Abstract:

The paper examines one feature of communication strategies, allo-repetition (two-party repetition) in conversation. It explores the many roles of repetition through a study of informal dyadic conversations between English as a Foreign Language (EFL) learners from Indonesia, Vietnam and Japan, and native speakers of English. Allo-repetition contributes to development, maintenance and coherence of a conversation; and is a principal strategy under the control of non-native speakers that enables them to communicate their positive involvement and interest in conversation where language skills are unequal. The data discussed in the paper confirm six categories of allo-repetition identified by Tannen (1989) and others: to indicate participatory listenership, to justify listenership, to request for confirmation, to request for clarification, to stall, and to indicate surprise. The study confirms the positive roles of repetition strategy in maintaining the conversation. The paper also presents an additional form of allo-repetition, to ensure correctness, which appears to take a more important role in cross-cultural conversations in which one party is a conscious learner of English. As the data indicate, in the interpretation of repetition strategies, prosodic characteristics such as intonation, the ensuing responses by the interlocutors, and the use of associated speech particles, can all help to explain the various functions of allo-repetition.

Keeping up with native speakers:
The many and positive roles of repetition in the conversations of EFL learners

INTRODUCTION

Celce-Murcia, Dornyei and Thurrell (1995) classify repetition as one of their taxonomy of communication strategies. Communication strategies form part of strategic competence which comprises verbal and non-verbal strategies that language learners utilise in order to compensate for lexical problems (Canale & Swain, 1980), to enhance the effectiveness of communication (Canale, 1983:11) and to sustain the continuity of a conversation in the face of communication difficulties, for example, by playing for time to think while searching for the intended meaning (Celce-Murcia, Dornyei & Thurrell 1995). The investigation of repetition in this paper is approached on the premise that repetition, with its varying forms and functions, contribute to the development, maintenance and coherence of a conversation. In Tannen's (1989:97) word repetition 'is a resource by which conversationalists together create a discourse, a relationship, and a world. It is the central linguistic meaning-making strategy, a limitless resource for individual creativity and interpersonal development'.

The study of conversation is very important in the area of second language acquisition. Contrary to the old view that repetition is an indication of lack of speaking skills, repetition is indeed a resource that language learners can utilise to enable them to engage in a conversation despite their language constraints. Wong (2000:408) argues that repetition is 'a human, social activity, clearly part of our every day conduct and behaviour and not just a marker of a disfluent or sloppy speaker'. This paper draws on data gathered in research of free-flowing conversations between English as a Foreign Language (EFL) learners from three Asian countries, studying in an English-speaking country (Australia), and native speakers of English, to study repetition in conversation. The objective of the study is to identify and examine one kind of repetition, allo-repetition in the EFL learners' conversations. (Allo-repetition is repetition that is inter-active in the sense that it involves both participants in a conversation). The paper first considers relevant literature and research on repetition strategy. It then presents the findings of the research, examining and discussing the functions that are performed by each act of allo-repetition in the data. The paper concludes with some of the implication of the findings.

BACKGROUND RESEARCH ON ALLO-REPETITION

There are two forms of repetitions depending on which speaker produced them. Self-repetition (or same-speaker repetition, Schnelby, 1994) is the repetition that occurs when the speakers repeat themselves. Other-repetition is a joint work between speakers and their interlocutors. Other-repetition has been labelled differently by different researchers: two-party repetition (Murata, 1995), second-speaker repetition (Schnelby, 1994; Simpson, 1994) and allo-repetition (Tannen, 1987). The latter term 'allo-repetition' is used in this paper.

According to Tannen (1989:48) the varied purposes served by repetition can be subsumed under the categories of production, comprehension, connection and interaction. The first three categories refer to the creation of meaning in a conversation. The latter category, an interactional function, goes beyond the level of meaning creation. It accomplishes social goals in a conversation. Tannen's (1989:51) data indicate the following interactional functions of repetition: to get or to keep the floor, to show listenership, to provide back-channel responses, to stall, to gear up to answer or speak, to indicate humor and play, to savour and show appreciation of a good line or a good joke, persuasion, to link ideas and to ratify another's contributions. Instances of repetition vary ranging from repetition of exact words with similar rhythmic pattern to paraphrases and other variation such as changing the person, tenses or the wording (Tannen, 1989). Brody (1994) states that repetition may take the form of full, reduced or expanded quotation.

