Please cite as...
© 1997 Second Language Teaching & Curriculum Center

'Can Pragmatic Competence Be Taught?' The simple answer to the question as formulated is "no". Competence, whether linguistic or pragmatic, is not teachable. Competence is a type of knowledge that learners possess, develop, acquire, use or lose. The challenge for foreign or second language teaching is whether we can arrange learning opportunities in such a way that they benefit the development of pragmatic competence in L2. This, then, is the issue I will address in this paper.

The pragmatic component in models of communicative competence

There are many definitions of pragmatics around. One I find particularly useful has been proposed by David Crystal. According to him, "Pragmatics is the study of language from the point of view of users, especially of the choices they make, the constraints they encounter in using language in social interaction and the effects their use of language has on other participants in the act of communication" (Crystal 1985, p. 240). In other words, pragmatics is the study of communicative action in its sociocultural context. Communicative action includes not only speech acts - such as requesting, greeting, and so on - but also participation in conversation, engaging in different types of discourse, and sustaining interaction in complex speech events. Following Leech (1983), I will focus on pragmatics as interpersonal rhetoric - the way speakers and writers accomplish goals as social actors who do not just need to get things done but attend to their interpersonal relationships with other participants at the same time.

Leech (1983) and his colleague Jenny Thomas (1983) proposed to subdivide pragmatics into a pragmalinguistic and sociopragmatic component. Pragmalinguistics refers to the resources for conveying communicative acts and relational or interpersonal meanings. Such resources include pragmatic strategies like directness and indirectness, routines, and a large range of linguistic forms which can intensify or soften communicative acts. For one example, compare these two versions of apology - the terse 'I'm sorry' and the Wildean 'I'm absolutely devastated. Can you possibly forgive me?' In both versions, the speaker apologizes, but she indexes a very different attitude and social relationship in each of the apologies (e.g., Fraser, 1980; House & Kasper, 1981; Brown & Levinson, 1987; Blum-Kulka, House, & Kasper, 1989).

Sociopragmatics was described by Leech (1983, p. 10) as 'the sociological interface of pragmatics', referring to the social perceptions underlying participants' interpretation and performance of communicative action. Speech communities differ in their assessment of speaker's and hearer's social distance and social power, their rights and obligations, and the degree of imposition involved in particular communicative acts (Takahashi & Beebe, 1993; Blum-Kulka & House, 1989; Olshtain, 1989). The values of context factors are negotiable; they can change through the dynamics of conversational interaction, as captured in Fraser's (1990) notion of the 'conversational contract' and in Myers-Scotton's Markedness Model (1993).

Pragmatic ability in a second or foreign language is part of a nonnative speakers (NNS) communicative competence and therefore has to be located in a model of communicative ability (Savignon, (1991, for overview). In Bachman's model (1990, p. 87ff), 'language competence' is subdivided into two components, 'organizational competence' and 'pragmatic competence'. Organizational competence comprises knowledge of linguistic units and the rules of joining them together at the levels of sentence ('grammatical competence') and discourse ('textual competence'). Pragmatic competence subdivides into 'illocutionary competence' and 'sociolinguistic competence'. 'Illocutionary competence' can be glossed as 'knowledge of communicative action and how to carry it out'. The term 'communicative action' is often more accurate than the more familiar term 'speech act' because communicative action is neutral between the spoken and written mode, and the term acknowledges the fact that communicative action can also be implemented by silence or non-verbally. 'Sociolinguistic competence' comprises the ability to use language appropriately according to context. It thus includes the ability to select communicative acts and appropriate strategies to implement them depending on the current status of the 'conversational contract' (Fraser, 1990).

Need L2 pragmatics be taught?

As Bachman's model makes clear, pragmatic competence is not extra or ornamental, like the icing on the cake. It is not subordinated to knowledge of grammar and text organization but co-ordinated to formal linguistic and textual knowledge and interacts with 'organizational competence' in complex ways. In order to communicate successfully in a target language, pragmatic competence in L2 must be reasonably well developed. But adopting pragmatic competence as one of the goals for L2 learning does not necessarily imply that pragmatic ability requires any special attention in language teaching. Before turning to the central question of my talk, i.e., whether L2 pragmatics can be taught, I will therefore address the logically prior question of whether L2 pragmatics needs to be taught. Because perhaps pragmatic knowledge simply develops alongside lexical and grammatical knowledge, without requiring any pedagogic intervention.

