Available online at www.sciencedirect.

com

System 37 (2009) 719–730

www.elsevier.com/locate/system

Codeswitching in the primary EFL classroom

in China – Two case studies
Xiaofang Qian *, Guisen Tian, Qiang Wang
School of Foreign Languages and Literatures, Beijing Normal University, Beijing 100875, China

Received 4 February 2009; received in revised form 26 May 2009; accepted 29 May 2009

Abstract

This paper presents the results of a small-scale study of codeswitching (CS) between Chinese and English in primary
English classrooms with a view to informing classroom interaction with young learners of English as a Foreign Language
(EFL). It analyzes CS of two teachers participating in the Primary English Curriculum Innovation (PECI) project in Bei-
jing, which adopts a holistic approach to innovation and implementation of curriculum. The data consist of 20 videoed
lessons, covering lessons from grade 1 to grade 4. The findings comprise syntactical identification of the switches and
the pedagogical and social functions that these switches serve.
The conclusions are: (1) CS is a discourse strategy that teachers use for promoting classroom interaction and ensuring
efficient classroom management. (2) A suitable quantity of CS use helps cultivate and reinforce good habits of learning and
foster a close student-teacher relationship. (3) Teachers have the ability to instruct in the target language; only when occa-
sions call for efficient instruction do they turn to L1. (4) The decrease of CS over the years proves the efficiency of a holistic
approach to curriculum innovations in the fulfillment of education objectives at the time of curriculum change.
Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Codeswitching; Young learners; Teacher talk; Classroom interaction; Foreign language learning

1. Introduction

Codeswitching (CS) is a common phenomenon of language contact in bilingual, multilingual and even
monolingual societies. It is generally understood as ‘‘the alternative use by bilinguals (or multilinguals) of
two or more languages in the same conversation” (Muysken, 1995, p. 7), or ‘‘in the unchanged setting, often
within the same utterance” (Bullock and Toribio, 2009, p. 2). Myers-Scotton (1997) indicates that codes
involved in CS are not necessarily those of standard languages and expands the term to subsume switching
between languages (Azuma, 2001; Platt and Webber, 1980; Treffers-Daller, 1998, etc.), dialects (Alfonzetti,

*
Corresponding author. Tel.: +86 10 5880 5140, mobile: +86 1364 1374 226; fax: +86 10 5880 9393.
E-mail addresses: amandaqxf@sina.com, amandaqxf@yahoo.com (X. Qian), tiangs@263.net (G. Tian), wang_qiang99@yahoo.com
(Q. Wang).

0346-251X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.system.2009.09.015
720 X. Qian et al. / System 37 (2009) 719–730

1998; Blom and Gumperz, 1972; Scotton and Yule, 1977, etc.), styles and registers (Farris, 1992). In this paper,
‘‘code” is used to refer to two distinct languages, English and Chinese.
Discussion of codeswitching in the classroom context in China draws data mostly from classrooms of
adults or secondary levels while codeswitching in classrooms of young learners is under-explored. Do teachers
of young learners engage in CS behaviors? And if they switch, what purposes and what functions do their
switches serve? Are they beneficial or detrimental to students’ language learning? By drawing on data from
two teachers’ classroom talk collected from a curriculum innovation project over 4 years’ time, we hope to
find answers to these questions. We also attempt to describe and explain CS in the primary EFL classroom
in the context of China.

2. Literature review

A large number of classroom CS studies, as opposed to social CS or general CS in no specific context, cen-
ter around functions and motivations for juxtapositions of two languages. They have been influenced by
research in classroom interaction, second language acquisition, teacher talk, conversational analysis, pragmat-
ics and the ethnography of communication. Many researchers devote their attention to bilingual classroom CS
of various types, while only a few study CS in the foreign language classroom. CS in the foreign language is a
more complicated issue, as the foreign language is both the means and the end of such classrooms.

2.1. Types of CS

Based on observations of many cases of CS, Sankoff and Poplack (1981) identify three types of CS syntac-
tically, namely, tag-switching, intra-sentential switching and inter-sentential switching. Tag-switching,
emblematic switching or extra-sentential switching (Muysken, 1995), involves the insertion of a tag or a short
fixed phrase in one language into an utterance which is otherwise entirely in the other language. Inter-senten-
tial switching involves a switch at a clause or sentence boundary, where each clause or sentence is either in one
language or the other. It may also occur when one speaker takes up where another leaves off. Intra-sentential
switching refers to switching within the clause or sentence boundary. This form involves the greatest syntactic
risk and requires that the speaker be fluent in both languages.

