Language Idiosyncrasies in Second Language Learners ’ Use of Communication Strategies

The term, “language idiosyncrasies” can generally be defined as what an individual typically says that becomes part of his or her personality. In doing so, this language behaviour can either positively or negatively impact communication. Based on this premise, the current study sets out to examine the occurrences of language idiosyncrasies in second language (L2) communication, particularly at the point when communication strategies (CSs) were employed to overcome communication problems. By employing non-participant observations of real university admission interviews, this study departs from the other studies related to CSs which involved non-authentic, simulated environments. The observed interview sessions were conducted in the English language involving 29 Malay candidates from 20 interview sessions. These sessions were video-recorded before the raw data were transcribed. The results revealed some occurrences of language idiosyncrasies in candidates’ utterances hence, supporting Paribakht’s (1985) research finding that speakers exhibited idiosyncratic patterns in the realization of communication strategies. The results also concurred with the findings of past studies on the influence of speakers’ L2 proficiency level on their use of CSs. As all participants were Malays communicating in an English speaking context, issues on cultural values added to the richness of the data and will also be discussed. While the findings may not be generalizable to all populations, it is hoped that this study will inform curriculum developers of language idiosyncrasies of Malay students, who form the bulk of the student population in Malaysia, so that awareness may be raised and appropriate interventions may be introduced.


Introduction
The official language in Malaysia is Bahasa Malaysia while English is the second language (L2) and is taught in Malaysian schools through the Communicative Language Teaching (CLT) approach (Ministry of Education Malaysia, 1987).This approach was adopted in 1979 to replace the structural situational approach (Rajaretnam & Nalliah, 1999).Since it aims to enhance learners' communicative competence, ideally, the CLT approach should provide learners with opportunities to interact meaningfully and makes real communication the focus of language learning (Richards, 2002).The learners are also expected to induce L2 linguistic rules from the various communicative classroom activities.Although there have been several major revamps in the Malaysian education system, CLT remains as the teaching approach of the English language in Malaysian schools until today.
Despite being claimed as an ideal language teaching approach, the CLT has many detractors.After decades of implementation, the standard of English among Malaysian students was reported to be declining (Ayub, 2012;Choy & Troudi, 2006;Mustapha, 2008;Asraf, 1996).Of the major ethnic groups in Malaysia, the decline is most serious among the Malays (Gill, 2005).In terms of stakeholders, potential employers have been lamenting about the lack of English language competency of Malaysian graduates (Bardan, 2012;Khoo, 2001;Woo, 2006).These graduates were found to be struggling when communicating in English during job interviews due to low proficiency level in the language.Despite this however, many would still make an effort to communicate as they are aware of the high-stake nature of the communicative event.This is mostly seen in many job interviews held in this country.For instance, in My Career Fair 2012 which offered almost 300 job placement on-the-spot (Khairudin, 2012), a large number of candidates attended walk-in interviews which were mostly conducted in the English language.To a certain extent, this indicates their willingness to communicate in the language despite the many complaints about their poor communication skills.In situations such as this, the use of communication strategies (CSs) would be useful for the candidates to express themselves.For instance, a speaker could rephrase or repeat interlocutors' preceding utterances in order to seek for clarifications and to gain time, respectively.In this regard, when CSs are employed in the same way, it could reflect the speakers' language idiosyncrasies or peculiar ways in handling communication problems.It was out of this curiosity that this study aims to investigate the presence of language idiosyncrasies in the use of CSs among students in Malaysia.