Research on repetition has been a fruitful area of investigation. Pioneered by Tannen (1987), repetition in a conversation has also been investigated by numerous other researchers (Murata, 1995; Lyster, 1998; Rieger, 2003; Jensen & Vinther, 2003). Some other researchers (Ferrara, 1994; Simpson, 1994; Dumitrescu, 1996; Perrin, et al. 2003) examine repetitions in accordance with particular prosodic characteristics, namely intonations. Murata, for example, conducts a cross-cultural study on repetition, comparing repetition strategies used in NSE/NSE (native speaker of English), NSJ/NSJ (native speakers of Japanese) and NSE/JSE conversations. She examines the immediate repetitions utilised at topic and subtopic boundaries. Her data indicated the following kinds of repetitions: interruption-oriented repetitions, solidarity repetitions, silence-avoidance repetitions, hesitation repetitions and reformulation repetitions. The first two repetitions were allo-repetitions and the last two were self-repetitions. Murata argues that the use of repetition is determined by socio-cultural values attached to its use. Her data indicate that Japanese styles of interaction are characterised by the use of solidarity repetitions, signalling to the interlocutors that what have been said by the first speaker is understood and even agreed, to show solidarity and to establish rapport in an interaction. Her data indicate that some kinds of repetitions are confined exclusively to either English language or Japanese language interaction. In an English interaction, silent-avoidance repetitions were more noticeable to indicate participation in a conversation. Simpson's (1994) study reveals that allo-repetitions with falling intonation function as an answer to a question, a supportive back-channel, a correction and an imitation. Repetitions with rising intonation serves as a question, and those with a rising-falling intonation are a sign of emphatic stress. Perrin (2003) identified four functions of allo-repetition: a taking into account function, confirmation request function, positive reply function and a negative reply function.

This paper uses Tannen's conception of the interactional functions of repetition focussing on allo-repetition in general regardless of types (i.e from one word to expanded quotation). Data used in this study were drawn from the author's doctoral research which looked at the communication strategies used by the EFL learners interacting with native speakers of English in free dyadic conversations. The study found that both self-repetition and allo-repetiton were strategies extensively employed by all learners. The present paper concentrates on allo-repetition as utilised by the EFL learners only, specifically identifies and examines their interactional functions. The allo-repetitons used by native speakers are not analysed in this paper.

METHODS
Participants
The participants selected for this study were two groups of students aged between 19 and 25 years of age. Prior to their involvement in the research, they were not acquainted with each other. The first group consisted of nine learners of English as a Foreign Language (EFL) from three countries in which English is learnt as a foreign language: Indonesia, Vietnam and Japan. Four were female and five were male. All of the EFL learners were undertaking an English course in the English Language Centre at Monash University. This was a bridging program of 10 weeks in length that was designed to supplement their International English Language Testing System (IELTS) score as required by the university, a test of the standard of English used as one of the requirements for entry to Australia. Some of the EFL learners had just finished their secondary schooling, while some had undertaken a degree course in their home country. The Head of the Language Centre stated that their level of English was 'intermediate'. It was assumed that learners at this level would be able to use their existing language knowledge communicate orally. The second group of students comprised native speakers, five female and four males, who were in the process of completing a teacher training qualification in English in the Education Faculty, Monash University.

The selection of these two groups of students was based on the criteria of accessibility and geographic proximity. The researcher had had prior contact with the head of the English Language Centre and with the senior lecturer in charge of the English Teaching Method in the Faculty of Education, Monash University. Though the native speakers were not yet fully qualified teachers, the fact that this research was conducted in an Education Faculty and that these participants were learning to become teachers may have caused them to interact with the non-native learners in 'teacherly' ways. This might have influenced the relationship between the two groups; and in turn, may have had an impact on the results of this study.

