Code-mixing: Linguistic form and socio-cultural meaning

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
Code-mixing: Linguistic form and socio-cultural meaning
Judy Woon Yee Ho
Lingnan University, Hong Kong
Introduction
Code-mixing is the change of one language to another within the same utterance or in the same oral/
written text. It is a common phenomenon in societies in which two or more languages are used. Stud-
ies of code-mixing enhance our understanding of the nature, processes and constraints of language
(Myers-Scotton, 1993a; Boeschoten, 1998; Azuma, 1998), and of the relationship between language
use and individual values, communicative strategies, language attitudes and functions within particular
socio-cultural contexts. (Auer, 1998; Jacobson, 1998; Myers-Scotton, 1993b; Lüdi, 2003)
The historical development of Hong Kong, a former British colony, has provided a solid ground for in-
ter-lingual and inter-dialectal contacts (Bolton, 2002). Even though the majority of Hong Kong Chinese
are native speakers of Cantonese, English has always played an important role in the government,
education and employment (Luke and Richards, 1982; Li, 1999; Pennington, 1998a), and the unifica-
tion with China in 1997 has not relinquished the position of English. The Basic Law, a mini-constitution
of Hong Kong, has accorded official status to both Chinese and English. This is meant to be a visible
sign that Hong Kong retains its own social and economic systems within the conceptualization of “one
country, two systems”. The use of English is also seen as an indispensable part of positioning Hong
Kong as an international financial centre. The regular use of Cantonese and English results naturally
in code-mixing.
Quite a number of papers have been written about this phenomenon in Hong Kong. Gibbons (1987) is
a classic study of university students’ code choice and code-mixing behaviour and has generated
much interest in this field. Luke (1998) presents a model of “expedient” and “orientational” code-mixing
and bases his arguments on the differentiation between high and low varieties of Cantonese and so-
cial motivation. Li (2000) finds this model “intuitively appealing” but challenges the assumption of
“translation equivalent”. Instead he proposes four motivations for codeswitching in Hong Kong:
euphemism, specificity, bilingual punning and principle of economy. More recently research of code-
mixing has been extended to computer-mediated texts. For example, functional complementarity be-
tween Chinese and English is analyzed in Ho (2006)’s study of ICQ histories. The paper investigates
how code-mixing enables bilinguals to effectively manage the pressure to achieve specific purposes in
on-line, real time communication.
So far relatively little has been done to examine code-mixing in different communicative contexts with
a qualitative approach and to unravel the finer details of code-mixing. The present study is aimed at
filling this gap. It examines the linguistic form of Cantonese-English mix and explores its socio-cultural
meaning, using both quantitative and qualitative methods.
Methodology
The present research is part of a larger project on code choice and code-mixing. The subjects were
52 tertiary students in Hong Kong. Students were asked to submit a language diary, modeled on Gib-
bons (1987). They recorded their linguistic behaviour for two days, one from Monday to Friday and the
other from Saturday to Sunday. The following details were required: time, place and circumstance of
the communicative event, style and topic of the talk, the interlocutor’s role, education, age, sex, ethnic
and linguistic background, and the code(s) used. Then they accounted for the choice of linguistic code
in a short essay. Fifty-one hours of verbal exchanges were recorded. For this paper, eighteen hours of
recordings covering a wide range of situations, topics and interlocutors were selected and studied in
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
detail. Students’ essays, responses in focus group interviews and individual comments were also ana-
lyzed.
Code-Mixing: Linguistic Form
Some subjects have reported difficulty in coding for their language diaries due to delicate differentia-
tion between pure Cantonese and a low level of English mix in Cantonese speech. In some cases,
English proper names and acronyms were not counted as English linguistic elements. In others, cer-
tain English words have become somewhat lexicalized in Cantonese and treated as pure Cantonese.
Examples include “Mummy” (pronounced as maa1mi4), “BB” (baby) pronounced as bi4bi1, “OK” and
“Bye-Bye”.
In fact, the difficulty in coding has highlighted a significant issue: how mixed is a mixed code? Defini-
tions of code-mixing or code-switching are reviewed in detail in. Li (1998; 2000). This paper adopts a
broader definition. Code-mixing refers to any admixture of linguistic elements of two or more language
systems in the same utterance at various levels: phonological, lexical, grammatical and orthographi-
cal. Due to constraints of space, the discussion will focus on lexical and grammatical code-mixing.
To explore the nature and extent of code-mixing, the English elements in the recorded materials have
been categorized according to the level of grammatical constituency (See Table 1). Unit types refer to
the letters/ words/ phrases etc found under each category, and tokens refer to the numbers of times
the letters/ words/ phrases etc were used by subjects or interlocutors.
Table 1: Levels of English in Cantonese Utterances
Subject
Interlocutor
Total
Linguistic units
Unit types
tokens
tokens
tokens
1. Letters of the alphabet
118
140
189
329
Letter names
28
40
30
70
Acronyms
90
100
159
259
2. Short Forms
35
65
91
156
3. Proper Nouns
220
355
341
696
Personal Names
136
265
255
520
Impersonal Names
84
90
86
176
4. Lexical Words
726
919
875
1794
5.
Phrases
252 172 180
352
6. Incomplete/Minor Sentences
34
66
91
157
7. Single Full Sentences
98
54
111
165
8. Two-Sentence Units
3
2
7
9

