The Effects of Explicit Instruction on EFL Students’ Production and Perception of Requests

As an exploratory study to expand the scope of pedagogical intervention to the teaching of pragmatics, the present research sought to examine the effects of explicit instruction on the learning of request forms. The effects were examined in two respects: the learners’ strategies in formulating request forms, and their perceptions of the pragmatic intervention. To this end, 50 third-year English students at the faculty of Arts and Humanities Sfax, Tunisia were exposed to explicit instruction in the form of awareness-raising tasks and written meta pragmatic feedback on the use of appropriate requests. They were required to take a pre-test, receive a one-month treatment, and take a post-test. In line with previous research, the study showed that learners’ production of requests benefited from the explicit instruction and the questionnaire data indicated that the explicit teaching has helped the learners understand the appropriate use of request strategies.


INTRODUCTION
In order to be successful in communication, it is essential for foreign and second language learners to know the grammar of the language, the text organization as well as the pragmatic aspects of the target language. The failure a speaker experiences in conveying an intended message is called pragmatic failure. In Thomas" (1983) [45] terms, "Pragmatic failure" refers to the speaker"s production of wrong communicative effects through the faulty use of speech acts or one of the rules of speaking. Thomas (1983) [45] uses the term "pragmatic failure" to refer to the inability of the individual to understand what is meant by what is said. In her description of pragmatic failure, Thomas distinguishes between two types of failure: "pragmalinguistic" and "sociopragmatic". Bachman (1990) [3] devised a "language competence" model (figure 1 below) subdivided into "organizational competence" and "pragmatic competence". Organizational competence comprises "grammatical competence" and "textual competence." Pragmatic competence includes "illocutionary competence", that is, knowledge of speech acts and speech functions, and "sociolinguistic competence", that is, the ability to use language appropriately in context. In Bachman"s model, pragmatic competence is not linked to knowledge of grammar and text organization but is tied to formal linguistic and textual knowledge. Pragmatic competence is specifically defined by Koike (1989) [26] as "the speaker"s knowledge and use of rules of appropriateness and politeness which dictate the way the speaker will understand and formulate speech acts" (p.279).