Procedure
Prior to the recording the participants were asked for their consent. On reading the explanatory statement, all participants were asked for their consent through an informed consent letter. The informed consent letter explaining the objective of the research was distributed to the participants to enable them to fully understand any potential risks that the research could cause them. Pseudonyms were used in data analysis in an attempt to conceal participants' names or identities for their privacy. After the participants had agreed to participate in the study at an agreed time and place, a video recording was conducted. Each conversation took approximately 45 minutes. It took place in an informal setting. The conversations were videotaped, transcribed and later analysed qualitatively.

Before the recordings the group was informed that the focus of the investigation was on the flow of conversation rather than any particular aspect of their conversation. They were asked to chat freely about everyday topics such as school, holidays, other experiences, and interests. The native speakers were asked to make every effort to keep the conversation going if necessary. This last aspect may also have affected the character of the study. For example when there were gaps in the conversation, the native speakers acted to ensure that the conversation would continue. The interactions were not completely spontaneous, but nor were they one-way, such as an interview. Rather they took the form of a controlled conversation. The control was not only exercised by the native speakers. The EFL learners wanted to demonstrate and improve their skills. If unable to maintain their part of the conversation, they would have lost 'face'. For different reasons, both parties had an interest in keeping the conversation going, and were contributing to its management. This was part of the context of conversational behaviour, including the use of repetition strategies.

ALLO-REPETITIONS IN THE RESEARCH DATA

The section that follows lists the various allo-repetition used by the learners. Note that some allo-repetitions can be interpreted as having more than one function. The following conventions are used in the transcript.
[ ] ..........................overlapping speech
@@ .......................laughter
(???).......................unclear speech
. ............................brief pause (one second)
.. ...........................two second pauses or longer
bold type ................indicates the focus of analysis which is allo-repetition
..............................[bold extracts are always EFL learners' speech]

1. To indicate participatory listenership
As found in other studies (Tannen, 1989; Murata, 1995, Perrin et al., 2003), allo-repetitions performing participatory listenership functions were observed frequently in these data. Murata (1995) calls this type of repetition 'solidarity repetition' in which no new information is added to the development of a topic. It indicates listenership and participation as illustrated in the following three extracts:

Extract 1:
N: I'm studying here I'm doing my Dip.Ed my diploma of education
D: diploma
N: it allows me to teach.in secondary school…so.. once I get out I can register as a teacher
D: mm
N: in Australia
D: mm
N: to teach English [which is]
D: [to teach] English
N: that's right.that's what I'll be teaching

Extract 2:
A: and she..how it works.. she um cos she spoke English she dealt with all the um with all the um English speaking clients?..that was in Nagasaki I think?
T: Nagasaki
A: yeh is that right
T: yes that's right

Extract 3:
N: I've got some.my great-grandfather was German
D: oh German
N: hm.so I'd like to go over there perhaps one day
D: hm

D (Indonesian) in extract 1 repeats the word 'diploma' and 'to teach English' just to indicate that he is following what the native speaker had just said, showing his listenership and involvement. These repetitions do not add any information to the development of the topic in the conversation, illustrating Murata's definition of solidarity repetition: 'the repetition of words and phrases in solidarity repetition does not add any new information to aid the development of a topic, or story, but shows the conversational interactants' involvement, cooperation, agreement, and participation in conversation' (p.349). This involvement or listenership is emphasized by an overlapping repetition. In extract 1 the learner's repetition 'which is' overlaps with the native speaker's utterances 'to teach' is indicated by the square brackets. According to Murata overlapping repetition indicates positive attitudes towards conversational partners. The expression 'oh' preceding the repetition 'German' by the Japanese learner as illustrated in extract 3 seems to accentuate the learner's interest in what has been said by the native speaker. This kind of repetition also functions as back-channel as it does not add significant meaning to a conversation except to show listenership and to enhance rapport between co-conversationalists (Simpson, 1994, Brody, 1994).