1486 1773 1885
3658
1. Letters of the alphabet
The first category is the smallest orthographical unit — letters of the alphabet. It is common in Hong
Kong to use letters to name or distinguish between objects e.g. (Hall) A/B/C or provide assessment
e.g. (Grade) A in an examination. This category also reflects another social practice in Hong Kong:
English acronyms are used to refer to objects or ideas that are common in a bilingual, business-
oriented international city. Examples include OT for “Overtime”, OL for “Office Lady” and MTR for
“Mass Transit Railway”. Students have formed the habit of abbreviating terms used within a specific
and familiar context, e.g SU for “Student Union” and GPA for “Grade Point Average”.
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
2. Short forms
These are formed by morphologically truncating lexical words and are chiefly related to courses of
study, e.g. “transla” for “translation”, “pre” or “present” for “presentation”, and “soci” for “sociology”.
3. Proper nouns
Personal names take up about 14% of the total number of tokens of English mix in the recorded mate-
rials. The prevalent adoption of an English name by Hong Kong individuals is an indication of their
Westernized orientation. Impersonal proper names which consist mainly of English names of brands,
companies, products and buildings also express this orientation at the societal level.
4. Lexical words
This category makes up the largest portion of code-mixed linguistic items and includes IT-related
terms.
As noted by Chan (1998, p.193),“Cantonese-English code-mixing most often takes the form of single
English words surrounded by Cantonese constituents.” Most of these English words are “content
words”, rather than “function words” (Gibbons, 1987). Pennington (1998b, p.9) uses the term “lexical
bilingualism” to describe this pattern of code-mixing and suggests that “the knowledge of English by
Hong Kong Chinese is more a matter of familiarity with a certain number of words and phrases than
fluency in a second language.”
The present data largely support these findings. Examples 1-3 show an increasing level of lexical-
grammatical complexity in code-mixing. The extent of code-mixing in Example 1 is quite minimal: an
insertion of a single English word into a Cantonese utterance.
Example 1
nei3 ho2 childish
you very childish
You are very childish.
Sometimes more English words are inserted but the overall morpho-syntactic structure remains Can-
tonese. Most of the linguistic evidences in this study are consistent with the findings in Chan (1998),
for instance, the English verb is usually in the infinitive without any tense or agreement markers, and it
is often inflected by Cantonese aspect markers. Other salient features include Cantonese classifiers
and sentence particles.
Example 2 presents an extracted conversation between two students A and B. They have been dis-
cussing the relatively light workload in a drama course and A starts to complain about the heavier
workload of another course Asian Voices.
A:
Asian Voices zik6cing4 hai6 soeng1faan2 wo3.

Asian Voices downright be opposite PART

But Asian Voices is just the opposite.

ngo5 jau5 sing4 sei3 joeng6 je5 wo3.