2.2. Development of classroom CS research

According to the understanding of Simon (2001), there were three phases of classroom CS summarized in
Martin-Jones (1995), shifting from the distribution of L1 and FL to the micro-ethnographic study of class-
room discourse. Early studies on CS, the first phase, attracted attention in the bilingual education programs
for linguistic minority pupils in the United States in the 1970’s and early 1980’s. These studies were quantita-
tive in nature and focused on the influence of CS in bilingual classroom communication on children’s linguistic
development. The second phase originated from the research conducted by Milk (1981, 1982) and Guthrie
(1984), both cited in Martin-Jones (1995, p. 92). They first broke through the purely quantitative study of
classroom communication by using audio-recordings and descriptive framework. They started to pay atten-
tion to the ways in which teachers and students fulfill tasks with two languages. Values conveyed through code
choice also became more significant for them. Guthrie’s (1984, p. 45) study identified several communicative
functions of L1 usage in the classrooms: (1) for translation; (2) as a ‘‘we” code; (3) for procedures and direc-
tions; (4) for clarification; and (5) to check understanding.
The third phase was ushered in by the distinctly ethnographic approach adopted originally by Zentella
(1981) and Erikson et al. (1983) in Martin-Jones (1995, p. 95). Within this approach, the researchers were
engaged in detailed situational analysis of teaching/learning events with regard to both the linguistic and cul-
tural background of the participants. Zentella (1981) also identified the pragmatic functions of CS, namely, to
alleviate the effect of admonition, to make asides, and to make metalinguistic commentaries. This identifica-
tion, like the one in Guthrie (1984), does not distinguish between communicative acts realized through code
choices for pedagogical or social reasons. Merritt et al.’s (1992) study in Kenyan primary classrooms found
 X. Qian et al. / System 37 (2009) 719–730 721

that teachers’ switching often provided a resource to focus or regain students’ attention, or to clarify, enhance,
or reinforce lesson material. Merritt also identified four syntactic types of CS. The first type consists of refor-
mulation across codes, with no new information and no new instructions. The second type involves codes-
witching as the content of the activity or the textual instruction is moving along. These two types often
involve a whole sentence or interactional move and are similar to inter-sentential switching as said above.
The third type consists of translation or word substitution within a sentence and mainly involve intra-senten-
tial switching. The last consists of interactional particles, including discourse markers, classroom management
routines, and terms of address and has commonalities with tag-switching
The Conversational Approach to CS as advocated by Peter Auer (e.g. Auer, 1998) suggests two ways
that CS can be used: discourse-related switching and participant-related switching (e.g. Liebscher and
O’Cain, 2005). In the classroom, discourse-related switching serves as a resource for accomplishing different
communicative acts, like marking topic changes. Participant-related switching takes into consideration the
hearer’s linguistic preferences or competence. Such switching is likely to happen in some contexts, for
instance, when the learners have great difficulty in understanding the teacher’s instructional language or
the target language.

2.3. Foreign language classroom as a specific CS context

Many of the studies cited above involve bilingual or multilingual classroom discourse. The foreign language
classroom may be considered ‘‘a microcosm of the community outside the classroom in terms of communica-
tion, but that it presents specific features Simon (2001, p. 314)”. Communication in a foreign language class-
room is multi-layered and more difficult to analyze. Simon also builds up a model which accounts for the
social dimension of CS, as well as the pedagogic dimension, for the primary goal of communication in this
context is to facilitate learning, and learning the foreign language in particular.
Many CS studies in China are situated in classrooms of secondary and tertiary education (e.g. Chen, 2004),
while few have investigated the foreign language classrooms of young learners. Codeswitching in young learn-
ers’ FL classrooms may have its own characteristics. The present study draws from previous researches and
aims to find out the patterns of L1 and TL distribution in teacher talk and the functions that teachers’ CS
serves in primary English.

3. Methodology

3.1. Research context

This research is a case study of two teachers participating in the Primary English Curriculum Innovation
(PECI) project in Beijing, China, which adopts a holistic approach to the innovation and implementation of
curriculum. It is a six year project spanning the years 2003–2009, headed by a university research institute, in
collaboration with a non-profit private educational institute. Altogether seven primary schools with approx-
imate 1500 pupils and 30 primary teachers participate in it (Wang et al., forthcoming).
For this study we adopt an approach based on conversational analysis. We pay great attention to the point
within the interaction episode where CS takes place, considering where in a sequence of actions switches occur
and their relevance to the turn-taking system.