2. To justify listenership
The allo-repetitions participatory listenership, and justifying listenership, are similar in nature. They both indicate involvement and listenership in a conversation. Justifying listenership repetition in extract 1 below demonstrates that the learner shows not only her listenership but also her agreement to what has been said, as emphasised by the 'yeh' preceding her repetition. This agreement is supported by the learner's knowledge of the subject under discussion. In extract 1 the learner is telling the native speaker that she has been to the Grampians, the Australian countryside. The learner is aware that the area is in the hills and it is cold. The learner (H) repeats part of the native speaker's utterances 'yeh the weather' and '[yeh in the] hills [yah]' signalling her acceptance or agreement to what the first speaker has just said. She does not just follow what the native speaker has said but also reasserts that that it took a little while to adjust to the weather and that the Grampians are in the hills. The agreement was based on the knowledge shared by the native speakers and learners. Extract 3 illustrates similar function. This kind of repetition is also called 'positive reply' (Perrin et al., 2003) and 'echoing' (Ferrara, 1994:70). Perrin, et al. (2003:1825) state that in this context 'the speaker qualifies the interlocutor's discourse ex post facto as a kind of anticipation of what he would have said himself, and reciprocally presents his own discourse as a reproduction of the interlocutor's'. According to Ferrara, echoing is the repetition of another person's statement using similar word choice and a similar basic intonation pattern to express agreement with or acceptance of the statement. In the therapy setting this echoed repetition has more emphatic meaning than just agreement if it is proceeded by the word 'yes' or 'exactly' (Ferrara, 1994:70). But the use of 'yeh' in the extract below simply indicates an emphasis of agreement. The overlapped repetition in extract 1 may indicate positive attitudes towards conversational partners, as in participatory listenership.

Extract 1:
A takes a little while to settle into like the weather and [all that sort of stuff]
H [yeh the weather@]
A its beautiful ah .countryside?
H yes
A yeh
H but its very cold too
A [oh yehh]cos its.its in the hills isn't it yeh in the [hills so]
H [yeh in the] hills [yah]
A [yeh] did you take lots of jackets [lots]

The repetition by learner P as in 'yes French colony' in extract 2 below was based on knowing that Vietnam had been a French colony. His repetition carried out two functions: to show his listenership and to agree with the native speaker plus to indicate his knowledge of Vietnam history.

Extract 2:
C: oh is that like um a different alphabet?
P: intonation
C: oh yes yes oh because Vietnam was part of France no it was a French colony
P: yes [French colony]
C: [so you'd have] little dots and everything like hash things oh yeh
P: yes this one (???)
C: oh yep yep yep

3. To ensure correctness
Repetition can also be used after correction as a remedying utterance. In extract 1 the learner explains to the native speaker that he is enrolling in a bridging course prior to his postgraduate studies. The native speaker asks him what course he will do after completing the bridging course. The learner seems to have lexical difficulty in explaining the name of the course, describing it not as 'finance' but as 'financial'. The native speaker corrects his utterances, followed by the learner's repetition of the same word to confirm correctness. Similarly, learner's repetition of 'hand-out' in extract 2 is to ascertain that he has the word right.

Extract 1:
S: aha and what are you actually gonna do once you finish well what course
H: ah..financial
S: finance
F: finance
S: yeh

Extract 2:
C: and in class during the day is that what you work on?
P: no we work on language we work on grammar we work [???]
C: [yep yep] from a textbook?
P: mm no they give some hand-in
C: ah hand-out
P: hand-out
C: yeh hand-out
P: and then read it and then
C: ahh

4. To request confirmation
Repetition is used as a request for confirmation when the speaker is unsure of what has been said by the interlocutor. The learner in extract 1 repeats the word 'Olympics' with a rising intonation because he is not so sure which Olympics the native speaker is talking about. The native speaker's ensuing utterances 'in Sydney' provides extra information which helps the learner understand the question. This repetition is normally used with a rising intonation, as illustrated below.

Extract 1:
S: you planning on going up to the Olympics when they're on?
H: Olympics?
S: Olympics
H: @
S: in Sydney
H: @no

In extract 2 below the native speaker asks the learner about her favorite television shows The learner says that she likes the TV serial 'Charmed'. Her repetition 'Charmed American? with a rising intonation suggests that she had thought 'Charmed' was an Australian show. She is repeating the word to confirm their truth.