I have ADV four CL things PART

I even have four things.
B: haa2？ sei3 fan6 assignment aa4？

PART Four CL assignment PART

What? Four assignments?
A:
loeng5 go3 Present，loeng5 go3 present，loeng5 fan6 paper wo3.

Two CL Present, two CL present, two CL paper PART
(There are) two presentations, two presentations, (and) two papers.
These few utterances show a predominantly Cantonese morpho-syntactic structure, with Cantonese
classifiers (joeng6, fan6 and go3) and sentence particles (wo3 and aa4).
Examples 3 contain a higher lexical density of English words.
Example 3
nei3 tai2 haa3 aa3 sir go2 jat6 send go2 fung1 e-mail,
you look ADV SUF sir that day send that CL e-mail
Just take a look at the e-mail which the teacher sent that day,
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
ngo5dei6 go3 schedule paai4 dou3 gam3 mat6
I-PL CL schedule arrange ADV so close
(And you will see that) our schedule is so full.
In Example 3, the mixed words are all content words “sir”, “send”, “e-mail” and “schedule”. These are
constituents of English, the “embedded language”, but the morphological markers and grammatical
structures are Cantonese, the “matrix language” (Myers-Scotton, 1992). The phrase structure is Can-
tonese: English head nouns “e-mail” and “schedule” modified by Cantonese classifiers “fung1” and
“go3” respectively. A common Cantonese diminutive suffix for proper names aa3 is combined with the
English word “sir”. The grammatical structures are Cantonese: there is no tense marker for the verb
“send”. The relative clause in Chinese is left-branching and the adverbial phrase go2 jat 6 (that day) is
pre-verbal.
In some cases, distinct syntactic properties of both Cantonese and English are mixed in a sentence
consisting of several clauses, as Example 4 demonstrates.
Example 4
Mass mail ngo5dei6 mei6 send ge3,
Mass mail I-PL ASP send PART’
We have not yet sent mass mail,
jan1wai6 ne1, jyu4gwo2 send mass mail
because PART if send mass mail
because if (we) send mass mail,
ngo5dei6 zau6 zou6 ng4 cit3 ge3 zaam si4
I-PL ADV do not in time PART temporary
We are unable to handle it for the time being.
What distinguishes Cantonese, as well as Putonghua typologically from many other languages is that
it is topic-prominent. (Li and Thompson, 1976; Matthews and Yip, 1994). In Example 4, the first clause
does not follow the English sentence-predicate structure. The noun phrase “mass mail” which is the
direct object of the verb “send” is topicalized in the initial position, followed by the comment “we have
not yet sent”. The second clause has an English predicate “send mass mail” but lacks a subject which
in Cantonese syntax may be deleted. The third clause, however, shows an English syntactic influence.
The Chinese adverb zaam6 si4 which is normally pre-verbal follows the English post-verbal position.
Example 5 shows a stronger English mix at a higher level of the clause.
Example 5
mo5 man6 tai4 gaa3 as long as go2 gin6 si6 hai6 zan1 ge3
no problem PART as long as that CL matter be true PART
It is all right as long as that is true.
Cantonese and English share a similar syntactic structure of mo5 man6 tai4 “It is no problem.” and
go2 gin6 si6 hai6 zan1ge3 “That is true” but the main clause followed by the conditional clause of “as
long as” is a typically English structure.
5. Phrases
Over 80% of the phrases found in this sample are related to the students’ learning experience and
occur in interactions with their peers. Names of courses are quite common, e.g. “Chinese Phonology”,
“Contemporary Social Issues and Politics”, and “China and Asian Pacific Affairs”. Grammatically most
of them are noun phrases, e.g. “continuous assessment”, “discussion technique”, and “clear instruc-
tion”.
6. Incomplete sentences
An incomplete or minor sentence here refers to an utterance unit that does not conform to normal sen-
tence rules. It includes statements or questions that lack a verb and /or subject, e.g. “What?” and sub-
ordinate clauses that lack the main clause, e.g. “if he signs a contract”. All content clauses are not re-
corded from natural conversations. Instead they are quotations from the scripted texts which students
read and discussed in class. Most of the minor sentences recorded from daily conversations are dis-
course markers and exclamations such as “Hello”, “Bye-bye”, “Sorry”, “OK”, “that’s why”, “that means”.
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
7. Single full sentences and two-sentence Units
There are more interlocutor tokens (118) than subject tokens (56) which is explained by the fact that
many of the single sentences and two-sentence units were uttered by the English-speaking professors
in class. Some sentences uttered by students in English were re-enactments of what had been said in
class in a fun-sharing chat. Others were quotations in some English learning activities.
The foregoing analysis shows that despite the intricate integration of English elements into Cantonese
syntax, mix remains largely a kind of intra-sentential lexical insertion.
Code-Mixing: Socio-Cultural Meaning
Choice of Cantonese mixed with English sometimes entails a risk of alienation in intra-ethnic commu-
nication. Subjects of this study generally make a conscious effort not to code-mix when the interlocu-