PRAGMATIC INSTRUCTION
In the fields of Interlanguage and Cross-cultural Pragmatics, previous studies have investigated different features of pragmatic acquisition as reviewed and classified by Kasper (2001) [23] such as discourse strategies (House and Kasper, 1981) [21], conversational routines (Tateyama, Kasper, Mui, Tay and Thananart, 1997;Tateyama, 2001) [44]. Additionally, various speech acts have been analysed including compliments (Billmyre, 1990 [8]; Roe and NG Kwai-fun, 2001 [39]), requests (Safont, 2003 [40]; Fukuya and Zhang, 2002[19]), apologies (Olshtain and Cohen, 1990[34]), refusals and complaints (Morrow, 1995) [33]. These studies were conducted in both ESL and EFL learning contexts. Regarding Pragmatic Instruction, previous studies from SLA (Long, 1983(Long, , 1993(Long, , 1996[29] [30]; Larsen Freeman, and Long, 1991[27], Schmidt, 1993[41]) have shown that instruction positively affects acquisition when compared with mere exposure to the target language. Additionally, studies that have attempted to teach pragmatic features of the L2 (Alcón, 2005 [38]). Several studies (Rose, 1994[35]; House, 1996[20]) have focused on developing awareness of pragmatics through explicit instruction. Their results generally supported the notion that pragmatic development can be enhanced through explicit awareness-raising techniques. Many other studies have compared implicit and explicit instruction of pragmatics in the ESL and EFL settings (Eslami-Rasekh, 2005 [18]; Fukuya and Zhang, 2002[19]; and Tateyama, 2001 [43]) found explicit instruction of pragmatic features to have a greater effect on pragmatic competence than implicit instruction. Operationally, explicit instruction has shown significant effects through a wide range of classroom activities that focus on description, explanation, discussion of pragmatic features as well as practice of those features. In consideration of such positive findings from previous research, it transpires that explicit instruction does in fact assist pragmatic development in second language learners.
The present study hopes to contribute to the large body of research in ILP by examining the effects of explicit instruction on the production and perception of a particular speech act that of requesting by learners of English as a Foreign Language in an instructional setting. The main aim is to ascertain the role of instruction in producing requesting strategies. The present study unfolds with a review of the literature most relevant to the current study. Then it describes the research methodology used, including the participants, the instruments and the study design. It concludes with the findings and discussion of the results. Austin (1962) [2] defines speech acts as acts performed by utterances such as giving order, making promises, complaining, requesting, among others. When we utter a sentence or a phrase, we are performing an act to which we expect our listeners to react with verbal or nonverbal behaviour (p.65). For Searle (1969) The speech act of requesting, which is the focus of the present study, falls into the category of directives. The illocutionary point of directives is to direct the hearer to do something, generally for a speaker"s goal. For Searle (1969) directives have been regarded as "an attempt to get hearer to do an act which speaker wants hearer to do, and which it is not obvious that hearer will do in the normal course of events or of hearer"s own accord" (p. 66). In addition, requests are "face-threatening acts (FTAs)" in which the speaker is imposing his will on the hearer (Brown and Levinson, 1987[12]: 65). They suggest that performing a FTA calls for a high level of appropriateness and linguistic knowledge on the part of the speaker. As such, he must choose between performing the FTA in the most direct and efficient manner or mitigating the effect of the FTA on the hearer"s face (p. 76). Blum-Kulka, House, and Kasper (1989) [11] studied the pragmatics of the request speech act and distinguished three degrees of directness in requests: direct requests, conventionally indirect requests, and non-conventionally indirect requests. The request strategies in the Cross-Cultural Speech Act Realization Project (CCSARP, Blum-Kulka et al., 1989 [11]) classification are ordered according to the degree of directness. In table 1 below, Request strategies are ordered on a scale, from most indirect (hint) to most direct (imperative) (Blum- Trosborg, 1995). The basic notion is that speakers choose the appropriate strategy from the scale according to their judgment regarding Power and Distance (Brown and Levinson, 1987) [12]. Power is defined as "the degree to which the hearer can impose his own plans and his own self-evaluation (face) at the expense of the speaker"s plans and self-evaluation" (p. 77). Distance refers to the "symmetric social dimension of similarity/difference within which the speaker and hearer stand for the purpose of an act and the kinds of goods exchanged between them (p. 76). Eslami-Rasekh (2005) [18] contends that the speech act of request is the most empirically-explored speech act in the cross-cultural or interlanguage pragmatics literature. Nonetheless, in the Tunisian context, no studies have been undertaken to date, which investigated the teachability of pragmatics to EFL learners. To fill in some of the existing gaps in ILP research in the Tunisian context, the present study seeks to examine the teachability of pragmatics with regard to the area of speech acts and more specifically, the speech act of requesting which could easily be incorporated into explicit metapragmatic instruction. As such, the present investigation may add another dimension to ILP instruction in the Tunisian EFL classroom.

IV Direct requests
Obligation You must/have to lend me your computer.

Performatives (hedged)
I would like to ask you to lend me your computer.
(unhedged) I ask/require you to lend me your computer.

Imperatives
Lend me your computer

Elliptical phrases
Your computer (please)

MOTIVATION FOR THE STUDY
The present work is an attempt to contribute to the growing literature in ILP research with data from Tunisian learners of English. To my knowledge, there are no existing studies focussing on the the teachability of the directive speech act of request in the Tunisian EFL context. Therefore, this research will hopefully fill a gap but it can not hope to cover all the issues involved in the development of the request speech act in Tunisian learners" interlanguage. It aims at being exploratory, and this is the overall purpose of this research. At a specific level, it examines what EFL learners say in the given request situations. It compares the learners" requesting strategies before and after explicit instructional treatment. It also looks at learners" perceptions of the instructional treatment, the request strategies and the tests" content

RESEARCH QUESTIONS
This study seeks to investigate the following research questions: 1. How is learners" production of the speech acts under investigation affected by explicit instruction to be measured by pre-and post-tests? 2. How do learners react to the use of explicit instructional techniques?

