Publishing Company.

Levesque, R. J. R. (2007). The ethnicity of adolescent research. Journal of Youth and

Adolescence, 36(4), 375-389.

Levinson, D. J. (1978). The seasons of a man's life. New York: Ballantine.

Li, D. (2005, 25-29 July). Attitudes, motivations and identities in learning Chinese as a

Heritage Language. Paper presented at the 14th World Congress of Applied

Linguistics, Madison, WI.

Li, D., & Duff, P. A. (2008). Issues in Chinese Heritage Language education and research at

the postsecondary level. In A. W. He & Y. Xiao (Eds.), Chinese as a Heritage

Language: Fostering rooted world citizenry (pp. 13-36). Honolulu: University of

Hawaii's Press, National Foreign Language Resource Center.

Li, W. (1994). Three generations, two languages, one family: Language choices and

language shift in a Chinese community in Britain. Clevedon, England: Multilingual

Matters.

Li, W., & Zhu, H. (2011). Changing hierarchies in Chinese language education for the British

Chinese learners. In L. Tsung & K. Cruickshank (Eds.), Teaching and learning

Chinese in global contexts: Multimodality and literacy in the new media age (pp.

11-27). London & New York: Continuum International Publishing Group.

Lin, H. (2004). The development of Hokkien and its relationship with the cultural identity in

Fujian Province in China. China: Southeast Research.

Lingard, B., Rawolle, S., & Taylor, S. (2005). Globalising policy sociology in education:

Working with Bourdieu. Journal of Education Policy, 20(6), 759-777.

Lo-Philip, S. W.-Y. (2010). Towards a theoretical framework of Heritage Language literacy

and identity processes. Linguistics and Education: An International Research Journal,

329

21(4), 282-297

Lo Bianco, J. (1987). National policy on languages. Canberra: Australian Government

Publishing Service.

Lo Bianco, J. (1990). Making language policy: Australia's experience. In R. B. Baldauf & A.

Luke (Eds.), Language planning and education in Australasia and the South Pacific

(pp. 47-79). Philadelphia: Multilingual Matters.

Lo Bianco, J. (2000). Making languages an object of public policy. Agenda, 7(1), 47-61.

Lo Bianco, J. (2011). Chinese: The gigantic up-and-comer. In L. Tsung & K. Cruickshank

(Eds.), Teaching and learning Chinese in global contexts: Multimodality and literacy

in the new media age (pp. xiii-xxiv). London & New York: Continuum International

Publishing Group.

Lodico, M. G., Spaulding, D. T., & Voegtle, K. H. (2006). Methods in educational research:

From theory to practice. San Francisco: Jossey-Bass.

Lokan, J., Ford, P., & Greenwood, L. (1997). Maths & science on the line: Australian middle

primary students' performance in the third international mathematics and science

study. Melbourne: Australia Council for Educational Research.

Louie, A. (2004). Chineseness across borders: Renegotiating Chinese identities in China and

the United States. Durham and London: Duke University Press.

Louie, K. (2008). Defining modern Chinese culture. In K. Louie (Ed.), The Cambridge

companion to modern Chinese culture (pp. 1-19). Cambridge: Cambridge University

Press.

Lu, X., & Li, G. (2008). Motivation and achievement in Chinese language learning: A

comparative analysis. In A. W. He & Y. Xiao (Eds.), Chinese as a Heritage Language:

Fostering rooted world citizenry (pp. 89-108). Hololulu: The University of Hawaii

Press.

Luke, A. (2009). Race and language as capital in school: A sociological template for language

education reform. In R. Kubota & A. M. Y. Lin (Eds.), Race, culture and identities in

second language education: Exploring critically engaged practice (pp. 286-308).

London: Routledge.

Luke, A. (2010). Documenting reproduction and inequality: Revisiting Jean Anyon's “social

class and school knowledge”. Curriculum Inquiry, 40(1), 167-182. doi:

10.1111/j.1467-873X.2009.00474.x

Luo, S.-H., & Wiseman, R. L. (2000). Ethnic language maintenance among Chinese

immigrant children in the United States. International Journal of Intercultural

Relations, 24(3), 307-324.

Lynch, A. (2003). The relationship between Second and Heritage Language Acquisition:

Notes on research and theory building. Heritage Language Journal, 1, 1-18.

Macknight, C. C. (1976). The voyage to Marege' : Macassan trepangers in Northern

Australia Melbourne: Melbourne University Press.