Extract 2:
P: yes they're earlier.cos they're Australian shows
S: yeh
P: but Charmed is American
S: Charmed American?
P: American show
S: yeh oh I like Australia Funniest Home Videos
P: ohh @@

According to Deen (1997) confirmation checks are used when the listener does not sufficiently understand what have been said by his/her interlocutor. A request for confirmation or in Deen's term 'confirmation check' is used as a double check. To do this the speaker can repeat with a rising intonation and he/she can also supply other words, as in extract 3 below. The learner does not just repeat the word 'Bali' to request confirmation but also supplies other words 'in Indonesia?' Repetition functions as confirmation checks may take the form of full repetition, partial repetition, additive repetition, substitutive repetition and complex repetition (Deen, 1997). In the data studies here, repetition takes the form of partial repetition, in which learner repeated part of the native speaker's utterances.

Extract 3:
S: this is the first time I go abroad
P: I've been to Bali
S: Bali? in Indonesia?
P: yes
S: ahh

5. To request clarification
Unlike a request for confirmation, repetition that functions as a request for clarification signifies that the speaker lacks understanding of the interlocutor's utterances. Both a confirmation request and a request for clarification trigger a repair which might take the form of a confirmation, a clarification, or a correction (Perrin, 2003). In the present study this lack of understanding is triggered by the lack of shared common ground between the native speaker and the learner. The learner repeats the troublesome utterances with a rising intonation, signaling confusion and requiring a clarification. In extract 1 the learner's repetition 'big move?' points out that he did not sufficiently understand the native speaker's utterances. His lack of full understanding is also shown by his further repetition as in 'makes new friends?' The native speaker's question 'was it a big move' triggered further exchange. This illustrates Roulet's model of discourse analysis in which confirmation request (in this context clarification request) 'open an exchange that is subordinate to the interlocutor's prior move' (Perrin et al. 2003).

Extract 1:
N: how did you find moving over.to Australia
D: ..ah from Indonesia
N: yeh..how did you find that
D: ahh
N: was it a.a big move?
D: big move?
N: yeh.moving from Indonesia to Australia?or.in terms of making new friends and um..
D: make new friends?
N: yeh
D: you mean moving here

A request for clarification can also be extended with a downward intonation, as illustrated in extract 2 below. It is relatively difficult to label repetition of this type. It can be interpreted either as showing listenership and involvement, or as an indication of confusion. If 'water' has been preceded by 'yes', it can be assumed that it is an involvement function. However, in this case the native speaker's ensuing response 'yeh you know bottle water' indicates that learner is actually seeking clarification. The native speaker assumes that the learner's repetition of 'water' is a signal of lack of sufficient understanding of what has been said. The native speaker's response then triggered the learner's response, 'ah', confirming his understanding.

Extract 2:
A: alright where are you from
T: I'm from Japan
A: Japan
T: yeh
A: my ex-girlfriend (she lived in Japan) she worked for a (don't know what it was) a company that sold water?
T: water
A: yeh you know bottled water? to um to other companies
T: ah
A: and she..how it works.. she um cos she spoke English she dealt with all the um with all the um English speaking clients?..that was in Nagasaki I think?

The learner's repetition of 'water' with a downward intonation implies a confusion which requires further clarification. Ferrara (1994) calls this kind of repetition 'mirroring', a repetition of part of the interlocutor's utterances stating that this repetition is an indirect request for elaboration. Ferrara's context was that of conversations between a therapist and a client but the point applies to a free conversational context.

6. To stall
Some repetitions function as stalling, that is to gain time while thinking about the next utterances, as the following examples show. The learners want to answer the questions but need time to think an appropriate answer. In extract 1 the learner's repetition 'in one class' has double function: as a request for confirmation indicated by a raising intonation; and as a stalling which signals the need for more time before producing the next utterance or response. Stalling is also indicated by repetitions preceded or followed by expressions such as 'oh', 'mm' in extracts that follow. These are particles used while searching for the answer. Perrin, et al. (2003) calls this type of repetition 'taking into account' which serves two functions: as a backchannel and as a preparation for an answer. Repetition in extract 1 does not function as a back-channel but as a request for confirmation as it is followed by a rising intonation. The learner repeats 'in one class?' to confirm. In extracts 2 and 3, allo-repetitions obviously function as stalling because allo-repetitions are uttered with downward intonation.