RESEARCH HYPOTHESES
The study attempts to check the validity of the following hypotheses: H 1 : Learners will significantly improve their production of pragmatically appropriate requests in the post-test over the pre-test. H 2 : Learners will positively react to the use of the explicit pragmatic instructional treatment to produce appropriate speech acts.

Participants
50 third year students at the University of Letters and Humanities in Sfax who had taken one Pragmatics course during the first semester were the subjects of the study. The selection of this sample took into consideration the convenience of the researcher and the accessibility of the students, thus meeting the criteria of "convenience or opportunity sampling" (Mackey & Gass, 2005[31]; Dornyei, 2007 [17]). It is worth mentioning that all participants" approval was provided through informed consent forms.

Data Collection
Data for this study was collected through Written Discourse Completion Tests (WDCTs) of eight scenarios (See Appendices A&B): four administered before treatment and four after treatment. The pre-and post-tests sought to identify students" performance after being exposed to the meta-pragmatic instruction (instructional material). The situations focus on requesting. The design of the DCT used in this study is based on the discourse completion tasks developed in the area of crosscultural pragmatic studies (e.g., House and Kasper, 1981[21]; Blum-Kulka et al., 1989 [10]). The content validity of the instrument was established by having the situations evaluated by two Native American teachers.

Written discourse completion test
WDCTs are written questionnaires including a number of brief situational descriptions, followed by a short dialogue with an empty slot for the speech act under study. WDCTs take different shapes: with or without a rejoinder. This is an example of a WDCT with a rejoinder taken from Blum-Kulka & Olshtain (1984) [9]: At the professor"s office A student has borrowed a book from her teacher, which she promised to return today. When meeting her teacher, however, she realizes that she forgot to bring it along. The WDCT made use of in the present study consists of descriptions of speech act situations followed by incomplete discourse sequences that the informants are requested to complete. This is an example of a prompt used in the written production tests: You are a student. You forgot to do the assignment for your Drama course. Before class, you decide to go to your teacher whom you know personally very well to ask his/her permission to bring the assignment at the end of the day. What would you say? You:……………………………………………………… …………………………………….……………………… ……………………………………………………………

Questionnaire
In order to elicit more information about their perceptions of the treatment, the instructional material, and the tests, students were also asked to respond to a questionnaire in the form of a student evaluation form. All the items in the questionnaire were designed for a Likert scale response using a four-interval scale of "easy", "very easy", "difficult", and "very difficult" (See Appendix C).