Communicative Strategies
"Strategic competence" which is commonly referred as "communication strategies" (CSs) (Faerch & Kasper, 1983, p. 212;Kellerman & Bialystok, 1997, p. 31;Tarone, 1980, p. 417) was firstly introduced by Selinker (1972) in the context of L2 communication.It is one aspect of what is termed as "communicative competence" by Hymes (1972).According to him, a speaker with this type of competency would have both knowledge of abstract linguistic rules and the ability to use language correctly and appropriately that suits the context of interactions (Hymes, 1972).
Although there has been no universally accepted definition of CSs, this notion can generally be viewed from two different perspectives: psychological and interactional.The psychological perspective views CSs as "mental processes that underlie learners' language behaviour when dealing with lexical and discourse problems" (Nakatani & Goh, 2007, p. 207).Despite its great influence on many CS researchers and scholars, this perspective is criticized for its lack of emphasis on interactional function of CSs.Tarone (1981, p. 288) for one, argues that language is "not an object which is used but a part of communication-a living organism created by both speaker and hearer".She therefore broadens the definition of CSs by referring them as "attempts to bridge the gap between the linguistic knowledge of the second-language learner and the linguistic knowledge of the target language interlocutor in real communication situations" (Tarone, 1983, p. 65), hence, leading to an interactional perspective of CSs.Two important elements of the use of CSs as emphasized by Tarone are linguistic knowledge and real communication situations.
A review of literature on CSs showed that there are many different types of CS typologies and taxonomies proposed by different scholars based on their differing concepts on CSs.The traditional conceptualization, for instance, would limit the use of CSs to compensate for gaps in the speaker's L2 proficiency while researchers of general communication hold that CSs are also useful in handling a wider scope of problems that surface during the course of L2 communication (Dörnyei & Scott, 1997).This has led to differing views with regard to the function of the same type of strategy.Repeating interlocutor's preceding utterance for instance, is viewed by (Dörnyei & Scott, 1997) to share the same function with self-repetitions (repeating own utterances) and the use of fillers (e.g."okay").All of them are employed as a time-gaining strategy.Nakatani (2010, p. 122) however, views repeating other's preceding utterances as "shadowing" that serves to indicate listener's understanding of important issues being discussed and hence, functions differently from self-repetitions.
Despite the various types of CS typologies and taxonomies mentioned in the current literature, the current study has referred to Dörnyei and Scott's taxonomy (1997) to identify the types of strategies employed by the candidates due to its general concept of CSs.They refer CSs to "every potentially intentional attempt to cope with any language problem of which the speaker is aware during the course of communication" (Dörnyei & Scott, 1997, p. 179).This reflects that CSs are useful to handle a wider range of problems in L2 communication than just to compensate for insufficient language resources.In addition, this taxonomy also appears as a "holistic" taxonomy as it includes the types of CSs mentioned by other CS scholars such as Tarone (1983), Faerch and Kasper (1983), Bialystok (1983), (Paribakht, 1985) and (Willems, 1987).Table 1 illustrates this taxonomy according to the main categories and types of CSs.
Table 1 shows that Dörnyei and Scott's taxonomy consists of three main categories: Direct Strategies, Interactional Strategies and Indirect Strategies.While Direct Strategies provide alternatives to compensate for linguistic gap, Interactional Strategies "carry out trouble shooting exchanges cooperatively, and therefore mutual understanding is a function of the successful execution of both pairs of the exchange" (Dörnyei & Scott, 1997, pp. 198-199).The third category, Indirect Strategies are not strictly problem-solving devices but are strategies in facilitating the conveyance of meaning directly and helping to prevent communication breakdowns by keeping the interaction channel open, for instance, through the use of fillers or feigning understanding (Dörnyei & Scott, 1997).(Dörnyei & Scott, 1997, p. 197) These three types of strategies are further divided according to the types of language problems in communication, namely resource deficit-related strategies, own performance problem-related strategies, other performance problem-related strategies or processing time pressure-related strategies.It is under each of these language problems that various types of strategies are listed as seen in Table 1 above.
While research on CSs largely relates this notion with speakers' L2 proficiency level (Lam, 2010;Khanji, 1996;Ismail & Musa, 2006;Tan, Nor Fariza, & Nayef, 2012;Wannaruk, 2003), others focus on the effectiveness of the use of CSs among L2 learners (Abdullah, 2002;Ismail, 2012;Awang, Maros, & Ibrahim, 2012).In general, these studies revealed that learners with high proficiency level were able to use oral communication strategies more effectively than the low proficient learners.In terms of the number of use, Tan et al. (2012) found that the total number of CSs employed by low proficient learners greatly outnumbered those employed by high proficient learners.
Aside from L2 proficiency level, a speaker's use of CSs could also be influenced by the kind of data analyzed as past studies have revealed some differences in the findings when the respondents were given different types of tasks (Mei, 2009;Chen, 1990;Poulisse & Schills, 1989) .In An Mei's (2009) study where a questionnaire and speaking tasks were used as instruments to collect data from two task types namely concept identification and role play, it was found that paraphrasing or circumlocution (Dörnyei & Scott, 1997, p. 197) was employed by all respondents, hence, appeared as the most used type of CSs.This is followed by restructuring and repetitions.Another task-based study that reported the use of paraphrasing or circumlocution as the most frequently employed strategy was conducted by Rushita Ismail and Muriatul Khusmah Musa (2006).The tasks involved in their study were object identification, picture story and role playing task.
While the above two task-based studies showed some similarities in their findings, a different finding was reported when more natural data were analyzed.In a study by Ting and Phan (2008) for instance, the respondents were involved in oral interactions on whether university students should date during university days.Since the respondents were paired with others they knew to allow them to draw upon shared knowledge in their discussion, it provides some real-life elements in the data.Analysis of data showed that restructuring was the most frequently employed strategy, followed by lexical repetitions.Among other possible reasons, this difference could be caused by the naturalness of data.As for the current study which employed natural data, the findings were expected to show some similarities with Ting and Phan's (2008).
Aside from different data types, another point worth highlighted pertaining to studies on CSs is the context of study which could also influence the naturalness of the collected data.In this respect, it was observed that many past studies on CSs were conducted in classrooms (Lim, 2004;Abdullah, 2002) or simulated situations (Omar, 2003, Nor, 2008;Tan et al., 2012;Ting & Lau, 2008;Ismail & Musa, 2006).To the researchers' best knowledge, there has yet one conducted in an authentic high-stake environment that is capable of eliciting real CSs.As this is the gap that this study aims to fill, it departs from non-authentic, simulated environments by conducting observations on authentic L2 interactions in real university admission interviews.

