Does Input Enhancement Work for Learning Politeness Strategies ?

The present study investigated the effect of input enhancement on the acquisition of English politeness strategies by intermediate EFL learners. Two groups of freshman English majors were randomly assigned to the experimental (enhanced input) group and the control (mere exposure) group. Initially, a TOEFL test and a discourse completion test (DCT) ensured homogeneity of the groups in terms of general proficiency and pragmatic competence. Then, the participants received the instructional treatments on English politeness strategies (PSs). They read and listened to dialogues containing the intended PSs. The target PSs were in bold typeface in enhanced input group’s dialogue booklet and in regular font in mere exposure group’s booklet. Subsequently, the groups took the DCT posttest. The analysis of the pretest and posttest data revealed the following results: Iranian intermediate EFL learners differed from English native speakers in their use of PSs; there was a low relationship, an insignificant correlation (rxy = .25), between general proficiency and pragmatic competence (i.e. competence in English PSs) of Iranian intermediate EFL learners; and input enhancement had a significantly greater effect on the acquisition of English politeness strategies (both comprehension and use of PSs) than simple exposure to PSs. The study implies that EFL learners should be provided with a specific instruction on English PSs and PSs should be presented in a noticeable way (e.g. typographically enhanced).


Introduction
To acquire the ability to communicate effectively and successfully, language learners need to develop all aspects of communicative competence.Pragmatic competence is an aspect of communicative competence and both Canale and Swain (1980) and Bachman (1990) include it in their models of communicative competence.Pragmatic competence consists of the "knowledge of what constitutes appropriate linguistic behavior in a particular situation" (Ellis, 2008, p. 956).One aspect of pragmatics is politeness, which concerns linguistic forms which language users employ to display respect and consideration for their addressees and to consider role relationship and social status between the interlocutors, the imposition of the speech act, and other social factors, such as age and sex.Linguistic politeness "has generally been considered the proper concern of pragmatics" (Holmes, 2006, p. 711).Richards and Schmidt (2002) define linguistic politeness as "(a) how languages express the social distance between speakers and their different role relationships; (b) how facework, that is, the attempt to establish, maintain, and save face during conversation, is carried out in a speech community" (p.405).According to Holmes (2006), linguistic politeness "is a matter of specific linguistic choices from a range of available ways of saying something" (p.711).For instance, expressions like Sir, Would you mind if …, and I was wondering if ... may be used to make a request more polite and more appropriate when speaking with a stranger or one's manager.
Research in pragmatics and interlanguage pragmatics (ILP) has generally shown that languages vary in their pragmatic features; language learners, regardless of their proficiency levels, differ from native speakers; instruction has a positive effect on learning pragmatic features; and explicit teaching has a greater effect on pragmatic development than implicit instruction.However, despite ample research in ILP, mostly investigating language learners' awareness and use of different speech acts, there has been a paucity of research on politeness strategies in second language pragmatics.Few studies have explored second or foreign language learners' knowledge and acquisition of politeness strategies.The present study pertains to linguistic politeness and aims to investigate Iranian intermediate EFL leaners' awareness and use of English politeness strategies (PSs), relationship between general proficiency and competence in English PSs and, most importantly, the effect of input enhancement on the acquisition of PSs.To that end, the following research questions were proposed.RQ1.Is there any significant difference between Iranian intermediate EFL learners and English native speakers in the range, frequency and types of politeness strategies they use?RQ2.Is there any significant relationship between Iranian intermediate EFL learners' general proficiency and their competence in English PSs? RQ3.Is there any significant difference in the effects of mere exposure and input enhancement on intermediate EFL learners' awareness of PSs? RQ4.Is there any significant difference in the effects of mere exposure and input enhancement on intermediate EFL learners' use of PSs?