3.2. Research questions

We intend to describe the codeswitching patterns occurring in teachers’ utterances and explain the functions
of switches in the primary EFL classroom setting. We address three questions:

(1) What types of CS are there in teachers’ talk in the primary English classroom?
(2) Is there any change in the quantity of teachers’ CS as students advance from one level to the next?
(3) What functions does teachers’ CS serve in classroom interaction?
722 X. Qian et al. / System 37 (2009) 719–730

3.3. Participants

The participants of the study are two young female teachers, May and Lily. They were selected because they
had participated in the innovative project for more than four years. Their lessons were periodically recorded
for studying various aspects of teaching and learning during the innovation process. Both Lily and May
worked as teachers in a primary school in Beijing, which has gained prestige since the PECI project was under-
taken there. Lily was a full-time teacher at the school but May was also a member of the research institute.
May had eight years’ experience of teaching primary English, and Lily had taught primary English for 11
years. The number of students in their classes was between 30 and 40, quite large for language learning pur-
poses. Throughout the project, they took part in teacher training activities like observation, discussion and
reflection of public or routine lessons, and mini-research workshops. Both were conscientious and excellent
teachers.

3.4. Materials

The materials of this study are the videotapes of both teachers. The lessons were videotaped by the project
staff. Ten of each of their lessons were chosen for this study, ranging from year one to year four, mainly gen-
eral English lessons, where there is a greater variety of classroom activities and more teacher–student interac-
tions going on, though there were other types of videoed lessons such as audio-visual lessons and reading
lessons. Each recorded lesson is about 40 minutes long.

3.5. Procedures

The whole lessons are transcribed. To get a rough estimate of the distribution of the two languages, word
count, an instrument in Microsoft WordÒ, is applied. Then, the unit of analysis is identified, that is, a turn of
speaking in teacher–student interaction. The units are further assigned to a subcategory including whole target
language (TL) units, L1 units and mixed units. Following Kim and Elder (2005), a functional language alter-
ation analysis framework is applied to identify the functions of the switches in the teacher talk. As elements of
a sequential flow, the switched units are coded in light of context.

4. Findings and discussion

The findings comprise syntactical analysis of the switches between English and Chinese and the pedagogical
and social functions these switches serve.

4.1. Much more inter-sentential switching than tag or intra-sentential switching

As the CS distribution pattern of both teachers is quite similar, the units consisting of mixed varieties of the
two teachers’ switches are combined for comparing the three types of switch patterns. As shown from Table 1,
among the three subcategories, inter-sentential switching (82%) far outweighs the tag-switching (2%) and
intra-sentential switching (16%).
Example 1 (Lily, Year 3).

T: Ok, at last, woman bees, ha, ha, they didn’t get, go out to work, but zen me yang?
(. . . but then what?)

Table 1
Types of switching in the teacher talk of both teachers.
Types Tag-switching Intra-sentential Inter-sentential
Frequency (%) 2 16 82
 X. Qian et al. / System 37 (2009) 719–730 723

S5: But they eat

T: But they eat the worker bees’ honey, so the worker bees ge, were getting mad, angry. Ok, very good
story. How about others?

Tag-switching involves discourse markers or sentence fillers, as in Example 1. The teacher used a tag zen me
yang (then what) to carry on questioning. In this story-telling class, one student (the fifth one to share stories
she made up herself) told a story of bees, in which the worker bees go out to collect honey and the counter bees
count the baskets of honey they collected. When they are about to have a party, they find the women bees are
eating the honey. Lily was amused by the story and asked the student to confirm what happened in it in Chi-
nese. This type of switching tends to be discursive; their occurrences are more of an impulsive nature and exist
when teachers are not ‘‘guarding” their choice of language for the sake of linguistic input.
Intra-sentential switching involves CS within the clause boundary, as shown in Example 2, where the tea-
cher inserted English words in her demonstration, which was predominately in Chinese to the first year pupils.
The English words inserted are always the target vocabulary to be learned. This is typical in primary English
teachers’ classroom instructions of lower grades.
In example 3, however, the target language learning unit was not just words; it was a sentence.
Example 2 (May, Year 1).

T: nimen huide dongzuo you shenme a? you march, clap, jump

(What action can you do, like march, clap, jump)

Example 3 (Lily, Year 1).

T: ruguo ni xiang dang shenme, ni jiu zhanqilai shuo I’m a police officer, I can help you.
(If you want to be someone, you then stand up and say.)

Inter-sentential switching occurs at the sentence boundary, or among sentences. In Example 4, the teacher
shifted her praise from English to Chinese. She repeated the praise and encouraged students to display their
work on the blackboard. The praise was a little difficult to be understood in English but was very meaningful.
By switching to the more familiar code, the teachers passed on the message efficiently, and the video shows
that the students were very happy to hear this.
Example 4 (Lily, Year 1).