tors’ educational level is lower than theirs and especially when the interlocutors are close to them, e.g.
parents, relatives and friends. This observation is supported by the minimal level of code-mixing in
most recorded conversations with interlocutors of these backgrounds. The major reason given is that
students do not want to be regarded as showing off or snobbish, as English is a prestige language in
the community and the social sanction against the use of English among Chinese in Hong Kong is
strong (Gibbons, 1987; Luke, 1998). An inappropriate extent of code-mixing is perceived to result in
alienation from the group one wishes to belong to. The following is a typical reflection:
Compare to my colleague, I use lesser English when communicate with my friends because not
all of them are university students. Since we have different background, I don’t want to act like I
am showing off because I am highly educated then they are. Therefore I use Cantonese most of
the time.
The caution to exclude English in private conversations is increased when the choice of Cantonese-
English poses a risk in jeopardizing a closer relationship:
When I talk to interlocutors with primary education only (like my parents and also my boyfriend’s
parents), I will talk to them in Cantonese only. I have to note that my parents receive primary
education in Mainland, so they did not learn English at that time. As my boyfriend’s education
level is up to Form 4 (in Chinese medium school) only, there is a considerable difference in the
codes used with him and my university classmates. When I speak to my boyfriend, I will try my
best to not to code mix. It is because I do not want him to have a hard feeling that we have some
distance psychologically, and he may think that I am showing off.
On the other hand, Cantonese-English mix is a strong binding force among educated bilinguals in
Hong Kong. It expresses a unique group solidarity and as a special register, facilitates easy communi-
cation among group members. It allows them to express ideas and feelings without fear of misunder-
standing and without the trouble of circumlocution or explanation in either English or Cantonese. Fo-
cus group interviews suggest that this is the most “comfortable” choice as pure English and pure Can-
tonese force them to speak cautiously. In speaking English, they are nervous about expressing them-
selves accurately and avoiding linguistic mistakes. In speaking Cantonese, they are worried about
inserting English words inadvertently and offending interlocutors who do not understand English.
Sometimes they find it a great mental burden to have to search for an equivalent in Cantonese and fail
to access it.
This tension results in an alertness in defining interpersonal relationships. As the foregoing discussion
reveals, Cantonese-English mix plays a paradoxical role in the bilingual’s social life. It alienates
him/her from those who are less educated, but integrates him/her into the speech community of the
educated. The decision to choose pure Cantonese or mix demands the individual to make judgments
about the interlocutor’s background and his/her relationship with self. As the language diaries show,
the bilinguals in Hong Kong have to make these choices rapidly and frequently. It may be argued that
these dynamic decision-making processes lead to greater emotional sensitivity and pragmatic aware-
ness.
Code choice does not only bear significance for the individual; it also indexes societal values and atti-
tudes. The use of English in Cantonese utterances delineates social stratification more clearly and
divides those with good education, great prestige and high social status from those without. While mix-
ing English is considered an act of snobbery when one uses it with those less educated than oneself
(especially those who are close to oneself), not mixing English when talking to educated people will
subject oneself to snobbery:
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
Whenever I spoke with the people I thought were highly educated, I tended to mix Cantonese
with English words in a sentence. It is because I believe that will give people a sense of being
educated and will not be looked down upon.
For this student, self-esteem is socially constructed out of a general attitude to English in Hong Kong.
English implies authority and professionalism and is highly regarded as the language of success in
employment and further studies (Li, 1999). It provides more opportunities for “upward” and “outward
mobility” (So, 2000). Therefore as “cultural and symbolic capital” (Chan, 2002), it is a very much
sought-after asset.
This explains some cases in which student’s code-mix with their parents. It has been reported that
their parents want them to use some English words despite their own limited knowledge of English.
The following example may provide a glimpse of the social motivation for code-mixing in working class
families:
My parents never receive higher education…. So one may suppose that I have to speak in pure
Cantonese with them otherwise they may feel bad about themselves because English in Hong
Kong is an identity marker for one’s education status. Nevertheless, I do not have to control my-
self to speak in Cantonese to avoid embarrassment as strictly as supposed to.