Mactavish, J. B., & Schleien, S. J. (2000). Beyond qualitative and quantitative data linking:

An example from a mixed method study of family recreation. Therapeutic Recreation

Journal, 34(2), 154-163.

Marks, G. N. (2009). The influence of cultural capital on educational and early labor market

outcomes of young people in Australia. In K. Robson & C. Sanders (Eds.),

Quantifying theory: Pierre Bourdieu (pp. 89-104). Breinigsville, PA, USA: Springer

Science + Business Media B. V.

Maton, K. (2008). Habitus. In M. Grenell (Ed.), Pierre Bourdieu: Key concepts (pp. 49-66).

Stocksfield: Acumen.

McClelland, K. (1990). Cumulative disadvantage among the highly ambitious. Sociology of

Education, 63(2), 102-121.

330

McGinnis, S. (2005). More than a silver bullet: The role of Chinese as a Heritage Language

in the United States. The Modern Language Journal, 89(4), 592-594.

McGinnis, S. (2008). From mirror to compass: The Chinese Heritage Language education

sector in the United States. In D. M. Brinton, O. Kagan & S. Bauckus (Eds.), Heritage

Language education: A new field emerging (pp. 229-242). New York: Routledge.

McLaren, A. (2011). Asian languages enrolments in Australian higher education 2008-09.

Canberra: The Asian Association of Australia.

Menard-Warwick, J. (2005). Intergenerational trajectories and sociopolitical context: Latina

immigrants in adult ESL. TESOL Quarterly, 39(2), 165-185.

Merriam, S. B. (1998). Qualitative research and case study applications in education. San

Francisco, CA: Jossey-Bass.

Mertens, D. M. (2005). Research and evaluation in education and psychology: Integrating

diversity with quantitative, qualitative, and mixed methods. U.S.A.: Sage.

Mertler, C. A., & Charles, C. M. (2005). Introduction to educational research (5th ed.).

U.S.A.: Pearson Education, Inc.

Mu, G. M. (2012, 24-27 November). Does ‘habitus’ count in Chinese Australians’

mathematics achievement? Paper presented at the 2nd International STEM in

Education Conference, Beijing.

Nardi, P. M. (2006). Doing a survey research: A guide to quantitative methods (2nd ed.).

U.S.A.: Pearson Education Inc.

Noels, K. A. (2005). Orientations to learning German: Heritage Language learning and

motivational substrates. The Canadian Modern Language Review, 62, 285-312.

Norton, B. (1994, March). Language learning, social identity, and immigrant women. Paper

presented at the Annual Meeting of the Teachers of English to Speaks of Other

Languages, Baltimore.

Norton, B. (1995). Social identity, investment, and language learning. TESOL Quarterly,

29(1), 9-31.

Norton, B. (2000). Identity and language learning: Gender, ethnicity, and educational change.

Harlow, England: Pearson Education.

Ochs, E. (1993). Constructing social identity: A language socialization perspective. Research

on Language and Social Interaction, 26, 287-306.

Oh, J. S., & Fuligni, A. J. (2010). The role of Heritage Language development in the ethnic

identity and family relationships of adolescents from immigrant backgrounds. Social

Development, 19(1), 202-220.

Ong, A. (1999). Flexible citizenship: The cultural logistics of transnationality. Durham, N.C.:

Duke University Press.

Ong, A., Dominguez, V. R., Friedman, J., Schiller, N. G., Stolcke, V., Wu, D. Y. H., & Ying,

H. (1996). Cultural citizenship as subject-making: Immigrants negotiate racial and

cultural boundaries in the United States [and comments and reply]. Current

Anthropology, 37(5), 737-762.

Orton, J. (2008). Chinese language education in Australian schools. Melbourne: Melbourne

Graduate School of Education, The University of Melbourne.

Pahl, K. (2008). Tracing habitus in texts: Narratives of loss, displacement and migration in

homes. In J. Albright & A. Luke (Eds.), Bourdieu and literacy education (pp.

187-208). London: Routledge.

Pao, D. L., Wong, S. D., & Teuben-Rowe, S. (1997). Identity formation for mixed-heritage

adults and implications for educators. TESOL Quarterly, 31(3), 622-631.

Patton, M. Q. (1990). Qualitative evaluation and research methods. Newbury Park, CA:

Sage.

Pavlenko, A. (2002). Poststructuralist approaches to the study of social factors in second

331

language learning and use. In V. Cook (Ed.), Portraits of the L2 user (pp. 275-302).