Treatment and Material
According to Rose (1999) [37], successful instruction of speech acts entails using pragmatic consciousness-raising (PCR) techniques. The purpose of the PCR is "to expose learners to the pragmatic aspects of language… and provide them with the analytical tools they need to arrive at their own generalizations concerning contextually appropriate language use" (Rose: 171). Rose further explains that PCR is a process in which pragmatic awareness is raised first by introducing students to a particular pragmatic feature, then by activating students" L1 knowledge of the pragmatic feature (Rose, 1999:180) [37]. As for Kasper (1997) [22], meta-pragmatic instruction might be combined with meta-pragmatic discussion with the active participation of students in various forms of teacher-fronted-format, peer work, small groups, role-plays, semi-structured interviews, introspective feedback, and meta-pragmatic assessment tasks. Rose (1999) [37] and Kasper (1997) [22] approaches to raising pragmatic awareness have guided the treatment design of the present study.
Based on a thirty sample pragmatics lessons including assorted speech acts which were collected by Bardovi-Harlig & Mahan-Taylor (2003) [6] and Cohen & Ishihara"s (2005) web-based approach to strategic learning of speech acts, the researcher prepared the materials to conduct the explicit treatment. The study included a set of programmed instructional materials explaining the realization patterns, rules and strategies of the request speech act intended to help English learners use socially appropriate language in a variety of informal and formal situations. The materials compiled started with presenting descriptions of specific points of speech acts including "Speech Acts", "Requesting", "Request Strategies", "Request Moves", "Request Head Acts", "Request Supportive Moves", and "Request Mitigators / Upgraders." The researcher gave each learner handouts with key terms defined with appropriate examples to ensure that the explanation was uniform to all learners.
As such, the four-session treatment was designed, organized, planned, and given to the subjects by the researcher. It should be mentioned that the pragmatic instruction took about 55 minutes each class period. This instructional treatment adopted a sequential method consisting of presentation of the selected listening excerpts, followed by the scripted versions of the excerpts, a set of explicit awareness-raising tasks, and discourse completion dialogues elaborated on the basis of the type of requests presented in the listening. Finally, the answers to the activities, together with written metapragmatic explanations on the use of appropriate requests were given to the learners. The different instructional activities used are listed below:  The explicit metapragmatic instruction began by a teacher-fronted discussion of various notions related to the request speech act, emphasizing the fact that a specific form can have several functions in the language and one function can be realized through different forms depending on contextual variables. During the instructional treatment, students were exposed to listening excerpts with a focus on requests. The excerpts included direct requests, conventionally indirect, and non-conventionally indirect requests. Then, students were provided with the scripts of the listenings and were asked to identify the phrases used to ask people to do something. They were also asked why they thought particular linguistic formulae were used. The aim of this task was to measure their awareness of requests. Secondly, students were asked to write a dialogue where somebody was making a request. The purpose of the task was to introduce the learners to the request realisation strategies taking into account two variables: social distance and power (Brown and Levinson, 1987) [12].
 Students were divided into different groups and asked to come up with examples of the target speech acts (requests) and to discuss the differences and similarities in the realization patterns of the speech act.  At the end of the instructional period participants were asked to individually write similar dialogues to the ones used in class and to explain the type of language used in making the requests. Students were asked to perform role plays of the intended speech acts for the whole class. These tasks were used to measure individual achievement as a result of explicit instruction.

Procedure
The study was carried out in the second half of the second semester and the experimental treatment was incorporated as a part of the curriculum. The written discourse completion data were collected in two classroom sessions during which participants were asked to provide a response that they think is appropriate in the given context. The situations were carefully designed to facilitate participants" identification with the roles they had to play. All the situations were student-life oriented in order to reflect subjects" real-life experiences and to ensure naturalness of data as much as possible. The participants completed the pre-and post-tests and filled in the questionnaire in the presence of the researcher.
To find out about students" production of the speech act situations, the following 8 situations were administered as shown in Table 2 below: 4 situations in the pre-test and 4 situations in the post-test. The request situations used in the present study varied systematically according to the interactants" relative power (P) and social distance (D). The power variable refers to the status relationship between the interactants where the hearer is either of higher status [+P], e.g. teacher/student, the interactants are equal in status [=P], e.g. student/student, or the hearer is of lower status [-P], e.g. senior student/ junior student. The distance variable refers to the interactants" degree of closeness. They either knew each other [-D] or did not know each other [+D].

Situation 4 [=P, -D]
Borrowing lecture notes Borrowing a book Data on the learners" production of the request speech act took into consideration only the head acts which were isolated and classified based on the levels of directness developed by the CCSARP project (Blum- Kulka et al., 1989: 17-19) [11], adopting Trosborg"s (1995) [46] modification with regard to the integration of strategy types and perspective which were presented in table 1 above. The frequency analysis was conducted to identify the proportion and percentage of request strategies used by the participants in the pre-and post-tests.
Quantitative data concerning the participants" perceptions of the instructional treatment, the request strategies and the pre-and post-tests" content was collected through a questionnaire in the form of a student evaluation form. It sought to identify students" perceptions by giving them a five-item questionnaire.
The content validity of the instruments was established by having the items evaluated by two Native American teachers. Data obtained from the DCTs and the questionnaire were analysed using percentages. Results obtained from the discourse completion tests and the questionnaire were presented in tables and figures. It is worth mentioning that only 42 out of 50 cases of students" responses were taken into consideration in the analysis of data. The remaining eight were excluded from the results data because 5 students did not attend the post-test and 3 copies of the questionnaire were returned incomplete.