My parents have invested in a great deal into things that will possibly make me a truly bilingual
person e.g. giving me money to participate the international student exchange program, applying
for phone plan which includes long-distance call service for me to chat with my English-speaking
friends. They are more then happy to hear me speak in English as they want me to claim for a
higher social status, so it is necessary for me to demonstrate this ability to them because they
would like to see their investment getting returns.
The discourse of investment and return is interesting here. As noted in Ho (2002), at an early age,
Chinese are socialized into a cultural suprastructure of values and relational hierarchies, as well as
roles and responsibilities. Under the tenet of filial piety, children “owe” their existence and growth to
the love and care of their parents, and should “repay” them by studying hard and achieving academic
success. In the present case, the student’s parents have made a substantial financial sacrifice. They
have “invested heavily” in the future success of their daughter, which is a proof of their love and care.
This is a “debt” that she has to “repay”. The important gate-keeping function of English in Hong Kong
has turned mastery of the language an essential part of the repayment and “return” for parental in-
vestment. On the parents’ part, the return is not just the material gain for the daughter and possibly for
the family (as it is a traditional cultural practice for grown-up, salary-earning Chinese to support their
parents’ living). It also carries an interest with it, to continue the metaphor, which is generated by the
English-based upward mobility. Parents who come from the lower socio-economic strata find a great
sense of fulfillment, honour and prestige in their children’s success. This socio-cultural interpretation
of the investment discourse unravels the puzzling phenomenon of parental preference for their chil-
dren’s code-mixing behaviour which looks like an intrusion of a foreign element into a private sphere
of social life – the family whose activities are normally best conducted in the mother tongue for under-
standing, identity and intimacy. In a similar way, it explains why, despite the obvious pedagogical ad-
vantages of mother tongue education, parents make substantial sacrifices to gain their children’s ac-
cess to an EMI (English as Medium of Instruction) education (So, 2000, pp.20-22, Chan, 2002,
pp.275-279).
Conclusion
Cantonese – the native language of the majority plays an important role in maintaining cross-
generation and cross-class cohesion in society. Cantonese-English mix plays a paradoxical role of
alienating and integrating group members, depending on differences in educational level and English
competence. Linguistic analysis shows that code-mixing in Hong Kong remains a form of English lexi-
cal insertion in Cantonese speech. The high number of unit types of lexical items and phrases, i.e.
two-thirds of the total, implies some English vocabulary acquisition. Yet higher language standards
have been set for the educated youngsters in Hong Kong, and they would have to work harder to at-
tain a level beyond the lexical.
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
Acknowledgments
The author is grateful to Lingnan University, Hong Kong for granting a research fund (DA06A1) for this
project, and for granting her study leave to complete it. She wishes to thank all the students who have
participated in this research.
References
Auer, P. (ed.) (1998). Code-Switching in Conversation: Language, Interaction and Identity. London &