Language Idiosyncrasies
"Language idiosyncrasies" can be defined as what an individual typically says that becomes part of his or her personality (Advanced English Dictionary, 2000).This notion is undeniably relevant in oral communication since some speakers have been found to display a peculiar way in handling communication problems, particularly in their use of CSs (Paribakht, 1985;Awang, Maros, & Ibrahim, 2014).As language idiosyncrasies can either positively or negatively impact communication (Awang et al., 2014), it is important to raise learners' awareness on this issue.
Although language idiosyncrasies can be present both in first language (L1) and L2 communication (Aguilar, Diego, & Sanchez-Lancis, 1999), the current study sets out to examine their occurrences in L2 communication, particularly at the point when the speakers had to employ certain strategies to overcome problems in communication.This was driven by Paribakht's statement that, "speakers may well exhibit idiosyncratic patterns in the realization of [strategic] competence" (Paribakht, 1985, p. 142).As language idiosyncrasies are part of one's language use which is greatly influenced by their L2 proficiency level (Lam, 2010;Tan et al., 2012;Wannaruk, 2003) and cultural values (Kramsch, 1998), it is reasonable to believe that the occurrences of language idiosyncrasies also reflect these two elements.It is for this reason that the findings of this study will also be discussed in relation to speakers' L2 proficiency level and cultural values.

Culture and Malay Cultural Values
By definition, culture refers to the norms, values, and beliefs of a particular group or community in a particular area or geographic location (Hofstede, 1980).The significance of culture in language studies has been highlighted by (Haesook, 2006) who, in reference to Kramsch (1998), reported that "[as] culture is acquired, socially transmitted, and communicated in large part by language, the language choice must be examined in light of culture and in relation to the specific interactions" (Haesook, 2006, p. 304).Driven by this statement, the analysis of data in this study will take into consideration the aspect of speakers' cultural values.As all respondents were Malays, discussions on their language use and idiosyncrasies will revolve around Malay cultural values.
Specifically on the use of CSs, Faucette (2001, p. 8) wrote that "just as Gumperz' 'discourse strategies' are cross culturally variable and a potential source of intercultural miscommunication, appropriate communication strategies use may also be culturally constrained".This statement does not only indicate the influence of culture on one's language use but also on their choice in employing CSs.
Being the largest population in Malaysia, the Malay ethnic group makes up the population by 50.59% of the total population, followed by the Chinese and Indians who make up the population by 21.97% and 6.63%, respectively (Department of Statistics Malaysia, 2014).According to the Federal Constitution of Malaysia (section XII, article 160), a person is considered a Malay when he or she is a Muslim, speaks the Malay language, adopts and practices the Malay culture.
The current literature on this ethnic group shows that there are some values, which are commonly associated with the Malays.These include being indirect (Zawawi, 2008;Maros, 2006;Salleh, 2005) and "subtle" in conveying their messages in order to avoid conflicts (Ali, 1995).In addition, the Malays are also known to be humble people (Abdullah, 1992), who would avoid self-praise by making understatement rather than overstatement (Teo, 1996) of one's achievements or personality.As one's language use could be influenced by his or her cultural values, it would be interesting to investigate further on how this factor leads to language idiosyncrasies.