Indeed, adult NNS do get a considerable amount of L2 pragmatic knowledge for free. This is because some pragmatic knowledge is universal, and other aspects may be successfully transferred from the learners' L1. To start with the pragmatic universals, learners know that conversations follow particular organizational principles - participants have to take turns at talk, and conversations and other speech events have specific internal structures. Learners know that pragmatic intent can be indirectly conveyed, and they can use context information and various knowledge sources to understand indirectly conveyed meaning. They know that recurrent speech situations are managed by means of conversational routines (Coulmas, 1981; Nattinger & DeCarrico, 1992) rather than by newly created utterances. They know that strategies of communicative actions vary according to context (Blum-Kulka, 1991); specifically, along such factors as social power, social and psychological distance, and the degree of imposition involved in a communicative act, as established in politeness theory (Brown & Levinson, 1987; Brown & Gilman, 1989). Learners have demonstrated knowledge of the directive and expressive speech acts that have been most frequently studied in cross-cultural and interlanguage pragmatics, such as requests and apologies, and they have been shown to understand and use the major realization strategies for such speech acts. For instance, in requesting, users of any language studied thus far distinguish different levels of directness; direct, as in 'feed the cat', conventionally indirect, as in 'can/could/would you feed the cat?', and indirect, as in 'the cat's complaining.' Furthermore, language users know that requests can be softened or intensified in various ways, as in 'I was wondering if you would terribly mind feeding the cat', and that requests can be externally modified through various supportive moves, for instance justifications, as in 'I have to go to a conference', or imposition minimizers, as in 'She only needs food once a day'. Studies document that these strategies of requesting are available to ESL or EFL learners who are NS of such diverse languages as Chinese (Johnston, Kasper, & Ross, 1994), Danish (Færch & Kasper, 1989), German (House & Kasper, 1987), Hebrew (Blum-Kulka & Olshtain, 1986), Japanese (Takahashi & DuFon, 1989), Malay (Piirainen-Marsh, 1995), and Spanish (Rintell & Mitchell, 1989). In their early learning stages, learners may not be able to use such strategies because they have not yet acquired the necessary linguistic means, but when their linguistic knowledge permits it, learners will use the main strategies for requesting without instruction.

Learners may also get very specific pragmalinguistic knowledge for free if there is a corresponding form-function mapping between L1 and L2, and the forms can be used in corresponding L2 contexts with corresponding effects. For instance, the English modal past as in the modal verbs could or would has formal, functional and distributional equivalents in other Germanic languages such as Danish and German - the Danish modal past kunne/ville and the German subjunctive könntest and würdest. And sure enough, Danish and German learners of English transfer ability questions from L1 Danish (kunne/ville du låne mig dine noter) and L1 German (könntest/ würdest Du mir Deine Aufzeichnungen leihen) to L2 English (could/would you lend me your notes) (House & Kasper, 1987; Færch & Kasper, 1989), and they do this without the benefit of instruction.

Positive transfer can also facilitate learners' task in acquiring sociopragmatic knowledge. When distributions of participants' rights and obligations, their relative social power and the demands on their resources are equivalent in their original and target community, learners may only need to make small adjustments in their social categorizations (Mir, 1995).

Unfortunately, learners do not always make use of their free ride. It is well known from educational psychology that students do not always transfer available knowledge and strategies to new tasks. This is also true for some aspects of learners' universal or L1-based pragmatic knowledge. L2 recipients often tend towards literal interpretation, taking utterances at face value rather than inferring what is meant from what is said and underusing context information. Learners frequently underuse politeness marking in L2 even though they regularly mark their utterances for politeness in L1 (Kasper, 1981). Although highly context-sensitive in selecting pragmatic strategies in their own language, learners may underdifferentiate such context variables as social distance and social power in L2 (Fukushima, 1990; Tanaka, 1988).

So, the good news is that there is a lot of pragmatic information that adult learners possess, and the bad news is that they don't always use what they know. There is thus a clear role for pedagogic intervention here, not with the purpose of providing learners with new information but to make them aware of what they know already and encourage them to use their universal or transferable L1 pragmatic knowledge in L2 contexts.