T: OK, very good, very nice, very nice, feichang bang, xiake yihou ba ta tiezai heiban shangmian, qita tongxue
keyi kan.
(I’ll stick them [flash cards] on the blackboard and other students can have a look.)

Inter-sentential switching always carries more social messages, as in Example 5. When teaching words of
occupations, Lily asked students about their future ambitions and encouraged them to work hard to make
them come true. In this sense, the language classroom is not only the place where students learn a foreign lan-
guage; it is also the space where teachers educate their students in social conduct, arts, etc. Such interaction
seems divergent from the main task of language teaching, but it fulfills the role of education.
Example 5 (May, Year 2).

T: Now you’ll be . . . jianglai nimen hui zenmeyang a? haohao xuexi, jianglai caineng zuo yisheng.
(What will you be in the future? Work hard, and then you may become a doctor.)

To sum up, teachers in this study were found to engage in tag-switching, intra-sentential switching and
inter-sentential switching. Inter-sentential switching has a higher frequency of occurrences, which could be
attributed to teachers’ intention of giving clearer instruction and eliciting more responses from students.
724 X. Qian et al. / System 37 (2009) 719–730

4.2. Drastic decrease of CS over the years

It is found that both teachers switched much more frequently in the first year, and the ratio of Chinese
utterances in one single lesson reached slightly more than 40% (as shown in Table 2). But the second year
saw a dramatic decrease of L1 use in lessons, the ratio dropped to below 10%. In the last two years, the
use of L1 was at a minimum. It needs to be made clear that ‘‘the one lesson” in Table 2 is the particular lesson
with the most occurrences of CS, a lesson representative of the year. The numbers of utterances are not means
of each year, either.
There is no significant change in the distribution of the three types of switching throughout the years, so no
detailed statistical analysis of the types was taken into special account.
Throughout the four years, the teacher talk in both teachers’ lessons had ‘‘evolved” towards minimal use
(as indicated in Chart 1). The dynamics and enthusiasm in the classroom, however, remained unchanged when
students reached higher levels, as shown from the whole-class videos. The smoothness and ease of teacher–stu-
dent communication in ‘‘whole” English can be easily detected. In one reading lesson, for example, the stu-
dents not only answered questions in English but also asked good questions in English. For instance, the
4th year pupils asked questions like ‘‘How long is the biggest lizard?”, ‘‘What is[does] he eat?”, ‘‘en, en, that,
on the head, it has two hole [holes], what are they?” when Lily tried to arouse their curiosity by asking them
what they wanted to know about lizards before she showed the whole story to them. The last question was
commented as ‘‘Good question!” by many students, in response to it, for it was more difficult.

4.3. Functions of CS in teacher talk

Teachers are often found to use CS as both a methodological and a social strategy in classroom interaction
(Simon, 2001). It is sometimes nevertheless hard to distinguish between the methodological and social func-
tions of some switches. This is particularly apparent in teacher talk to young learners.

Table 2
Ratio of Chinese utterances in one lesson over the years.
Year Case Words total Chinese Characters English words Percentage (C/total)
1st May 2623 1113 1510 42.4
Lily 2897 1175 1722 40.6
2nd May 2499 289 2210 11.9
Lily 3024 249 2775 8.2
3rd May 2856 70 2786 2.5
Lily 4405 173 4232 3.9
4th May 1624 8 1616 0.5
Lily 3841 16 3825 0.4

45.00%
40.00%
May
35.00%
Lily
30.00%
25.00%
20.00%
15.00%
10.00%
5.00%
0.00%
1 2 3 4
Chart 1. Distribution of CS of each teacher.
 X. Qian et al. / System 37 (2009) 719–730 725

4.3.1. Methodological functions of CS

Both May and Lily switched to Chinese to ensure maximum effect of teaching, in contrast with the intention
to adjust teacher–student relationship via CS. They used it for translation, which generally involved intra-sen-
tential switching. They clarified or highlighted in L1 to draw students’ attention. When the occasion called for
efficient communication, they switched to the L1. Inter-sentential switching would then become more preva-
lent in their speech.

4.3.1.1. For translation. Lily used CS as a means to ensure students’ understanding by means of translation
(Guthrie, 1984) as in Example 6, where she translated into Chinese students’ response ‘‘It’s not fair” to what
happened in a story so as to reinforce their judgment. Teachers do have other ways of explaining such phrases,
like miming or paraphrasing, but translation is no less effective than those methods. In another class, when
telling stories, Lily used a few words unfamiliar to students; she asked ‘‘What’s the meaning of it”. Students
then came up with a Chinese translation. This happened 4 times in that class, but the translations were the
only L1 utterances detected in the 40-min lesson.
Example 6 (Lily, year 3).