Research Design
This qualitative study was set to investigate the presence of language idiosyncrasies among L2 speakers at the time when CSs were being employed to tackle communication problems in a real communication situation.The data were collected from observations on real university admission interviews conducted at Universiti Teknologi MARA (UiTM) Shah Alam after consent was obtained from the deans of two faculties namely the Faculty of Communication and Media Studies (FCM) and the Faculty of Sports Science and Recreations (FSSR).As English is the medium of instruction at UiTM, all interviews were conducted in English, making the communicative event a high-stake one.
The data of this study were based on 20 interview sessions, which were video-recorded.They were then transcribed using Jefferson's (2004) notation convention (see Appendix A).The choice for this notation convention was due to the fact that it allows natural conversations to be transcribed as they are conversed (Schiffrin, 1994) and among the most frequently used notation convention (Wengraf, 2001).In order to allow full understanding of the conversations, the transcriptions were prepared based on complete interactions although the data analysis was focused only on the candidates' language use as it is their performances that have been the concern of many.
To establish a profile of the candidates, the candidates' English result in Malaysian Certificate of Examination (SPM) from the respective faculties were obtained.No retrospective data however were available in this study as the researchers were not given access to them.As these data can provide useful information regarding how and why the candidates chose such strategies (Nakatani, 2005), they can help to enhance the validity of data interpretation.Their absence therefore has become a major limitation of this study.
Data analysis of this study commenced with the examination, identification and classification of the types of CSs employed by the candidates based on (Dörnyei & Scott, 1997) perspective and taxonomy of CSs.This is followed by a deeper analysis to elicit (any) evidence of language idiosyncrasies in the ways the CSs were being employed.

Respondents
The candidates in this study consisted of applicants for various degree programmes at FCM and FSSR in UiTM. 10 interview sessions were observed in each faculty.At FSSR, one candidate was called in in each session, giving a total of 10 FSSR candidates.Meanwhile, at FCM, the first session involved one candidate.However, when the interviewers realized that they might not have enough time to conduct the interviews on one-to-one basis, they decided to call in two candidates for the remaining 9 sessions, hence giving a total of 19 FCM candidates.All these gave a total of 29 candidates from 20 sessions observed in this study.
In order to measure the candidates' proficiency level of the English language, the researchers referred to the results of their English language in SPM obtained from the respective faculties.SPM is a standardized public examination taken at the end of upper form level in Malaysian schools.For the purpose of this study, the following categorizations have been referred to in describing the candidates' level of proficiency in the English language.
Referring to Table 2, candidates who gained distinction results (1A and 2A) were considered highly proficient while those who obtained B3 or B4 were regarded as proficient users of English.Meanwhile, candidates with weak credit of C5 and C6 were considered moderately proficient while those with lower grades than 6C were categorized as low proficient users.In the examined interview sessions, there was one panelist head who was assisted by another panelist.Both of them were Malays.For the purpose of this paper, both panelists were denoted by Int. 1 and Int. 2, respectively.("Int." is short for 'interviewer").As for the candidates, they were identified by a pseudonym to provide anonymity.

Data Analysis
After transcriptions were prepared based on full interactions that occurred in each interview session, an analysis was carried out with the help of Nvivo software (version 8.0) to examine the use of CSs and any presence of language idiosyncrasies.Based on Dörnyei and Scott's perspective and taxonomy of CSs (1997), the analysis started with identifications of the use of CSs by the candidates.In doing this, the researchers viewed the video recordings in tandem with the transcriptions in order to avoid mistakes in classifying the types of CSs.Upon identifications of such strategies, the portions of data that contain the use of CSs were noted and further examined to identify the presence of language idiosyncrasies.Here, the candidates were considered to display language idiosyncrasies when their language use became typical that showed their personality (Advanced English Dictionary, 2000).This includes the use of similar utterances and ways of employing CSs.
The following section illustrates how the above analysis was done on an excerpt taken from candidate Anisah's data which resulted in identifications of the use of code switching.As seen in the above excerpt, words/phrases in italic are the examples of code switching (Dörnyei & Scott, 1997, p. 197) that occurred in Anisah's utterances when she was describing about her background.Notice that many of the switched words/phrases are not difficult words/phrases.Rather, they are commonly used in conversations.
Hence, it may be concluded that, unlike other speakers who usually code switch when they could not figure out the words in the target language, Anisah seemed comfortable switching back and forth between the English and Malay languages, hence, resulting in the occurrences of language idiosyncrasies.