The most compelling evidence that instruction in pragmatics is necessary comes from learners whose L2 proficiency is advanced and whose unsuccessful pragmatic performance is not likely to be the result of cultural resistance or disidentification strategies (Kasper, 1995, for discussion). In a study of a large sample of advanced ESL learners, Bouton (1988) examined how well these students understood different types of indirect responses, or implicature, as in the following dialog:
Sue: How was your dinner last night?
Anne: Well, the food was nicely presented.
Bouton found that in 27% of the cases, implicatures were understood differently by native speakers (NS) and NNS. A re-test of 30 students after 4 1/2 years demonstrated that their comprehension now showed a success rate of over 90%. But some implicature types resisted improvement through exposure alone. These included the Pope question (as in Is the Pope Catholic?) and indirect criticism as in the Sue & Anne dialogue. Students' comprehension of implicature may thus profit from instruction, and as we will see shortly, this has indeed proved to be the case.

Turning to production, candidates for pedagogic intervention can be sorted in four groups: (1) choice of communicative acts, (2) the strategies by which an act is realized, (3) its content, and (4) its linguistic form. Drawing on her and Beverly Hartford's data from academic advising sessions (Bardovi-Harlig & Hartford 1990, 1993), Bardovi-Harlig (1996) noted that NNS students tended to leave suggestions about their coursework to their advisor and then react to them. Consequently, the NNS performed more rejections of advisor suggestions than the NS students, who were more initiative in making suggestions and thereby avoided rejections. Both NS and NNS regularly offered explanations when they rejected their advisor's course suggestion, but the NS would also suggest alternatives ('how about I take x course instead'), something the NNS never did. For their rejections, the NNS sometimes used inappropriate content, such as claiming the course suggested by their advisor was either too easy or too difficult, or even evaluating their advisor's course as 'uninteresting'. Finally, even at the end of the observation period, the NNS had not learnt how to mitigate their suggestions and rejections appropriately. By using mitigating forms such as 'I was thinking' or 'I have an idea... I dont' know how it would work out, but...', the NS would cast their suggestions in tentative terms. By contrast, the NNS tended to formulate their suggestions much more assertively, as in 'I will take language testing' or 'I've just decided on taking the language structure' (all examples from Bardovi-Harlig, 1996, 22f.).

Two things need to be emphasized in assessing the implications of Bouton's and Bardovi-Harlig and Hartford's studies. First, the participating advanced students were ESL learners, yet the target environment either did not provide students with the input they needed, or they did not notice it. Secondly, the recorded differences in NS and NNS pragmatic comprehension and production may lead to serious miscommunication and compromise the NNS's goals. Bardovi-Harlig and Hartford (1990) found that when students' contributions were pragmatically inappropriate, they were less successful in obtaining their advisor's consent for taking the courses they preferred.

A further aspect of students' pragmatic competence is their awareness of what is and is not appropriate in given contexts. Bardovi-Harlig and Dörnyei (1997) reported that Hungarian and Italian EFL learners recognized grammatically incorrect but pragmatically appropriate utterances more readily than pragmatically inappropriate but grammatically correct utterances, and this was true for learners of all proficiency levels. This finding strongly suggests that without a pragmatic focus, foreign language teaching raises students' metalinguistic awareness, but it does not contribute much to develop their metapragmatic consciousness in L2.

Can L2 pragmatics be taught?

All of the 10 studies report on classroom-based research on pragmatics. I excluded studies conducted in a lab type situation because I wanted to make sure that the chosen approaches are ecologically valid in actual L2 classrooms.

As you can see from the second column to the left, the teaching goals in these studies extend over a large range of pragmatic features and abilities. Some studies examine the discourse markers and strategies by which conversationalists get in and out of conversations, introduce, sustain, and change topics, organize turn-taking and keep the conversation going by listener activities such as backchanneling. Many of these conversational activities are implemented by pragmatic routines which regularly occur in spoken discourse, yet foreign language learners may have little exposure to them. A number of discourse markers and strategies are illustrated in the following conversational sequence.

A telephone conversation (Sacks, 1995, vol. II, p. 201f; transcript slightly modified)
A: Hello.
B: Vera?
A: Ye:s.
B: Well you know, I had a little difficulty getting you. (1.0) First I got the wrong number, and then I got Operator, [A: Well.] And uhm (1.0) I wonder why.
A: Well, I wonder too. It uh just rung now about uh three ti//mes.
B: Yeah, well Operator got it for me.
A: She did.
B: Uh huh. So //uh
A: Well.
B: When I- after I got her twice, why she [A: telephoned] tried it for me. Isn't that funny?
A: Well it certainly is.
B: Must be some little cross of lines someplace hh
A: Guess so.
B: Uh huh,uh, am I taking you away from yer dinner?
A: No::. No, I haven't even started tuh get it yet.
B: Oh, you have//n't.
A: hhheh heh
B: Well I- I never am certain, I didn't know whether I'd be m too early or too late // or ri-
A: No::. No, well I guess uh with us uhm there isn't any - [B: Yeah.] p'ticular time.