T: . . .Did you remember? In Bob’s story?

[Ss had no response]
T: It’s. . .?
Ss: It’s not fair.
T: Not fair. tai bu gongping le, why? Because worker bees?
(It’s not fair)
Ss: The worker bees, worker bees take the honey.
S7: The worker bees find the honey.

4.3.1.2. For clarification. Both teachers used CS to make things clear to understand, as Chinese is the code
shared by everyone and much easier to understand. In Example 7, May asked about the meaning of ‘‘pilot”,
and students responded with a translation. It is worth noting that May clarified in Chinese and repeated the
definition partly in TL, which differed from switches in the first year, when she basically offered L1 equivalents
and did not switch back to English. Students had not been able to understand definitions in English at that
time, while now they could. She also intended to build up students’ capability of comprehending words in Eng-
lish. In the videoed lessons of year 4, in fact, explanation or clarification of vocabulary was always in TL in the
lessons of both teachers.
Example 7 (May, year 2).

T: Tom wants to be a pilot. Do you know a pilot?

Ss: feixingyuan. (pilot)
T: jiashi feiji de, fly a plane. (Who flies a plane)

4.3.1.3. For highlighting. CS was also a means applied by both teachers to highlight key learning points or
important task requirements, that is, to get students’ attention or to emphasize a point which teachers consid-
ered significant. In Example 8, Lily invited one more pair of students to present their paired work ‘‘I Ask You
Answer” to the class. She emphasized that one student would act as the teacher, the other as the student.
Example 8 (Lily, year 2).

T: next one, I want a teacher, he is a teacher. yao wende tebie shulian, lianxu wen wuge wenti.
(You must be very skilful in asking the questions, five questions altogether.)

In this example the teacher emphasized the requirement in L1, for this was crucial to the activity, which was
also a group competition. Teachers seemed to be constantly worried that students might not be able to
726 X. Qian et al. / System 37 (2009) 719–730

understand them clearly, so they often resorted to students’ L1 to draw their attention. If this was overdone,
however, students would have less chance of getting TL input. However, these two teachers seemed to have
mastered the ‘‘art” of using L1 properly in instruction; they skilfully channelled their students to total English
instruction and interaction.

4.3.1.4. For efficiency. When speakers are in pressing situations, they may switch to their native language to
express their ideas. They need to save time and effort and ensure the optimal effect of communication. Teach-
ers are no exception in this regard. A 40-min lesson can be an especially pressing situation as teachers might be
hard pressed to complete a variety of tasks. This is particularly true of Example 9, when the bell rang but the
task remained half completed. It was the last lesson of the unit and the final session before a 7-day holiday.
The bell rang but May had not talked about homework yet. Homework constitutes an important part of
schoolwork in China. In this case May asked pupils to look for colors and shapes in and around their home
as a way of reviewing those words. The only practical way to regain the impatient little pupils’ attention (Mer-
ritt et al., 1992) and finish the assignment was switching to L1. Here May not only shifted to Chinese, but she
also spoke much faster.
Example 9 (May, year 1).

T: Ok, sit down. haole, xianzai tinghao, laoshi shuo zuoye a.

(ok, now listen. It’s about homework.)
S1: laoshi daling le.
(Miss, the bell rings.)
T: wo zhidao le, cong jintian kaishi, wo ba zuoye gaosu ni yibian, huijia yihou ni gen baba mama shuo zuoye a. . .
(I know it. From today, I’ll tell you about homework, after go home you tell your mum and dad about your
homework. . .)

4.3.2. Social functions of CS

The analyses above center around functions and motivations of CS from methodological considerations.
This is where classroom CS differs from social CS, as proved by Simon (2001). In the former cases, teachers
are more concerned with the teaching effect, but CS instances can also imply social effects, in terms of increas-
ing or decreasing social distance (Myers-Scotton, 1997). Teachers use CS as a strategy to establish or reestab-
lish certain relationships, to strengthen solidarity or authority. Here the similarities that classroom CS has
with social CS become more evident.

4.3.2.1. For praise. Both teachers are found to have used CS as a strategy to praise their students. For instance,
as in Example 10, May praised a student whose English name was Fiona for the words she remembered and
her performance in that class. To make sure the student would not miss any part of the praise, she turned to
L1. The video showed that Fiona was quite pleased by it. By using L1 the maximal effect of teachers’ com-
ments are achieved, which is beneficial to classroom rapport and student–teacher relationship.
Example 10 (May, year 1).

T: Fiona
Fiona: Kitchen, bathroom
T: Ok, bathroom jintian biaoxian feichang hao a
(you did very well today.)