RESULTS AND DISCUSSION
Findings related to the learners" production data and their perception data were analysed, discussed and presented in this study segment.

Strategies used in the Head Act
In the present study, the participants" written responses (elicited by the pre-and post-tests) of request expressions were compared in terms of the types of strategies used in the Head Act portion. Using these formulas, the written responses were analyzed by order and frequency. The coding methods used by the CCSARP project have been found effective in categorizing learners" expressions used for requests (Blum-Kulka et al., 1989) [11]. Results of the data analysis are displayed in figures 2 and 3. As displayed graphically in figures 2 and 3 above, request strategies used in Head Acts varied between the pre-test and the post-tests. First, the types of strategies used were slightly different. Figure 2 shows that want statements is by far the most preferred strategy choice in the pre-treatment test, accounting for 30.9% of all requests. If we exclude hints, which were entirely absent from the data, the three most frequently used strategies are the most direct. They account for almost 83% of all strategies used in the pre-test. This figure is much higher than the result of the post-test that showed that direct strategies constituted only 41% of the data. Following Brown and Levinson (1987) [12], direct requests appear to be inherently impolite and face-threatening because they intrude in the addressee"s territory (p. 17). This explains that learners made the requests employing non-target forms in the pre-test, where as in the post-test they employed appropriate target forms. In that, more than 80% of the request strategies employed in the pre-test included direct strategies used for both students and professors. However, in the post-test, learners employed a variety of different target forms addressed during the instructional treatment. This can be observed in their responses in the post-test which included two appropriate forms (i.e. strong and mild hints) that were not found in the learners" production data in the pre-test. Although strategies from Want statement category (e.g., "I want to borrow your computer"), Explicit performative category (e.g., "I am asking you to lend me your computer") and hedged performative category (e.g., "I ask you to lend me your computer") were found in both tests, query preparatory category (e.g., "could you lend me your computer") and Strong hint categories (e.g., "I forgot my laptop") were used mainly in the post-test. Strategies from strong hint, and mild hint categories did not appear in the pre-test data.
A second difference between the two tests was the frequency of strategies from each category (see figure  4). Some strategies were used more often for the pre-test, and vice-versa. The pre-test contained more direct strategies than the post-test. Furthermore, when we compare the participants" responses within the same test across different situations, it appears that the participants tended to use more direct strategies for friends than for professors, and more indirect strategies for professors than for friends in the post-test responses, but not in the pre-test responses. This tendency in the post-test data was closer to the requesting behaviour reported in previous research (e.g., Blum-Kulka and House, 1989[11]; Tateyama (2001) [43], and Eslami-Rasekh (2005) [18].
As illustrated in figure 4, the conventionalised request forms were the most frequent strategy employed by the students in the post-test which was mostly conveyed by only one indirect sub-strategy: the query preparatory (44%). Almost exclusively this strategy constituted about half of all requests produced by the learners in the posttests responses. Following Brown and Levinson (1987) [12], higher levels of indirectness may result in higher levels of politeness.
It is important to note that in the present study non-conventionally indirect strategies are the least made request types. Indeed, strong and mild hints were missing in the pre-test data (0%) and vaguely used in the post-test data (10%). The underuse of this strategy by the students might be due to the highly inferential nature of the nonconventionally indirect request strategies which require an effort from the hearer to deduce the speaker"s intention. Besides, students are probably not sufficiently equipped to use this type of requesting that is considered the most complex indirect strategy.
In general, the results of this study indicated that after being exposed to explicit instruction, learners" use of direct requesting strategies decreased and at the same time conventionally and non-conventionally indirect types of requesting increased (see figure 4). It is possible that the greater frequency of request strategies in the post-test responses constitutes evidence of language change. This result provides another piece of evidence to support the claim that instructional intervention affected the participants" production of the request speech act positively. In this way, we may assume that instruction was effective in providing learners with greater variety of appropriate forms to be used when making requests.