Literature Review
Studies comparing pragmatic behavior of nonnative speakers and native speakers have mostly indicated that learners differ from natives in their use and recognition of pragmatic features, including politeness strategies (Olshtain & Blum-Kulka, 1985, as cited in Ellis, 2008;Eisenstein & Bodma, 1986;Schmidt, 1994).Bardovi-Harlig & Hartford (1993, as cited in Bardovi-Hatrlig, 2001) studied the pragmatic performance of nonnative speaker (NNS) and native speaker (NS) students in the context of academic advising sessions.NNSs usually did not employ the mitigators used by the NSs to soften their rejection of the advisors' suggestions and they often used aggravators, which were never used by the NS peers.NNSs have also been shown to differ from NSs in the use of routines or "typical expressions", such as "Could you ………?" and "How clumsy of me, ………", which make the speech act or the semantic formula immediately recognizable to the hearer (Scarcella, 1979;Takahashi & Beebe, 1978).Some studies have explored teachability of pragmatic rules, including politeness strategies, most of which indicate that pragmatic features are teachable; language learners who received instruction on pragmatic features outperformed those who received no instruction on, or only mere exposure to, these features (Kubota, 1995;Lyster, 1994;both cited in Ellis, 2008;Billmyer, 1990;Eslami-Rasekh, Eslami-Rasekh & Fatahi, 2004;Yoshimi, 2001).Eslami-Rasekh et al. (2004) investigated the effect of explicit teaching of request, apology and complaint speech acts on Iranian advanced EFL learners' comprehension of these speech acts.The results of the study manifest that pragmatic competence is not impervious to instruction.Takahashi's (2010) meta-analysis of 49 pragmatic interventional studies revealed that intervention has the potential to enhance pragmatic knowledge of language learners.However, some studies provide the evidence to the contrary and indicated that instruction has no significant effect on the acquisition of pragmatic rules (King & Silver, 1993;LoCastro, 1997).
Another group of studies investigated the influence of different teaching methods on the learning of pragmatic features and the majority of the studies indicate that explicit instruction of pragmatic features lead to a higher level of acquisition than implicit teaching (House, 1996;Koike & Pearson, 2005;Rose & Kwai-fun, 2001;Salemi, Rabiee & Ketabi, 2012;Takahashi, 2001).House (1996) found that the advanced learners of German in the explicit teaching group outperformed the learners in the implicit teaching group in the areas of discourse strategies, speech acts and gambits.Takahashi's (2010) meta-analysis also revealed that explicit intervention seems to be more effective than implicit instruction.However, some studies have proved the opposite, indicating that explicit instruction is not significantly more effective than implicit teaching (Takimoto, 2006;Tateyama, 2001).Tateyama (2001) compared the implicit and explicit instruction of formulaic expressions for some speech acts and found no difference between the two types of instructions.
The effect of input enhancement on pragmatic development has provided mixed results.For instance, Fukuya and Clark (2001) did not demonstrate any significant effect of input enhancement on pragmatic development (knowledge of mitigators in requests); whereas, Vahid Dastjerdi and Rezvani (2010) showed that input enhancement technique exerts a significant effect on learners' pragmatic development; they also found no significant difference between the participants who received explicit instruction and those who received enhanced input in their production of request strategies in English.Martı´nez-Flor (2004) demonstrated that the combination of input enhancement and recasts could be as effective as explicit instruction (teachers' explanation) in development of production and comprehension of the speech act of suggesting.
In short, research has generally shown that teaching methods which are more explicit and involve more noticing have greater effect on the acquisition of pragmatic features (e.g., politeness markers) than implicit teaching methods.As a result, input enhancement seems to be more effective than mere exposure, since the former involves more consciousness-raising and noticing.However, sufficient research is required to make the claim.

Participants
The participants of the study were 41 second-semester English majors, mainly female, in the 18-24 age range.The participants were from two intact university classes, which were randomly assigned to the control (mere exposure) group and the experimental (enhanced input) group.The control group comprised 21 participants and the experimental group included 20 participants.The students were considered as intermediate language learners.

Instruments
The instruments of the study included a general proficiency test (TOEFL test, 2002, excluding the listening and writing sections); two discourse completion tests (DCTs), one as the pretest and the other as the posttest, each including a written discourse completion test (WDCT) and a multiple-choice discourse completion test (MDCT); and the materials for the treatment of English PSs.

Proficiency and DCT Tests
The TOEFL test (2002) was used to measure the participants' general English proficiency and ensure the homogeneity of the two groups prior to the study.The listening and writing sections of the TOEFL test were excluded due to practicality issues.The pretest and posttest were two separate researcher-made DCT tests, which included both WDCTs and MDCTs.The pretest and posttest each consisted of 12 WDCT items and 12 MDCT items (see Appendix).The described situation in each DCT item required realization of an appropriate politeness strategy and the participants were asked to identify appropriate PSs to use or choose.The DCT tests were developed and validated with the assistance of 22 English native speakers, the majority of whom were American NSs in the 20-30 age range.First, two NSs read and revised the first version of the two DCT tests.Then the revised versions were administered among 20 NSs and their responses to DCTs were used as baseline data for the evaluation of the participants' performance.The reliability analysis carried out on the TOEFL test, MDCT pretest and posttest scores assured that the tests were reliable measures (Cronbach's Alpha for the TOEFL test, the pretest and the posttest were .84,.68 and .71respectively).