4.3.2.2. For encouragement. Similarly, CS is also often applied to encourage students, especially when they are
faced with difficult tasks. Teachers encourage their students, and weak students in particular, in L1 in order to
sound less intimidating. In Example 11, with a gentle touch, Lily encouraged the student to say more, as she
was standing in the front of the classroom and was probably a little scared. Compared with TL, L1 could bet-
ter achieve solidarity.
 X. Qian et al. / System 37 (2009) 719–730 727

Example 11 (Lily, year 2).

[S2 came to the front; T helped her to put the picture on the projector]
S2: My story title is: It’s Not Fair.
T: En, yes, very good. ta hai jieshao yixia ta de shenme a, title. It’s Not Fair. Do you know what’s the mean-
ing? [T touched S2]
(What else did he just introduce about him?)

4.3.2.3. For disapproval. When teachers are dissatisfied with students’ behavior or performances, they tactfully
use L1 to express their disapproval. As English is the foreign language and sounds more distanced, the teacher
uses the code that both parties share, or the ‘‘we” code (Guthrie, 1984). In Example 12, for instance, students
were about to be engaged in guessing games in groups. They held different pictures of people and places and
could not look at each other’s. While walking around, May noticed that one pair of students intended to swap
pictures. She immediately switched to Chinese and prevented them from doing so.

Example 12 (May, Year 2).

T: Good picture, face down. Face down, ok, Amy’s group is the first, and second group. Put the pen in the
box, Lucy, put the pen in the box. Which group is the first?
[Two Ss tried to change their seats]
T: Buxing, nimen xianzai buneng huan. Yihuier huan hui gaomi de. Xian zheme zuozhe ba. Yes, um, this
group.
(No, you cannot exchange now. If you do, you’ll let out the secret. Sit like this for now.)

4.3.3. Multiple functions of CS

The functions of CS have been classified into the two general categories, and it is obvious that they overlap
in many instances. It has to be admitted that the methodological functions are not absolute; they are interre-
lated. As a matter of fact, in one passage of speech the speaker might have integrated both methodological and
social motivations.

4.3.3.1. Kind reminder and efficiency. One single switch can be a kind reminder, and it saves time. Apart from
learning TL, students’ emotions have to be considered. Lily (in Example 13) introduced new vocabulary
accompanied with actions, and tried to elicit words learned previously from students. After several repetitions
of the word ‘‘sleep”, students were unable to produce any words related to ‘‘sleep”, as they were not quite sure
of the teacher’s intention. Then she reminded them in Chinese. The reminder is a friendly prompt and is effi-
cient, for the student produces the desired word.

Example 13 (May, year 1).

[The teacher mimed the action of sleeping by putting both hands beside her cheeks.]
Ss: Sleep.
T: Sleep.
Ss: Sleep.
T: Sleep.
Ss: Sleep.
T: Sleep.
T: na ni xiangqi shenme lai le?
(Then, what comes into your mind?)
Ss: Bedroom
T: Ok, good, bedroom
728 X. Qian et al. / System 37 (2009) 719–730

4.3.3.2. Kind reminder and highlighting. In Example 14 May invited individual students to tell stories to the
class based on the drawing they had completed at home. They needed to show the picture, tell the title they
had made up, and invite fellow students to answer questions like: ‘‘What can you see in the picture?” They
jointly told the story. Two girls had done this very well, but when Philip came to the front, his story ‘‘Which
one?” seemed to have perplexed his peers due to the light coloring of his pictures. May stepped in and kindly
reminded him. As this point was important in completing the task, the L1 use rightly served the purpose of
drawing their attention and warning them against such flaws.
Example 14 (May, Year 2).

T: Ok, xiaci yao hua zhong yidian.Thank you. Ok, now, let’s stop.
(OK, next time draw with more force.)

4.3.3.3. Kind reminder, praise and highlighting. Teachers can expertly ‘‘kill three birds with one stone”. Their
purposeful use of one single act of switching can serve three purposes, as in Example 15. May was reviewing
vocabulary of family members and phrases like ‘‘in the living room”. She asked students to repeat the phrases
and do the action of putting family member cards on the corresponding room picture. Then she noticed that a
girl did not have any cards with her. She reminded her to bring her cards next time and praised her for her
gesture of pointing to the picture. Such a gentle reminder in L1 can leave a deep impression on students’
minds.
Example 15 (May, year 1).