Results and Discussion
This study aims to shed light on the occurrences of language idiosyncrasies among L2 speakers at the time when CSs were being employed based on the transcribed oral data taken from 29 candidates attending real university admission interviews.This is a high-stake context where real L2 communication takes place that many past studies on CSs lack of.Upon identifications of the types of CSs employed by the candidates, the data were further examined to see how the candidates' language use turned into an idiosyncrasy.The findings will then be presented in relation to the influence of speakers' L2 proficiency level and cultural values, particularly in the light of Malay culture of respect (Asma Abdullah, Singh, & Gill, 2001;Md Zabid Abdul Abdul Rashid & Jo Ann Ho, 2003).
On the use of CSs by L2 learners, the findings of this study showed similarities with those of some past studies.
In general, there are four types of CSs that significantly emerged from the data.Except expressions of non-understanding that could be interpreted by other researchers as seeking for clarifications or asking for interlocutors' help, the other three types of CSs identified in this study namely repetitions of others' preceding utterances, code switching and restructuring have been quoted in the literature as common strategies employed by L2 speakers.Ting and Lau (2008) and Ting and Phan (2008) for instance, reported an extensive use of restructuring and lexical repetition of others' preceding utterances while Tan et al. (2012) quoted code switching as the most frequently employed strategy in their study.
It is worth highlighting at this point that Ting and Phan's study (2008) contains some real-life elements which also characterize the current study.In their study, the participants were involved in a discussion on whether university students should date during their university days which should elicit natural conversations.Naturalness in the data was further added by pairing the participants with partners whom they knew and hence, were expected to be able to draw upon shared knowledge in their discussion.The naturalness of their data was further evident by extensive use of code switching that usually occurred in natural interactions such as informal conversations among friends.
As both Ting and Phan's (2008) and the current study contain the elements of naturalness in the observed oral data, it was not surprising that both studies reported similar results in terms of the types of CSs employed by the respondents.Both studies reported the occurrences of code-switching, restructuring and lexical repetitions of interlocutor's preceding utterances as strategies employed by the respondents.In addition, the current study also identified some expressions of non-understanding.It should be noted however, this strategy could be categorized differently by other researchers.It could be interpreted as a strategy to seek for clarifications or to ask for interlocutors' help, Unlike Ting and Phan's (2008) analysis which focused mainly on the use of CSs, the current study extended its data analysis into examining the occurrences of idiosyncrasies in the respondents' language use.At this point, it is worth highlighting that studies on idiosyncrasies could be carried out from various perspectives such as stylistic idiosyncrasies (Koppel, Schler, & Zigdon, 2005) and phonetic regularities that form speakers' idiosyncrasies (Aguilar et al., 1999).To the researchers' best knowledge, there are limited studies that focus on language idiosyncrasies which occurred from L2 speakers' peculiar ways of employing CSs that the current study is set for.

Expressions of Non-understanding
The first type of language idiosyncrasy identified in this study occurred at the point when one of the candidates expressed his non-understanding of what was asked or said earlier by the interviewer.Candidate Shamsuddin was found to utter 'hah?" at least four times when he wanted to express his non-understanding as seen in the following four excerpts which took place when the candidate was giving his background information.((Shamsuddin laughed)) As shown in the above four excerpts, Shamsuddin who obtained a weak credit of C6 in his English SPM result appeared to be comfortable uttering "hah?" to indicate his non-understanding of what was said by Int.1.To a certain extent, this indicates that, despite being exposed to CLT approach which teaches students the various ways to communicate their intended messages, Shamsuddin's way of communication was rather rigid, thus, leading to language idiosyncrasies.Despite many other ways of expressing non-understanding such as "pardon" or "excuse me" (Wannaruk, 2003, p. 7), Shamsuddin's repeated use of expression "hah?" could be caused by his low competency level in using the language as studies have shown that learners of different proficiency levels employ CSs at varying degrees (Lam, 2010;Tan et al., 2012;Wannaruk, 2003).In addition, this expression also seems to display a kind of language behaviour which may not correspond with Malay culture of respect (Abdullah et al., 2001;Rashid & Ho, 2003).
On the relationship between learners' L2 proficiency and their strategic competence, (Chen, 1990) reported that frequency, type and effectiveness of communication strategies used by the learners vary according to their proficiency level.In high-stake interactions such as in university admission interviews, expression "hah?" to express non-understanding would appear inappropriate considering the fact that the candidate was talking to someone of higher position (in this case, the interviewer was a university lecturer while the candidate was a potential university student) in which polite expressions were expected from the candidates.
When a comparison is made between Shamsuddin's and another candidate's expression of non-understanding whose name was Anisah, the influence of language proficiency level on the way one employs CSs is evident.Unlike Shamsuddin, Anisah, appeared to be more polite in expressing her non-understanding by repeating the interlocutor's preceding utterances in a rising tone as seen below.This candidate was expected to be a more competent L2 speaker than Shamsuddin since her moderate credit of B4 for her English in SPM is simply a better result than Shamsuddin who obtained a weak credit of C6.Observe her utterances in the following excerpt.
Excerpt 3.1 (e): (Anisah was asked about her SPM subjects) Int.1: Aa.During the SPM, [did] you take the History, subject?
In the above utterance, Anisah repeated the phrase "history subject" twice in a rising tone to indicate her lack of comprehension of what was being talked about.In order to help her, Int. 1 translated his question into Bahasa Melayu before Anisah started to understand.She therefore laughed and apologized for her confusion.
Based on the above excerpts, it may be safe to conclude that the ways CSs were employed by the candidates in this study did not only indicate their language idiosyncrasies, but also relates to their L2 proficiency level.Thus, despite what have been claimed by many that Malays are polite people, the case is different when it involves low competent speakers speaking in the English language whereby their inability to respond appropriately could result in inappropriate expressions.Following this, L2 speakers ought to be exposed to various ways of employing a particular type of CS to prevent them from repeatedly uttering the same inappropriate expression as seen in Shamsuddin's data.