Clevedon, Buffalo, Toronto, Sydney: Multilingual Matters Ltd.

Pennycook, A. (2003). Global Englishes, rip slime, and performativity. Journal of

Sociolinguistics, 7(4), 513-533.

Peräkylä, A. (2004). Reliability & validity in research based on naturally ocurring social

interaction. In D. Silverman (Ed.), Qualitative research: Theory, method and practice

(2 ed., pp. 283-304). London: Sage.

Peyton, J. K., Ranard, D., & McGinnis, S. (2001). Charting a new course: Heritage Language

education in the United States. In J. K. Peyton, D. Ranard & S. McGinnis (Eds.),

Heritage Languages in America: Preserving a national resource (pp. 3-26).

Washington, DC: Center for Applied Linguistics.

Phinney, J. S., & Ong, A. D. (2007). Conceptualisation and measurement of ethnic identity:

Current status and future directions. Journal of Counseling Psychology, 54(3),

271-281.

Phinney, J. S., Romero, I., Nava, M., & Huang, D. (2001). The role of language, parents, and

peers in ethnic identity among adolescents in immigrant families. Journal of Youth

and Adolescence, 30(2), 135.

Price, C. A. (1975). Foreword. In C. Y. Choi (Ed.), Chinese migration and settlement in

Australia (pp. ix-x). Sydney: Sydney University Press.

Purdey, J. (2006). Anti-Chinese violence in Indonesia, 1996-1999. Honolulu: University of

Hawai'i Press.

Reay, D. (1995). 'They employ cleaners to do that': Habitus in the primary classroom. British

Journal of Sociology of Education, 16(3), 353-371.

Reay, D. (2004). 'It's all becoming a habitus': Beyond the habitual use of habitus in

educational research. British Journal of Sociology of Education, 25(4), 431-444.

Root, M. P. (1992). Within, between, and beyond race. In M. P. Root (Ed.), Racially mixed

people in America (pp. 3-11). Newbury Park, CA: Sage.

Roscigno, V. J., & Ainsworth-Darnell, J. W. (1999). Race, cultural capital, and educational

resources: Persistent inequalities and achievement returns. Sociology of Education,

72(3), 158-178.

Rosenthal, D. A., & Feldman, S. S. (1992). The nature and stability of ethnic identity in

Chinese youth: Effects of length of residence in two cultural contexts. Journal of

Cross-Cultural Psychology, 23(2), 214-227.

Rowe, K. (2002, October 11). The measurement of latent and composite variables from

multiple items or indicators: Applications in performance indicator systems. Paper

presented at the Royal Molbourne Institute of Technology Statistics Seminar Series,

Melbourne.

Rowsell, J. (2008). Improvising on artistic habitus: Sedimenting identity into art. In J.

Albright & A. Luke (Eds.), Bourdieu and literacy education (pp. 233-251). London:

Routledge.

Rummens, J. A. (2003). Conceptualizing identity and diversity: Overlaps, intersections, and

processes. Canadian Ethnic Studies Journal, 35(3), 10-25.

Ryan, F., Coughlan, M., & Cronin, P. (2009). Interviewing in qualitative research: The

one-to-one interview. International Journal of Therapy & Rehabilitation, 16(6),

309-314.

Sadovnik, A. R. (2011). Theory and research in the sociology of education. In A. R. Sadovnik

(Ed.), Sociology of education: A critical reader (pp. 3-22). New York: Routledge.

Salili, F., Chiu, C. Y., & Lai, S. (2001). The influence of culture and context on students’

achievement orientations. In F. Salili, C. Y. Chiu & Y. Y. Hong (Eds.), Student

motivation: The culture and context of learning (pp. 221-247). New York: Plenum.

332

Shen, H. H. (2003). A comparison of written Chinese achievement among Heritage Learners

in homogeneous and heterogeneous groups. Foreign Language Annals, 36(2),

258-266.

Shin, S. J. (2010). "What about me? I'm not like Chinese but I'm not like American”: Heritage

Language learning and identity of mixed-heritage adults. Journal of Language,

Identity & Education, 9(3), 203-219.

Silverman, D. (1993). Interpreting qualitative data: Methods for analyzing talk, text and

interaction. London: Sage.

Smolicz, J. J. (1981). Language as core value of culture. In H. B. Beardsmore (Ed.), Elements

of bilingual theory (pp. 104-124). Brussels: Vrije Universiteit.