Internal modification strategies
The only types of internal modifiers used within the Head Act and reported in the result data of the present study are the conditional syntactic downgraders "could" and "would", and the politeness marker "please" which were used by the majority of the participants. The participants used fewer internal modifications than reported in other studies (Matsumura, 2001and Safont, 2003 [32], probably because their proficiency on average was not high enough to allow them to use a variety of modifications. In the relatively infrequent use of internal modifications, however, some variation was still observed across different situations, as well as across different methods. The participants tended to use the politeness marker "please" more often in both tests. The participants may have felt more need to mitigate the tone of their requests when they actually interacted with an interlocutor. They also used the conditional downgraders (could/would) more often when the degree of imposition was greater. This suggests that the participants were sensitive to differences in the impositive force of the request, showing a tendency to soften its force.
The findings of the study, thus, seem to prove the effectiveness of instruction, since subjects significantly improved their use of pragmatically appropriate requests in the post-test as compared to the pre-test (see figure 4 below). This result is in line with previous research that has focused on the effects of instruction on the production of particular speech acts (Olshtain & Cohen, 1990[34]; Morrow, 1995[33], Safont, 2003[40], Eslami Rasekh, 2005 [18]).

Students' perceptions of the explicit treatment
To find out about students" views on the instructional treatment, the request strategies and the pre-and post-tests" content, the following 5 items were analysed as shown in Table 3. Also, the responses were merged into two: very easy and easy for easy responses, and difficult and very difficult for difficult responses ( Table 3). The perception questionnaire responses were evaluated in terms of students" views on the instructional treatment (the lessons and the pragmatic instruction), the pre-and post-0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Frequencies Strategies tests and the request strategies presented in the different stages of the treatment.  Table 3 shows that the highest number of students (66.66%) indicated that the request situations used in the post-test are much easier than the ones employed in the pre-test as a result of the instructional treatment. This is followed by many students opting for difficulties encountered when exposed to the instructional treatment and the request strategies respectively: (59% of the students reported having difficulties with the pragmatic instruction and 64.28% having trouble with the request strategies). These difficulties probably came about due to learners" inability to comprehend key notions they did not come across before. Besides, the complex nature of some pragmatic categories, like strong and mild hints, necessitates a high proficiency on the students" parts to understand and convey such notions. From table 3 it can be seen that approximately half of the respondents found that the pragmatic lessons easy (52.3%) and around 47.6% of the students found them difficult. This difficulty could be explained by the fact that learners were not going to be tested on these lessons as they did not count toward the final grade in the course which reduced their interest and seriousness.
Even though more than half of the respondents indicated that the instructional treatment and the request situations were difficult, these problems did not affect their performance in the post-test treatment. In fact, more than half of the learners found the lessons and the post-test strategies easy and helpful. More importantly, students" answers were significant because they help shed more light on the findings of the present study which give further evidence to Eslami Rasekh"s (2005) [18] claim that "in order for noticing to happen, input might have to be made salient through input enhancement, which will raise the learners" consciousness about the target features."

CONCLUSION
The present study compared 2 production tests as data elicitation methods and as testing tools of EFL 2 Percentages are indicated in brackets. students" production of the speech act of request. The results provided somewhat different samples of request strategies in the two tests. Compared to the pre-test results which showed a high degree of directness in students" responses, the post-test responses contained a greater variety of strategies. In addition, students" responses to the questionnaire revealed that they showed great awareness of the appropriate target strategies employed when requesting, and that they approached the explicit instructional treatment with high degree of sensitivity.
However, results of the present study should be interpreted within the context of its methodology limitations. First, this was a small scale study involving only a small sample of learners. Second, while the present study showed that explicit instruction was highly beneficial, the instructional period which spanned over one month is rather short. Moreover, the results of the study concerning the effect of the explicit instructional conditions would have worked better if the treatment design had included a control group which may have yielded more reliable results.
Accordingly, future research needs to investigate the effects of the pragmatic instruction in longer duration of the treatment while including a control group. Unless we do so, the findings of the present investigation remain exploratory and questionable.