Repeating Others' Preceding Utterances
This strategy is termed "other-repetitions" by (Dörnyei & Scott, 1997, p. 190) to refer to a situation whereby a speaker repeats what his interlocutor just said as a means to gain time before he is able to proceed with the conversations.Since this strategy occurred extensively in the data of the current study, it could be one form of candidates' language idiosyncrasies.
The substantial use of this strategy concurs with what was reported by Rajai Khanji (1996) who studied the relationship of the use of CSs with L2 learners' proficiency level based on oral data taken from 36 Jordanian EFL learners at the Intensive English Programme of the University of Jordan.He found that repetitions appeared as the most frequently used strategies (27%) followed by other types of CSs such as message abandonment (21%), transliteration, topic shift (11%), code switch (6%) and appeal for assistance (4%).The significance of this strategy was further acknowledged by Rushita Ismail and Muriatul Khusmah Musa (2006) who reported repetition as the second most frequently employed strategy in their study after circumlocution or paraphrasing.
In the current study, repeating other's preceding utterances appeared the most in the data taken from candidate Muiz who obtained a strong credit of B3 for his English result in the SPM exam.This candidate was found to employ this strategy 12 times as compared to an average of 2.5 times by other candidates.While (Dörnyei & Scott, 1997) classify all types of repetitions as a time-gaining strategy, (Nakatani, 2010) holds different view on the function of different types of repetitions.She categorizes the use of repeating others' preceding utterances as "shadowing" which "[consist] of exact, partial, or expanded repetition of the interlocutor's preceding utterance to show the listener's understanding of important issues" (Nakatani, 2010, p. 122).Labeling "shadowing" as a maintenance strategy, (Nakatani, 2010) asserts that this strategy functions differently from self-repetitions.In this respect, the latter serves more as a time-gaining strategy (Dörnyei & Scott, 1997) which could also be replaced by the use of vocalized fillers such as "er" and "erm" (Faerch & Kasper, 1983, p. 215) in order to hold the floor during conversations.
While it is admitted that the absence of retrospective data makes it difficult to ascertain candidate Muiz's real intention of repeating interviewer's preceding utterances (whether to gain time or to indicate understanding) before he gave his answer, extensive use of this in his data clearly indicates his language idiosyncrasies.The following are some excerpts taken from Muiz's data that show these occurrences.The repeated phases are italicized for ease of reference.
Int. 2: …the British Parliament, when [did they enact] this OSA, when was it?Muiz: When was it?Err::In about 1948 ((laughing as he knew that his answer was wrong)) Int.1: ((laughing)) In the above excerpt, notice that Int.2's utterance "when was it" was repeated by Muiz, followed by a filler "err" before he answered "1948".Here, after repeating the phrase "when was it", Muiz did not wait for Int.2's response to confirm his understanding of the question.This indicates that his repetition was a time-gaining strategy instead of merely asking for confirmation from Int. 2 due to communication problems that occurred during the interactions.
A similar response by Muiz could be seen again in the following excerpt which contains two cases of other-repetitions.Observe excerpt 3.2 (b) below, particularly the italicized utterances.
Excerpt 3.2 (b): (Muiz was asked to define the word "communication") Int.2: What is actually the meaning of <"communication">?Muiz: <Communication> er:: ((brief silence)) Int.2:Where was this word derived from?Muiz: Derived from.((brief silence)) Int.1: (inaudible) Int.2: Sociology.This is the basic in sociology."Communication", if you ask the layman, right?So, the layman will say, "communication" is an interactive activity between one person to another person but you, have undergone the, processes.What is actually the meaning as <"communication">.Originally.Where is actually this word being taken, derived from what word?Muiz: That is tough question.
When asked to give the meaning of the word "communication", Muiz responded by repeating the word with a drawl (indicated by <…>) which allowed him to "[plan] his subsequent speech unit" (Faerch & Kasper 1983, p. 215).This therefore, became his strategy to gain time before he could respond to Int. 1's question.From a different perspective however, this could also be viewed as "shadowing" (Nakatani, 2010, p. 122) which helped to indicate that the candidate was paying attention and understand the question asked by Int. 1. Again, the absence of retrospective data here makes is difficult to be certain of the candidate's real purpose of repeating the interviewer's preceding word.
Muiz's repetition of the word "communication" was then followed by a non-lexicalized filler "er::" and a short silence before Int. 2 rephrased his question by asking "where was this word derived from?".In responding to the rephrased question, Muiz repeated Int.2's utterance once again, that was "derived from", followed by another short silence.This clearly indicated Muiz's difficulties to give immediate response that caused Int. 2 to further explain his question before Muiz finally admitted that it was a difficult question.
Muiz's tendency to repeat the interviewers' previous utterances was not limited to short words and phrases but also a full phrase as seen in excerpts 6.2(c) and (d) below.In these excerpts, Muiz was asked to name "Menteri Besar of Terengganu" and "Menteri Kewangan Malaysia", respectively.Notice that in the above two excerpts, Muiz's first response to Int.1's questions was a repetition of the latter's preceding utterance (in this context "Menteri Besar Terengganu" and "Menteri Kewangan Malaysia") which indicated the candidate's peculiar way in dealing with questions.His tendency to repeat the questions revealed his language idiosyncrasies.
Here, aside from "[gaining] the speaker time at a lexical selection point" (Faerch & Kasper, 1983, p. 215), repeating others' preceding utterances could also be a speaker's strategy to show that he is following the conversation and comprehends it well.Again, this relates to (Nakatani, 2010, p. 122) concept of "shadowing".If this were true, this strategy signifies respect in communication as this language behaviour indicates that the speaker is taking interlocutor's words seriously.While this behaviour is rather common among the Malays, it becomes one example of how cultural background influences one's language use.