T: ta meiyou dai kapian, danshi, ta youdian manhao, laoshi shuo put the grandmother in the bedroom, ta
mashang jiu zhidao BEDROOM, rang laoshi kandao, ta huizuo, xiaci dai kapian, haoma? danshi jintian zhege
shoushi henhao.
T: She didn’t bring her card today, but there’s one thing she did well. The teacher said ‘‘Put the grand-
mother in the bedroom”, she then pointed to the BEDROOM to let me see it. She can do it, and bring
the card next time, will you? But today the pointing gesture is very good.

4.4. Summary

The two teachers in this research do not show significant differences in the CS patterns in their talk, whether
in the syntactic types or the functions of their switches. The similarities of their switching are in fact astonish-
ing, especially when their students move up from year 1 to year 4. Their switching also differs from the switch-
ing in the teacher talk of more advanced learners in that they seldom switch to L1 for quantification of the
message or for some social effects like humor. Both discourse-related and participant-related switching were
present in their utterances (Liebscher and O’Cain, 2005).

5. Conclusions and implications

We can reach the following conclusions from the analysis and discussion above.
First, CS is a readily available and frequently applied strategy for teachers in classroom interaction and
classroom management. The classroom is a special context, but it is also an extremely important social setting
for every student’s growing-up and socialization. In such a setting, teachers, the most powerful participants,
will use every means to fulfill their duty of educating their students and ensuring the smoothness of commu-
nication. Codeswitching is an available resource for bilingual speakers and hearers, in the foreign language
classroom context as well. It is even more congenial to use CS as a discourse strategy in such contexts where
learners and teachers have homogeneous ethnic and linguistics backgrounds, as apart from some L2 class-
rooms where students’ and teachers’ backgrounds are more diverse and the target language is the only lingua
franca. This point may have some relevance to the issue of instructional language choice in teacher education.
 X. Qian et al. / System 37 (2009) 719–730 729

Second, it is obvious that teachers’ CS does not impede language acquisition of young learners. A prudent
use of it helps cultivate and reinforce good habits of learning for students and foster healthy and close rela-
tionship, especially for the lower levels. It is also essential to cultivate good learning habits and routines in
young learners, right from the very beginning of FL learning. Sometimes simple explanations in L1, accom-
panied with non-verbal hints, can serve the purpose, but other times it would be cumbersome and futile to
explain a word or highlight a point in TL. This is another characteristic of FL learning where the learners’
cognitive levels are far beyond their foreign language level.
Third, we can assume that the teachers had the ability to instruct mostly in TL or in mixed code. Only when
occasions call for efficient instruction in the classroom do they turn to L1 to communicate the message, or to
maintain a favorable social relationship. But it should be noted by teachers or teacher trainers that only lim-
ited or selected use of L1, or well-contemplated CS, should be applied in the classroom. Teachers should not
habitually switch to L1 to explain themselves whenever there seem to be obstacles.
Last, the drastic decrease of CS from level to level in the activity of both teachers also provides strong evi-
dence of the success and impact of the PECI Project. It also proves that the holistic approach to curriculum
innovations has great advantages in the fulfillment of the education objectives at the time of curriculum
change. The changes in language behavior in teachers do not happen all at once. They go hand in hand with
the development of students’ language abilities. In other words, teachers are able to adapt to students’ lan-
guage capacity. This ability should be emphasized in teacher training.
The limitation of the study is the lack of evidence from the teachers’ perspectives. Interviews of the teachers
would have convincingly demonstrated the motives and functions of their switches. This deficiency is due to
the nature of the research. As it is longitudinal, it is difficult to ask teachers to recall their motives. Further
study could be done right from the fifth year to complement the motivational studies.

Acknowledgements

The authors wish to thank Beijing Qianyan Education Research Institute for their kindest support through-
out the study, Dr. Andrew Thornton of Saint Anselm College for his aid in polishing the paper, and the
reviewers for their insightful comments.