Instructional Materials
For the development of the instructional materials, some English course books, conversation books and the internet were searched to find dialogues containing the intended PSs.Fifty such dialogues were selected and used in seven lessons.There were seven dialogues in each lesson, except the last lesson which comprised eight dialogues.Each dialogue was accompanied by its audio file, which was played for the participants as they were reading the dialogues in their booklets.The target politeness strategies were in bold type in input enhancement group's dialogue booklet and in regular font in the mere exposure group's booklet.

Procedure
First, to ensure homogeneity of the two groups in general proficiency and pragmatic competence, a TOEFL test and a DCT pretest were given to the participants.A week later, the participants started to receive the instructional treatment of the intended PSs, which consisted of seven sessions and lasted seven weeks.The participants in the two groups read and listened to the same dialogues and, after each dialogue, the teacher explained the new grammatical and vocabulary points, if necessary.The participants in the enhanced input group observed the PSs in bold font while the participants in the mere exposure group observed them in regular font.A week later, the participants took the DCT posttest and the analysis of the results revealed the effects of the two teaching methods (mere exposure and input enhancement) on the acquisition of English PSs.

TOEFL Test Results
The independent samples t-test analysis of the TOEFL test results (observed t = .099,df = 39, p = .922)indicated that there was no significant difference between the mere exposure and enhanced input groups and the uniformity of the two groups in general proficiency was confirmed.

Natives and EFL Learners
The Chi Square analysis comparing the range and frequency of PSs used by the native speakers and EFL learners on the WDCT pretest revealed that there was a significant difference between the natives and learners in terms of the number of PSs they used (x 2 (1) = 14.90, p = .000),though there was no significant difference between the two groups in the rage of employed PSs (x 2 (1) = .004,p = .949).The NS participants used significantly more PSs than the Iranian EFL learners.The natives used 69 PSs more than expected while the learners used 69 PSs fewer than expected (table 1).Moreover, the learners were different from the native speakers in the types of PSs they frequently employed.The native speakers used the following politeness strategies twice to four times as many as the learners: terms of address, intensifiers, appealers, understaters, suggestory formula, pseudoconditionals, disarmer, apology, past tense, and query preparatory.Whereas, the learners used the following PSs more frequently than the NSs: imperatives, subjectivezers, politeness markers (please), and Yes/No question.There were also some cases of L1 transfer in the learners' responses to WDCT items.The learners translated common Farsi sentences and expressions into English and used them when responding to the described scenarios.
The following are some examples: 1) Is it possible for you to put a time for me to work on test?
2) Let me change it.(After spilling a friend's cup of coffee.) 3) The grass is grown under my foot, why are you so late?4) Let me tell to bring you another coffee.
Also, there were some cases where the learners lacked sufficient grammatical knowledge to use appropriate politeness strategies.The following examples demonstrate this fact.
1) I will be wondering if you help me to get ready for the test.
2) Would you mind if I leaving the work place earlier?
3) I was wondering if you order another thing.

4) Would you mind pass the remote?
Last but not least, the learners rarely or never used the routines or formulaic expressions that were frequently employed by the natives.The following are some examples: 1) Would you mind if I sat here?
2) Sorry, I'll get you another coffee.
3) Let me get you another one.(After spilling a friend's cup of coffee.)4) Could you take our order please?We've been waiting for a long time.
The comparison of the frequency and types of PSs used by the native speakers and Iranian EFL learners revealed that their pragmatic behavior was significantly different.Therefore, the first research question was answered by rejecting the null hypothesis that there is no significant difference between Iranian intermediate EFL learners and English native speakers in the range, frequency and types of politeness strategies they use.This finding is in line with the results of the majority of the studies comparing NSs and learners.It confirms that pragmatic behavior of language learners differ from that of native speakers and learners may not have the required pragmatic knowledge to display politeness and express social distance and role relationship between the interlocutors.

The Experimental and Control Groups
The independent samples t-test analysis of the mere exposure and enhanced input groups' MDCT pretest scores (observed t = .053,df = 39, p = .958),the p value of which considerably exceeded .05,revealed that there was no significant difference between the groups in their knowledge of English PSs.Also, the results of the Chi Square analysis of the range (x 2 (1) = .309,p = .579)and frequency (x 2 (1) = 1.37, p = .240) of PSs used by the two groups on the WDCT pretest revealed no significant difference between them.Therefore, the two groups were shown to be homogeneous in the awareness and use of PSs prior to the instructional treatment.