Extract 1
A: when um.like you start the class about how many people.like in one class
H: in one class? Oh about…hm forty.can more than forty
A forty.[wow]

Extract 2:
A wow.. what are some of.like your like favourite movies
H: oh favourite music movie..mm..ah romantic or love everything sad movie
A: yeh.have you seen um…Sleepless in Seattle?
H noh@

Extract 3:
D: do you have pets?
H: pets ah in Japan I don't have but my host-family have one dog
D: oh yeh friendly

There are occasions when repetition is not accompanied by particles such as 'mm', 'oh','ah' but with a rising intonation. Extract 4 is one example of this. Clearly the learner understood the question, but needs some time to produce his answer. He repeats 'holidays' while trying to think of what he did in the holidays. Through a succession of turns by the native speaker it becomes established that the question asked by the native speaker is too general. The learner's repetition 'in Australia?' provides a clue for him. Even so the learner still needs more time before producing his answer as seen in his response 'oh…' which signals this to the native speaker. While he is searching for the answer, the native speaker specifies by saying 'have you been outside of Melbourne? or'. After a series of exchange, the learner eventually mentions that he has been to Phillip Island.

Extract 4:
N: yeh..what do you do for.holidays
D: holidays?
N: holidays
D: in Australia?
N: in Australia
D: oh..
N: have you been outside of Melbourne?or
D: [ahhh]
N: [for a look around]
F: well I think just on the.Phillip Island is it outside Melbourne?

Stalling in a conversation can be achieved by either repeating the interlocutor's utterances or self-repeating. Rieger's (2003) study, to gain time while searching for the next utterances, the subjects (English-German bilinguals) utilize repetition as a self-repair strategies thereby gaining time in the conversation. In Rieger's study, repetition as self-repairs was used differently depending on the language of use. Subjects were reported to use more repetitions on pronoun-verb combination, more personal pronouns and more prepositions in English than in German. Differences on the structure of a particular language contributed to the differing use of repetition as self-repair.

7. To indicate surprise
The native speaker (P) explains to the learner (S) that she has six sisters (extract 1). The learner's repetition with rising intonation is not a request for confirmation, as the native speaker's utterances are very clear and contextual. The repetition with a rising intonation 'six sister?' was to indicate surprise and disbelief. This is in line with the finding by Bolinger (1989) that a confirmation request with a prosodic feature can be used to express surprise and astonishment. Following Jefferson (1972) Simpson (1994) notes that repeated questions that are characterised by disbelief or surprise signals a problem in the previous utterances suggesting that is to remedy the problem. To ensure the correctness of the native speaker's statement the leaner generates further questions to seek clarification, as can be seen in the consecutive lines in bold type.

Extract 1:
P: yeh but in my family I've got my mum and dad? And I've got six sisters
S: six sister?
P: yeh
S: ohh
P: all older than me and then I've got one younger brother
S: younger
P: mm
S: so you have six sister and one brother
P: yeh
S: ahh so your family have eight children ahh
P: mm big family
S: big family yeh
P: yeh
S: there only six people in my family but I think it is a big family@
P: @@
S: but your family is bigger than mine
P: yeh much bigger@

The learner's repetition 'six sister?' could be substituted by such expressions as 'really' or 'is it true' with a rising intonation. By repeating the same word in this context 'six sister' can embrace disbeliefs and surprise at the same time.

IMPLICATIONS

Functions of repetition

In the data discussed here, seven functions of allo-repetitions have been identified: repetitions that indicate participatory listenership, justify listenership, ensure correctness, request for confirmation, request for clarification, stall, and indicate surprise.

Both repetition as an indication of participatory listenership and repetition to justify listenership are utilised to show involvement and solidarity. The latter also carries another function; to agree with what the interlocutor has just said, based on a shared common knowledge between speaker and listener. Repetition is also used as a remedying utterance to ensure correctness. Repetition can be employed as a way of requesting confirmation or clarification. The latter triggers more explanation by the interlocutor. In relation to stalling, a lack of linguistic knowledge or knowledge of the topic under discussion may lead the speaker to use repetition to stall, so as to gain time while thinking of the next utterance. Repetition to indicate surprise occurs in cases where the speaker has understood what he/she has heard, but finds it difficult to accept. All of these functions of repetition were quite active in the data, indicating the very significant and complex role played by repetition in conversation.