References

Alfonzetti, G., 1998. The conversational dimension in code-switching between Italian and dialect in Sicily. In: Auer, P. (Ed.),
Codeswitching in Conversation: Language, Interaction and Identity. Routledge, London, pp. 180–211.
Auer, P., 1998. Introduction: bilingual conversation revisited. In: Auer, P. (Ed.), Codeswitching in Conversation: Language, Interaction
and Identity. Rutledge, London, pp. 1–24.
Azuma, S., 2001. Functional categories and codeswitching in Japanese/English. In: Jacobson, R. (Ed.), Codeswitching Worldwide II, . In:
Trends in Linguistics: Studies and Monographs, vol. 126. Mouton de Gruyter, Berlin, pp. 109–124.
Blom, P., Gumperz, J., 1972. Social meaning in linguistic structure: code-switching in Norway. In: Gumperz, J.J., Hymes, D. (Eds.),
Directions in Sociolinguitics: The Ethnography of Communication. Basil Blackwell, Cambridge, pp. 407–434.
Bullock, B.E., Toribio, A.J., 2009. Themes in the study of code-switching. In: Bullock, B.E., Toribio, A.J. (Eds.), The Cambridge
Handbook of Linguistic Code-switching. Cambridge University Press, Cambridge, pp. 1–17.
Chen, L., 2004. A study of codeswitching in the classroom of teachers of English majors. Journal of PLA University of Foreign Languages
27 (5), 34–40.
Erikson, F., Cazden, C.B., Carrasco, R., Maldonado-Guzman, A., 1983. Social and Cultural Organization of Interaction in Classrooms of
Bilingual Children. Final Report to the NIE.
Farris, S., 1992. Chinese preschool codeswitching: Mandarin babytalk and the voice of authority. In: Eastman, M. (Ed.), Codeswitching.
Multilingual Matters Ltd., Claredon, pp. 187–213.
Guthrie, L.F., 1984. Contrasts in teachers’ language use in a Chinese–English bilingual classroom. In: Handscombe, J., Orem, R.A.,
Taylor, B.P. (Eds.), On TESOL 1983: The Question of Control. TESOL, Washington, DC.
Kim, S., Elder, C., 2005. Language choices and pedagogic functions in the foreign language classroom: a cross-linguistic functional
analysis of teacher talk. Language Teaching Research 9 (4), 355–380.
Liebscher, G., O’Cain, D., 2005. Learner code-switching in the content-based foreign language classroom. The Modern Language Journal
89, 234–247.
Martin-Jones, M., 1995. Code-switching in the classroom: two decades of research. In: Milroy, L., Muysken, P. (Eds.), One Speaker, Two
Languages: Cross-disciplinary Perspectives on Code-switching. Cambridge University Press, New York, pp. 90–111.
730 X. Qian et al. / System 37 (2009) 719–730

Merritt, M., Cleghorn, A., Abagi, J., Bunyi, G., 1992. Socialising multiligualism: determinants of codeswitching in Kenyan primary
classrooms. In: Eastman, M. (Ed.), Codeswitching. Multilingual Matters Ltd., Claredon, pp. 103–121.
Milk, R., 1981. An analysis of the functional allocation of Spanish and English in a bilingual classroom. California Association for
Bilingual Education: Research Journal 2, 11–26.
Milk, R., 1982. Language use in bilingual classrooms: two case studies. In: Hines, M., Rutherford, W. (Eds.), On TESOL ’81. TESOL,
Washington, DC, pp. 181–191.
Muysken, P., 1995. Code-switching and grammatical theory. In: Milroy, L., Muysken, P. (Eds.), One Speaker, Two Languages: Cross-
disciplinary Perspectives on Code-switching. Cambridge University Press, New York, pp. 177–197.
Myers-Scotton, Carol, 1997. Code-switching. In: Coulmas, F. (Ed.), The Handbook of Sociolinguistics. Blackwell, London, pp. 217–237.
Platt, J., Webber, H., 1980. English in Singapore and Malaysia. Oxford University Press, Oxford.
Sankoff, D., Poplack, S., 1981. A formal grammar for code-switching. Papers in Linguistics: International Journal of Human
Communication 14, 3–45.
Scotton, C., Yule, W., 1977. Bilingual strategies: the social functions of code-switching. In: Zhu, W. (Ed.), 1985. Translated Articles of
Sociolinguistics. Peking University Press, Beijing, pp. 199–217.
Simon, L., 2001. Towards a new understanding of codeswitching in the foreign language classroom. In: Jacobson, R. (Ed.), Codeswitching
Worldwide II, . In: Trends in Linguistics: Studies and Monographs, vol. 126. Mouton de Gruyter, Berlin, pp. 311–343.
Treffers-Daller, J., 1998. Variability in code-switching styles: Turkish-German codeswitching patterns. In: Jacobson, R. (Ed.),
Codeswitching Worldwide, . In: Trends in Linguistics: Studies and Monographs, vol. 108. Mouton de Gruyter, Berlin, pp. 177–200.
Wang, Q., Sun, L., Ma X., forthcoming. An impact study of a TEYL innovation project in Beijing, China. In: Enever, J., Moon, J.,
Raman, U. (Eds.), Young Learner English Language Policy and Implementation: International Perspectives. IATEFL Young Learner
Special Interest Group, Kent.
Zentella, A.C., 1981. Ta bien, you could answer me en cualquier idoma: Puerto Rican codeswitching in bilingual classroom. In: Duran, R.
(Ed.), Latino Language and Communicative Behavior. Alberx Publishing Corporation, Norwood, NJ, pp. 109–132.