Stevens, J. P. (2002). Applied multivariate statistics for the social sciences (4th ed.). Hillsdale,

NJ: Erlbaum.

Studenmund, A. H., & Cassidy, H. J. (1987). Using economietrics: A practical guide. Boston:

Little Brown.

Sturak, K., & Naughten, Z. (2010). The current state of Chinese, Japanese, Indonesian and

Korean language education in Australian schools: Four languages, four stories

(ISBN: 978 1 74200 105 0). Carlton South, Vic.: Education Services Australia

Retrieved from http://www.deewr.gov.au/Schooling/NALSSP/Pages/Resources.aspx.

Sullivan, A. (2001). Cultural capital and educational attainment. Sociology, 35(4), 893-912.

Sullivan, A. (2002). Bourdieu and education: How useful is Bourdieu's theory for researchers?

The Netherlands' Journal of Social Sciences, 38(2), 144-166.

Sun, H., Hu, Z., & Huang, X. (2007). Languages in China. Beijing: The Commercial Press.

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston:

Allyn and Bacon.

Tajfel, H. (1981). Human groups and social categories. Cambridge: Cambridge University

Press.

Tan, S.-H. (2008). Modernizing Confucianism and ‘new Confucianism’. In K. Louie (Ed.),

The Cambridge companion to modern Chinese culture (pp. 135-154). Cambridge:

Cambridge University Press.

Tashakkori, A., & Teddlie, C. (2003). Handbook of mixed methods in social and behavioral

research. Thousand Oaks, CA: Sage.

Thomson, P. (2008). Field. In M. Grenfell (Ed.), Pierre Bourdieu: Key concepts (pp. 67-81).

Stocksfield, UK: Acumen.

Tod, A. (2006). Interviewing. In K. Gerrish & A. Lacey (Eds.), The research process in

nursing (pp. 337-352). Oxford: Blackwell Publishing.

Tse, L. (2001). Why don’t they learn English? Separating fact from fallacy in the U.S.

language debate. New York: Teacher’s College Press.

Tsung, L., & Cruickshank, K. (2011). Teaching and learning Chinese: A research agenda. In L.

Tsung & K. Cruickshank (Eds.), Teaching and learning Chinese in global contexts:

Multimodality and literacy in the new media age (pp. 213-224). London & New York:

Continuum International Publishing Group.

Valdés, G. (2001). Heritage Language students: Profiles and personalities. In J. K. Peyton, D.

A. Ranard & S. McGinnis (Eds.), Heritage Languages in America: Preserving a

national resource (pp. 29-36). McHenry, IL: Centre for Applied Linguistics / Delta

Systems.

Van Deusen-Scholl, T. (2003). Toward a definition of Heritage Language: Sociopolitical and

pedagogical considerations. Journal of Language, Identity & Education, 2(3),

211-230.

Veenstra, G. (2009). Transmutations of capitals in Canada: A ‘social space’ approach. In K.

Robson & C. Sanders (Eds.), Quantifying theory: Pierre Bourdieu (pp. 61-74).

http://www.deewr.gov.au/Schooling/NALSSP/Pages/Resources.aspx

333

Breinigsville, PA, USA: Springer Science + Business Media B. V.

Von Glasersfeld, E. (2001). The radical constructivist view of science. Foundation of Science,

6(1-3), 31-43.

Vryonides, M. (2007). Social and cultural capital in educational research: Issues of

operationalisation and measurement. British Educational Research Journal, 33(6),

867-885.

Wacquant, L. J. D. (1989). Towards a reflexive sociology: A workshop with Pierre Bourdieu.

Sociological Theory, 7(1), 26-63.

Walonick, D. (1993). StatPac gold IV: Marketing research and survey edition. Minneapolis,

MN: StatPac, Inc.

Wang, S. C. (2007). Building societal capital: Chinese in the US. Language Policy, 6(1),

27-52. doi: 10.1007/s10993-006-9043-2

Watson, J. L. (1975). Emigration and the Chinese lineage: The Mans in Hong Kong and

London. Berkeley: University of California Press.

Webb, J., Schirato, T., & Danaher, G. (2002). Understanding Bourdieu. Australia: Allen &

Unwin.

Webb, J. B., & Miller, B. L. (2000). Teaching Heritage Language Learners: Voices from the

classroom. Yonkers, NY: ACTFL.