Issues of interpretation

It is not always easy to identify what functions are performed by a certain instances of allo-repetition. Some allo-repetitions perform a clear-cut function, and others require further clues to complete the interpretation. Much is said in the literature that the interpretation of repetition in a conversation depends on many factors including the type of situation; the nature of the relationship between the interlocutors, and previous interactions between them; the mood of the participants; prosodic features such as tone (Perrin, et al., 2003), intonation pattern (Simpson, 1994), and other parameters such as gestures and facial expression (Bolinger, 1989; Selting, 1996; Paradis et al., 2001 in Perrin et al, 2003); the culture that provides the options, speech event or setting in which they occur; and on the practices of the interlocutors in choosing among discourse options (Ferrara, 1994:69). Similar to previous findings, this study also identifies the importance of intonation pattern in interpreting the function of allo-repetition. For example, repetition of the other's utterances with a rising intonation may be a confirmation request rather than just an indication of listenership. There are many possible explanations on the use of intonation. Its use can be understood in terms of participants' first language. The Vietnamese language, for example, is a tonal language in which a word can carry various meaning depending on how it is pronounced. In the conversation of Vietnamese learners, in the context of this study, first language intonation transfer might present in their use of allo-repetition strategy. (This area may constitute a useful avenue for further research).

As noted in the preceding analysis, conversational data can also provide other clues which can be helpful in interpreting allo-repetition analysed through transcript. For example, in cases when there is no clear-cut function performed by a particular allo-repetition, ensuing utterances by the interlocutor help interpret the function. Particles such as 'yeh', 'yes' preceding allo-repetition can also provide clues for interpretation.

The use of allo-repetition may at some points reflect the cultural values of its users towards the conduct of everyday conversation. This study is too small to draw such a discussion. However, certain findings from this study support other research in the area. The Indonesian and Vietnamese learners mostly used allo-repetition as a confirmation check. Japanese learners tended to repeat the native speakers mostly to show listenership and solidarity. This conforms to Murata's (1995) study of Japanese conversations with native speakers of English, which found many instances of solidarity repetitions. She noted that it is typical of Japanese interaction to show respect for the territoriality of the other speaker; although the opposite may be the case with native speakers of English, who show their solidarity by intruding on the territory of the other. There is some confirmation of both behaviours in the present study, as shown in some of the extracts quoted above. However, a more extensive study that was specifically focused on the range of cross-cultural aspects would be fruitful here.

CONCLUSIONS

The study has acknowledged the importance of research on repetition, in association with other intonation pattern and other features. The varied functions performed by allo-repetition strategies provide ample resources for EFL learners to achieve their communicative goals, while simultaneously sustaining their involvement in and interest in the conversation. Repetition is one of the many communication strategies that constitute part of successful cross-cultural conversations.

It is important to acknowledge the positive role of repetition, again confirmed by this study. In the conversations discussed here, it is clear that repetition is not a sign of conversational inadequacy or absence, but the contrary: it indicates the will to sustain conversational presence. It is a constructive learner-generated and learner-managed device. It enables the EFL learners, taking part in a cross-cultural conversation in English with more accomplished speakers, to maintain the momentum of conversation while simultaneously learning, developing and using their emerging language skills.

It can be hypothesised that in the context of cross-cultural conversation, repetition is more frequent, and a more important strategy (particularly for non-native speakers) than in conversations between native speakers. To establish this would require a further study comparing repetition used by native speakers in native/native conversation with repetition as used by learners of English, from different cultural backgrounds. As Murata (1995) notes, repetition strategies are cultural specific.

Contribution of this study

A study is valuable if it confirms what other studies have found, in a new context, and also if it adds new knowledge. Six of the functions of repetition strategies as discussed above were also discussed in Tannen (1989) and are referred to in some other research in the area (Murata, 1995; Lyster, 1998; Jensen & Vinther, 2003, Ferrara, 1994; Simpson, 1994; Dumitrescu, 1996; Perrin, et al. 2003). This study confirms the relevance of the previous findings in the context of cross-cultural conversation.