Another group of studies explores whether students benefit from instruction in specific speech acts. So far, speech acts examined are compliments, apologies, complaints, and refusals. There is a research literature on all of these speech acts, documenting how they are performed by native speakers of English in different social contexts. Based on this literature, students were taught the strategies and linguistic forms by which the speech acts are realized and how these strategies are used in different contexts. As one example, consider the realization strategies (or 'speech act set') for apologies (adapted from Blum-Kulka, House, & Kasper, 1989):

• Apologetic formula: I'm sorry, I apologize, I'm afraid
• Assuming Responsibility: I haven't read your paper yet.
• Account: I had to prepare my TESOL plenary.
• Offer of Repair: But I'll get it done by Wednesday.
• Appeaser: Believe me, you're not the only one.
• Promise of forbearance: I'll do better after TESOL.
• Intensifier: I'm terribly sorry, I really tried to squeeze it in.

Whereas most of these pragmatic features were taught to intermediate or advanced learners, participants in Wildner-Bassett (1994) and Tateyama et al. (1997) were beginning learners. These two studies thus address the important question of whether pragmatics is teachable to beginners or whether there needs to be some threshold of linguistic L2 competence first.

Wildner-Bassett's (1994) and Tateyama et al.'s studies are also the only ones in which the target language is not English - in Wildner-Bassett's study, the L2 is German, in Tateyama et al., it is Japanese. Note that in some studies, the target language is a foreign language whereas in others, it is a second language. This has consequences for the learning outcomes, as I will show a bit later.

The studies differed in their research goals. Olshtain and Cohen (1990), Wildner-Bassett (1994) and Morrow (1996) explored whether the features under investigation were teachable at all. These studies did not employ control groups but compared students' test performance before and after instruction to that of NS of the target language, referred to as 'L2 baseline' in the 'design' column in Table 1. Billmyer (1990) and Bouton (1994) examined whether students who received instruction in complimenting and implicature did better than controls who did not. Yet another group explored the effectiveness of specific teaching approaches. In these studies, two or more student groups received different types of instruction. House and Kasper (1981), House (1996), and Tateyama et al. (1997) compared explicit with implicit approaches. Explicit teaching involved description, explanation, and discussion of the pragmatic feature in addition to input and practice, whereas implicit teaching included input and practice without the metapragmatic component. Wildner-Bassett (1984, 1986) compared an eclectic approach with a modified version of suggestopedia, and Kubota (1995) compared an inductive approach, where students had to figure out in groups how implicatures in English work, to a teacher-directed deductive approach and zero instruction in implicature. Information about the designs and assessment procedures and instruments is provided in the two rightmost columns in Table 1, but I'm not going to comment on those. Instead, let's proceed to the findings of the studies.

First of all, the studies that examined whether the selected pragmatic features were teachable found this indeed to be the case, and comparisons of instructed students with uninstructed controls reported an advantage for the instructed learners. Secondly, the studies comparing the relative effect of explicit and implicit instruction found that students' pragmatic abilities improved regardless of the adopted approach, but the explicitly taught students did better than the implicit groups. Thirdly, with respect to other teaching approaches, Wildner-Bassett (1984, 1986) found that both the eclectively taught students and the suggestopedic group improved their use of conversational routines considerably, however the eclectic group outperformed the suggestopedic group. Kubota (1995) reported an advantage for students receiving either deductive or inductive instruction over the uninstructed group, with a superior effect for the inductive approach, this initial difference had evaporated by the time a delayed post-test was administered.

Wildner-Bassett (1994) and Tateyama et al. (1997) demonstrated that pragmatic routines are teachable to beginning foreign language learners. This finding is important in terms of curriculum and syllabus design because it dispels the myth that pragmatics can only be taught after students have developed a solid foundation in L2 grammar and vocabulary. As we know from uninstructed first and second language acquisition research, most language development is function-driven - i.e., the need to understand and express messages propels the learning of linguistic form. Just as in uninstructed acquisition, students can start out by learning pragmatic routines which they cannot yet analyze but which help them cope with recurrent, standardized communicative events right from the beginning.