New York: Routledge.
Azuma, S. (1998). Meaning and form in code-switching. In Jacobson, R. (ed.) Codeswitching World-
wide, pp.109-123. Berlin/New York: Mouton de Gruyter.
Boeschoten, H. (1998). Codeswitching, codemixing, and code alternation: What a difference. In Ja-
cobson, R. (ed.) Codeswitching Worldwide, pp.15-24. Berlin/New York: Mouton de Gruyter.
Bolton, K. (2002). Introduction. In Bolton, K. (ed). Hong Kong English – Autonomy and Creativity, pp.1-
25. Hong Kong : Hong Kong University Press.
Chan, B. (1998). How does Cantonese-English code-mixing work? In Pennington, M. (ed.) Language
in Hong Kong at Century’s End, pp.191-216. Hong Kong: Hong Kong University Press.
Chan, E. (2002) Beyond pedagogy: Language and identity in post-colonial Hong Kong. British Journal
of Sociology of Education, 23(2), 271-285.
Gibbons, J. (1987) Code-Mixing and Code-Choice: A Hong Kong Case Study. Bath: Multilingual Mat-
ters.
Ho, J.W.Y. (2002) Narrative Writing in Australian and Chinese Schools: A Study of Text in Context.
Bern: Peter Lang.
Ho, J.W.Y. (2006) Functional complementarity between two languages in ICQ. International Journal
of Bilingualism. 10(4),429-451.
Jacobson, R. (ed.) (1998) Codeswitching Worldwide. Berlin/New York: Mouton de Gruyter.
Li, C. and Thompson, S. (1976) Subject and topic: A new typology of language. In Li, C. (ed.) Subject
and Topic, pp.457-489. New York: Academic Press.
Li, D. (1998) The Plight of the Purist. In Pennington, M. (ed.) Language in Hong Kong at Century’s
End, pp. 161-190. Hong Kong: Hong Kong University Press.
Li, D.C.S (1999) The functions and status of English in Hong Kong: A post-1997 update. English
World-Wide, 20(1), 67-110.
Li, D.C.S. (2000) Cantonese-English code-switching research in Hong Kong: a Y2K review. World
Englishes, 19(3), 305-322.
Luke, K.K. (1998) Why two languages might be better than one: Motivations of language mixing in
Hong Kong. In Pennington, M. (ed.) Language in Hong Kong at Century’s End, pp.145-159.
Hong Kong: Hong Kong University Press.
Luke ,K.K. and Richards, J.C. (1982) English in Hong Kong: functions and status.
English World-wide, 3(1), 47-64.
Lüdi, G. (2003) Code-switching and unbalanced bilingualism. In Dewaele, J., Housen, A., and Li,
W.(eds.) Bilingualism: Beyond Basic Principles, pp.174-188. Clevedon: Multilingual Matters.
Matthews, S. and Yip, V. (1994) Cantonese: A Comprehensive Grammar. London and New York:
Routledge.
Myers-Scotton, C. (1992) Constructing the frame in intrasentential codeswitching. Multilingua, 11, 101-
127.
Myers-Scotton, C. (1993a) Duelling Languages: Grammatical Structures in Codeswitching. Oxford:
Clarendon Press; New York: Oxford University Press.
© LSC-2007
Issue 21

---

The International Journal of Language Society and Culture

Editors: Thao Lê and Quynh Lê
URL: www.educ.utas.edu.au/users/tle/JOURNAL/
ISSN 1327-774X
Myers-Scotton, C. (1993b) Social Motivations for Codeswitching. Oxford: Clarendon Press; New York:
Oxford University Press.
Pennington, M. (ed.) (1998a) Language in Hong Kong at Century’s End. Hong Kong: Hong Kong Uni-
versity Press.
Pennington, M. (1998b) Colonialism’s aftermath in Asia: a snapshot view of bilingualism in Hong
Kong. Hong Kong Journal of Applied Linguistics. 3(1), 1-16.
So, D.W.C. (2000) Achieving biliteracy and trilingualism without MOI-based bifurcation of the schools:
A plea for third alternatives. In Li, D.C.S, Lin, A. and Tsang W.K. (eds.) Language and Educa-
tion in Postcolonial Hong Kong, pp.9-33. Hong Kong: Linguistic Society of Hong Kong.
© LSC-2007
Issue 21

---