Code Switching
Code switching refers to the inclusion of L1 words and their pronunciations in L2 speeches (Dörnyei & Scott, 1997, p. 189).This strategy occurred at least 31 times in the data of the current study and was noted among the most frequently employed strategy by the candidates.This concurs with many past studies which reported the presence of this strategy in oral data (Omar, 2003;Abdullah, 2002;Tan et al., 2012).
Among all candidates, this strategy was seen the most in candidate Anisah's data with 10 occurrences.This is a relatively greater number than the average of 2 times by the majority of other candidates.The candidate obtained a moderate credit B4 for her English SPM result and was applying for a Bachelor of Mass Communication in Broadcasting.Some occurrences of code switching taken from Anisah's are shown in Table 3. Referring to Table 3, it was found that many cases of code switching in Anisah's utterances occurred at word and phrasal levels which could be caused by a few reasons.One example of code switching at a word level is seen in Excerpt 3.3 (a).Here, the use of the word "ini" was more like a false-start before the candidate was able to proceed with her utterances.Similarly, in Excerpt 3.3 (b), the Malay word "rezeki" was used to replace the word "provision" or "sustenance" which might not come across the candidate's mind, resulting in the sentence "maybe not my rezeki".
A code switching at a phrasal level is seen in Excerpt 3.3 (c) when Anisa was giving a reason why TV1 and TV2 channels were rather boring.In trying to say that this was due to their old-fashioned delivery technique, Anisa switched to her L1 by saying "dia cakap" instead of "the way they speak".Upon realizing that she had code switched, she immediateoy apologized and corrected her phrase by saying "they speak", before continued communicating in English.This however, did not last when she incurred another code switching by saying "dah lama benda tu" to mean "it has been outdated".
One important finding derived from the study is that, many of the "switched" words are not difficult words.Instead, they are the common words used in conversations.For instance, the terms "diaorang" and "selalu" (see Excerpt 3.3 (d)) are the Bahasa Melayu terms for "they" and "always" respectively which occur regularly in daily conversations.When these two words were translated into the candidate's first language, it shows that code switching is more of her habitual language behaviour or idiosyncrasies rather than a strategy employed when the speaker experiences a halt in communication.Perhaps this is the reason why at one point in the conversation, Anisah was reminded by Int. 2 to speak in the English language (see Excerpt 3.3 (e)).Considering the above cases, code switching seems controllable by L2 speakers.Hence, they should be trained not to easily resort to their L1when they are engaged in L2 communication.