Correlation Analysis
Subsequently a Pearson correlation analysis was carried out on the participants' TOEFL test and MDCT pretest results.The analysis indicated a low relationship, a statistically insignificant correlation (rxy = .25,p = .112),between the general proficiency and pragmatic competence (competence in English PSs) of the Iranian intermediate EFL learners.Hence, the second research question was answered by confirming the null hypothesis that there is no significant relationship between Iranian intermediate EFL learners' English general proficiency and competence in English PSs.This further confirms the notion that grammar proficiency does not guarantee pragmatic competence and learners should be provided with specific instruction on pragmatic features.

Posttest Results
Eventually, the performance of the control (mere exposure) and experimental (enhanced input) groups on the posttest was compared to investigate the effect of input enhancement on the acquisition of PSs.Table 2 displays the descriptive statistics of the two groups' performance on the MDCT posttest.The results of the independent samples t-test comparing the groups' MDCT posttest scores (observed t = 2.14, df = 39, p = .03)indicate that the enhanced input group significantly outperformed the mere exposure group.This reveals that the enhanced input group had a greater awareness of English PSs after the treatment, although the two groups were virtually the same on the MDCT pretest.Therefore, the third research question was answered by rejecting the null hypothesis that there is no significant difference in the effects of mere exposure and input enhancement on Iranian intermediate EFL learners' awareness of PSs.The Chi Square test comparing the range of PSs used by the mere exposure and enhanced input groups revealed no significant difference between them (x 2 (1) = 1.41, p = .285).However, the Chi Square analysis comparing the number of PSs used by the two groups on the WDCT posttest indicated a significant difference between the enhanced input and mere exposure groups (x 2 (1) = 5.21, p = .02).The enhanced input group significantly outperformed the mere exposure group and used a significantly greater number of PSs on the WDCT posttest, although the two groups had used almost the same number of PSs on the pretest.
Moreover, there was a difference in the types of PSs used by the two groups on the WDCT posttest.The enhanced input group used the following PSs twice to several times more than the mere exposure group: consultative devices, pseudoconditionals, grounders, politeness markers, May I …, and Would it be ok if …..The mere exposure group made more use of imperatives and address terms.The enhanced input group was closer to the native speakers in the types of PSs they used.Thus, the last null hypothesis that there is no significant difference in the effects of mere exposure and input enhancement on Iranian intermediate EFL learners' use of PSs was rejected and the last research question was answered.
Since the two groups were almost the same in their performance on the pretest, their diverse performances on the posttest could be attributed to the instructional treatments.The participants in the enhanced input group performed significantly better on both MDCT and WDCT posttests, which suggests that input enhancement method had a significantly greater effect on learners' awareness and use of PSs, i.e., their acquisition of politeness strategies.

Pedagogical Implications and Further Research
The study indicated that intermediate EFL learners differ from native speakers in their use of politeness strategies; there is a low relationship (a statistically insignificant correlation) between grammar proficiency and competence in politeness strategies; and input enhancement has a significantly greater effect on the acquisition of PSs (both knowledge and use of PSs) than mere exposure.The first implication of the study is that language learners should be provided with instruction on politeness strategies, as their knowledge and use of PSs is far from appropriate and grammar proficiency does not guarantee competence in PSs.The second implication is that PSs are more effectively learned when they are presented in a more highlighted and noticeable manner (i.e., in bold type), as they may attract more attention from the learners and lead to more acquisition.Exposing learners to PSs in regular font is less likely to attract learners' attention and help them to learn these pragmatic features.Therefore, the suggestion is that PSs should be taught in a more conspicuous way, so that learners have more chance of noticing and learning them.
Finally, every study is doomed to some limitations and the present research was not an exception.Therefore, the generalizations should be made with caution and further research is required to overcome limitations.First, the study was limited to intermediate level of proficiency (freshman students), so the findings are applicable to learners at this level.Further research is required to investigate the effect of pragmatic instruction on learners at other levels of proficiency (e.g.beginners and advanced learners).Second, the present study investigated 30 PSs, which were mainly lexical and syntactic PSs; further research can explore the effect of pragmatic instruction on the acquisition of the other 15 PS, which are mainly discourse level PSs.Finally, the present study was limited to two teaching methods (mere exposure and input enhancement); interested researchers can investigate the effect other teaching methods, such as explicit teaching and classroom discussion of PSs, on the acquisition of these pragmatic features.

Table 1 .
Frequency of PSs used on the WDCT pretest

Table 2 .
Descriptive statistics of the groups' MDCT posttest scores