Weedon, C. (1997). Feminist practice and poststructuralist theory (2nd ed.). Oxford:

Blackwell.

Weerakkody, N. (2010). Roots and routes. In M. Kumar, H. D'Cruz & N. Weerakkody (Eds.),

Where are you from? Voices in transition (pp. 25-37). Altona, Vic.: Common Ground.

Weger-Guntharp, H. (2006). Voices from the margin: Developing a profile of Chinese

Heritage Language Learners in the FL classroom. Heritage Language Journal, 4(1),

29-46.

Wen, X. (1997). Motivation and language learning with students of Chinese. Foreign

Language Annals, 30(2), 235-251. doi: 10.1111/j.1944-9720.1997.tb02345.x

Wen, X. (2011). Chinese language learning motivation: A comparative study of Heritage and

non-Heritage Learners. Heritage Language Journal, 8(3), 41-66.

Williams, R. (1988). Keywords: A vocabulary of culture and society. London: Fontana Press.

Wong-Fillmore, L. (1991). When learning a second language means losing the first. Early

Childhood Research Quarterly, 6(3), 323-346.

Wong, K. F., & Xiao, Y. (2010). Diversity and difference: Identity issues of Chinese Heritage

Language Learners from dialect backgrounds. Heritage Language Journal, 7(2),

153-187.

Woodward, K. (1997). Identity and difference. London: Sage.

Woodward, K. (2003). Understanding identity. London: Arnold.

Xi, X., & Gao, C. M. (2005). A brief history of the Cultural Revolution (2nd ed.). Beijing:

Chinese Communist Party History Press.

Xiao, Y. (2006). Heritage Learners in the Chinese language classroom: Home background.

Heritage Language Journal, 4(1), 47-56.

Xiao, Y. (2008). Home literacy environment in Chinese as a Heritage Language. In A. W. He

& Y. Xiao (Eds.), Chinese as a Heritage Language: Fostering rooted world citizenry

(pp. 151-166). Hololulu: University of Hawai’i Press, National Foreign Language

Resource Center.

Xing, J. Z. (2006). Teaching and learning Chinese as a Foreign Language: A pedagogical

grammar. Hong Kong: Hong Kong University Press.

Yamamoto, Y., & Brinton, M. C. (2010). Cultural capital in East Asian educational systems:

The case of Japan. Sociology of Education, 83(1), 67-83.

Yang, J. S. R. (2003). Motivational orientations and selected learner variables of East Asian

334

language learners in the United States. Foreign Language Annals, 36(1), 44-56. doi:

10.1111/j.1944-9720.2003.tb01931.x

Yang, K. S. (1986). Chinese personality and its change. In M. H. Bond (Ed.), The psychology

of the Chinese people (pp. 106-170). Hong Kong: Oxford University Press.

Yuan, K. H., Marshall, L. L., & Bentler, P. M. (2002). A unified approach to exploratory

factor analysis with missing data, nonnormal data, and in the presence of outliers.

Psychometrika, 67(1), 95-122.

Yurdugul, H. (2008). Minimum sample size for Cronbach's coefficient alpha: A Monte-Carlo

study. H. U. Journal of Education, 35, 397-405.

Zacher, J. (2008). Social hierarchies and identity politics. In J. Albright & A. Luke (Eds.),

Bourdieu and literacy education (pp. 252-278). London: Routledge.

Zhang, D., & Slaughter-Defoe, D. T. (2009). Language attitudes and Heritage Language

maintenance among Chinese immigrant families in the USA. Language, Culture and

Curriculum, 22(2), 77-93.

Zhang, J. (2009). Mandarin maintenance among immigrant children from the People’s

Republic of China: An examination of individual networks of linguistic contact.

Language, Culture and Curriculum, 22(3), 195-213.

Zhang, Q. Y. (2009). Confucianism for today. Beijing: Peking University Press.

Zhao, D., & Singh, M. (2011). Why do Chinese-Australian students outperform their

Australian peers in mathematics: A comparative case study. International Journal of

Science and Mathematics Education, 9(1), 69-87. doi: 10.1007/s10763-010-9214-7

Zumbo, B. D., & Zimmerman, D. W. (1993). Is the selection of statistical methods governed

by level of measurement? Canadian Psychology, 34(4), 390-399.

Date post:	09-May-2020
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Heritage Language for Chinese Australians: …eprints.qut.edu.au/63295/1/Guanglun_Mu_Thesis.pdfThis...

Documents