Restructuring
The final type of CS employed which indicates the presence of language idiosyncrasies is restructuring.This strategy occurred 35 times in the data and was noted as one of the most frequent strategies in this study.While others might just abandon their messages or omit the unknown words when experiencing a difficulty in communication, candidate Naddia was found to make the most effort to restructure her sentences to ensure that her intended message was still delivered.While the average number of restructuring made by each candidate was two times, Naddia appeared to restructure her sentences at least seven times.Observe the following excerpts that show how these were done.Int.2: So:: have you ever had this so-called, experience of handling the the event (with) this?Naddia: Er:: (yes of course).Every:: er: (.) first semester, they will, they have to:: it is necessary to join the event.So, I:: my experience was when, when:: I was in my:: part, I mean, semester one, I:: I want to:: I was in the protocol for this event.
In Excerpt 3.4 (a), the candidate decided to change her sentence structure from "I really..." to "I'm curious…" before she continued explaining her interest working behind the scenes.Most possibly, the candidate wanted to say "I really want to learn how to conduct programme."However, she then restructured her sentence.
A similar case was identified in Excerpt 3.4 (b).In this excerpt, restructuring occurred at least twice in the interactions before the candidate reached her final sentence.While a change from "they will" to "they have to" could be regarded as a false start, her next sentence "it is necessary to join the event" was obviously a restructuring of her earlier sentence.Similarly, a change from "So, I..." to "my experience was when.." could be a false start.But changing a phrase from "I want to" to "I was in the protocol for this event" looks more like restructuring a sentence if not an abandonment of the earlier message.At this point, retrospective data which this study lacks of would be greatly helpful to ascertain the candidate's intention of adjusting her sentences as shown in the above excerpts.
The candidate's choice to restructure her sentences is seen to greatly relate to her L2 proficiency level.As noted earlier, communicative competency comprises both knowledge of abstract linguistic rules and the ability to use the language appropriately (Hymes 1972).In Nadddia's case, her high competency level was evident by her English SPM result in which she obtained a distinction 2A and her ability to restructure her sentences instead of abandoning them or resorting to her L1 (code switching) when she faced a difficulty in conveying her message.
In terms of cultural values, unlike the earlier three types of CS, restructuring does not seem to relate to cultural values.Extensive use of this strategy however, is a good indicator of language idiosyncrasies in a speaker.

Conclusions and Implications
There are at least two conclusions derived from this study.First, in addition to what is concluded from past studies that the types of CSs employed by L2 learners are influenced by their L2 proficiency levels (Tan et al., 2012;Khanji, 1996;Wannaruk, 2003), L2 proficiency level also influences the way a particular type of CSs is employed by L2 learners.In expressing non-understanding for instance, candidate Anisah who obtained a strong credit B4 for English subject in SPM opted to repeat interviewer's question to indicate her lack of understanding (see Excerpt 3.1 (e) as compared to Shamsuddin who would ask bluntly using an inappropriate expression (see Excerpt 3.1 (a) -(d)) (note that Shamsuddin obtained a weak credit of B6 for his English SPM result).This could not only affect the results of the interviews, it also reflects disrespect which contradicts with Malay cultural values.
Second, candidates' L2 proficiency level does not only relate to their use of CSs but also their politeness in communication.This is proven when Muiz, who obtained a strong credit B3 for English subject in SPM was found to play an active role in communication by repeating interviewers' preceding utterances in order to gain time before he could give his response.This did not only give an impression that he was following the conversation well but also a signal for interviewers that their words were being taken seriously, hence inserting the elements of respect and politeness towards interviewers.The case is different with Shamsuddin who seemed ignorant of the fact that he was dealing with individuals of higher position resulting in expressing "hah?" repeatedly when he did not understand interviewers' questions or statements.As Shamsuddin obtained the lower grade than Muiz in his English SPM result, this reflects that although Malays are commonly perceived as polite people, their L2 proficiency level influences their L2 language use more than their cultural background.
Finally, as this study revealed the existence of language idiosyncrasies rooted from candidates' ways of employing CSs, issues of politeness and appropriateness must be taken seriously in L2 teaching in this country in order to avoid negative implications of language idiosyncrasies on learners' L2 communication.Hence, this study would strongly recommend that the current CLT approach in Malaysian schools place greater emphasis on the social aspects of communication.One way to do this is by raising learners' awareness that they need to consider context and purpose of communication by analyzing some relevant social factors such as culture, social distance and power difference that occur between them and the interlocutors as this helps to enhance effectiveness in communication.
Er:: so, I:: I would like to:: further my [studies] in Mass Com.Erm:: because er::I, I really like broadcasting.Ini (This) I choose broadcasting.I, I got (inaudible) ehem ((clearing throat)):: and then:: I don't know much about broadcasting.Yang saya tahu (As far as I am concerned) which is about:: ermm:: penerbitan sesebuah rancangan (the production of a certain programme), erm:: journalism:: erm: apa (what)?Kewartawanan (journalism).Erm:: working on backstage, in front of stage.So, I hope (that) er, I can, I can improve my >communication< skills when I (.) involved in Mass Com.
Menteri Kewangan Malaysia.Yang Berhormat Dato' Seri Najib, Mohd.>Najib Razak< hhh Excerpt 3.4 (a): (Naddia was explaining about her interest) Naddia: … I really want to learn what (.) is broadcast all about.And I-I really, I'm curious about how to (.) how to er:: how to, do a programme and, to work behind the scenes.Excerpt 3.4 (b): (Naddia was explaining about her experience in organizing an event)

Table 2 .
Categorizations of English language proficiency according to SPM English results

Table 3 .
Examples of code switching in Anisah's utterances