In addition, the study has added an additional function of repetition to Tannen's and other researchers' list of functions. This additional function of repetition is that of remedying utterances so as to ensure correctness. In this respect, the present study adds to our stock of knowledge about repetition strategies in English-language conversation. This particular function becomes more important in cross-cultural contexts - especially when the non-native speaker is learning English - than in conversations between two native speakers. That is because issues of correctness and usage are always highlighted in the cross-cultural context. The use of repetition as a remedying strategy is an example of 'learner-generated attention to form' (Williams 1999), an approach to non-didactic language teaching and learning that has been emphasised in some recent research on English-language instruction. The present study suggests that conversation between native and non-native speakers is likely to provide a fruitful environment in which this use of repetition as a self-learning strategy can be developed. Further research with a different emphasis would be needed to identify the pedagogical potential of such conversations.

REFERENCES

Bolinger, D. 1989. Intonation and Its Use: Melody in Grammar and Discourse. Stanford: Stanford University Press.

Brody, J. 1994. Multiple repetitions in Tojolab'al conversation. In B. Johnstone (Ed.) Repetition in Discourse. Interdisciplinary Perspectives, vol. 2. (pp. 3-14). Norwood, New Jersey: Ablex Publishing Corporation.

Canale, M & Swain, M. 1980. Theoretical bases of communicative approach to second language teaching and testing. Applied Linguistics, 1 (1), 1-47.

Canale, M.1983. From communicative competence to communicative language pedagogy. In J. C. Richards, & R. W. Schmidt (Eds.). Language and Communication. (pp. 2-27). London: Longman.

Celce-Murcia, M, Dornyei, Z. & Thurrell, S. 1995. Communicative competence: A pedagogically motivated model with content specification. Issues in Applied Linguistics, 6 (2), 5-35.

Deen, J. 1997. Dealing with Problems in Intercultural Communication. Studies in Multilingualism 5, Amsterdam: Tilburg University Press.

Dumitrescu, D. 1996. Rhetorical vs. nonrethorical allo-repetition: The case of Romanian interrogatives. Journal of Pragmatics, 26, 321-354.

Ferrara, K. 1994. Repetition as rejoinder in therapeutic discourse: Echoing and mirroring. In B. Johnstone (Ed.) Repetition in Discourse. Interdisciplinary Perspectives, vol. 2. (pp. 66-83). Norwood, New Jersey: Ablex Publishing Corporation.

Jensen, E. & Vinther, T. 2003. Exact repetition as input enhancement in second language acquisition. Language Learning, 53, (3), 373-428.

Johnstone, B. 1994. Repetition in Discours. Interdisciplinery Perspectives Vol. 1. Norwood, New Jersey: Ablex Publishing Corporation.

Johnstone, B. et al. 1994. Repetition in discourse: A dialogue. In B. Johnstone (Ed.) Repetition in Discourse. Interdisciplinary Perspectives, vol. 1. (pp. 1-20). Norwood, New Jersey: Ablex Publishing Corporation.

Lyster, R. 1998. Recasts, repetition, and ambiguity in L2 classroom discourse, SSLA, 20, 51-81.

Murata, K. 1994. Intrusive or co-operative? A cross-cultural study of interruption. Journal of Pragmatics, 21, 385-400.

Perrin, L., Deshaies, D., & Paradis, C. 2003. Pragmatic functions of local diaphonic repetitions in conversation. Journal of Pragmatics, 35, 1843-1860.

Rieger, C. L. 2003. Repetitions as self-repair strategies in English and German conversations. Journal of Pragmatics, 35, 47-69.

Selting, M. 1996. Prosody as an activity-type distinctive cue in conversation: The case of so-called astonished question in repair initiation. In E. Couper-Kuhlen & M. Selting (Eds.) Prosody in Conversation. (pp. 231-270). Cambridge: Cambridge University Press.

Simpson, J. M. 1994. Regularised intonation in conversational repetition In B. Johnstone (Ed.) Repetition in Discourse. Interdisciplinary Perspectives, vol.1. (pp.1-20). Norwood, New Jersey: Ablex Publishing Corporation.

Tannen, D.1987. Repetition in conversation: Toward a poetics of talk, Language, 63, (3), 574-605.

Tannen, D. 1989. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse. New York: Cambridge University Press.

Williams, J. 1999. Learner-generated attention to form. Language Learning, 49, (4), 583-625.