There is little evidence for aspects of L2 pragmatics that resist development through teaching, but the few documented cases are instructive. One such study is Kubota's replication of Bouton's (1994) research on the teaching of implicature. Kubota's Japanese EFL learners were able to understand the exact implicatures that were repeated from the training materials but were unable to generalize inferencing strategies to new instances of implicature. However, these students' English proficiency was much less advanced than that of the learners in Bouton's studies, and with more time, occasion for practice, and increased L2 input, the students' success rate might have improved.

The other study that suggests limitations to teachability in L2 pragmatics is House's (1996) investigation on improving the pragmatic fluency of advanced German EFL students. All but one feature of pragmatic fluency gained from consciousness raising and conversational practice; the resistent aspect was to provide appropriate rejoinders, or second pair parts, to an interlocutor's preceding contribution, as in this exchange:

NS: Oh I tell you what we go shopping together and buy all the things [we need]
NNS: [Of course] of course
NS: Okay then and you try and call Anja and ask her if she knows somebody who owns a grill
NNS: Yes of course (House, 1996, p. 242)

More appropriate acceptances of the NS' suggestions would have been 'ok/good idea/let's do it that way then' or the like. Why would inappropriate rejoinders persist in these advanced learners' discourse despite instruction? A plausible explanation is Bialystok's (e.g., 1993) notion of control of processing: fluent and appropriate conversational responses require high degrees of processing control in utterance comprehension and production, and such complex skills may be very hard to develop through the few occasions for practice that foreign language classroom learning provides.

But despite those few limitations, the research supports the view that pragmatic ability can indeed be systematically developed through planful classroom activities. In order to address the next question -

How can language instruction help develop pragmatic competence?

- we need to consider for a moment what opportunities for pragmatic learning are offered by traditional forms of language teaching.

L2 classrooms as impoverished learning environments
It is a well-documented fact that in teacher-fronted teaching, the person doing most of the talking is the teacher (e.g., Chaudron, 1988, for various analyses of teacher talk). This is to the detriment of students' speaking opportunities, but it could be argued that through the sheer quantity of teacher talk, students are provided with the input they need for pragmatic development. However, studies show that compared to conversation outside instructional settings, teacher-fronted classroom discourse displays

• a more narrow range of speech acts (Long, Adams, McLean, & Castaños, 1976)
• a lack of politeness marking (Lörscher & Schulze, 1988)
• shorter and less complex openings and closings (Lörscher, 1986; Kasper, 1989)
• monopolization of discourse organization and management by the teacher (Lörscher, 1986; Ellis, 1990), and consequently,
• a limited range of discourse markers (Kasper, 1989).

The reason for such differences is not that classroom discourse is 'artificial'. Classroom discourse is just as authentic as any other kind of discourse. Rather, classroom interaction is an institutional activity in which participants' roles are asymmetrically distributed (Nunan, 1989), and the social relationships in this unequal power encounter are reflected and re-affirmed at the level of discourse. Teacher's and students' rights and obligations, and the activities associated with them, are epitomized in the basic interactional pattern of traditional teacher-fronted teaching - the (in)famous pedagogical exchange of elicitation (by the teacher) - response (by a student) - feedback (by the teacher) (cf. discussion in Chaudron, 1988, p. 37). The classic scenario is consistent with a knowledge-transmission model of teaching, according to which the teacher imparts new information to students, helps them process such information and controls whether the new information has become part of students' knowledge. Such functions can be implemented through a very limited range of communicative acts.

If we map the communicative actions in classic language classroom discourse against the pragmatic competence that nonnative speakers need to communicate in the world outside, it becomes immediately obvious that the language classroom in its classical format does not offer students what they need - not in terms of teacher's input, nor in terms of students' productive language use. In a comparison of teacher-fronted teaching and small group work, Long et al. (1976) demonstrated over 20 years ago that student participation increases dramatically in student-centered activities. Importantly, student-centered activities do more than just extend students' speaking time: they also give them opportunities to practice conversational management, perform a larger range of communicative acts, and interact with other participants in completing a task.

But despite its unique structure, even teacher-fronted classroom discourse offers some opportunities for pragmatic learning. One important learning resource is classroom management, because in this activity language does not function as an object for analysis and practice but as a means for communication. If classroom management is performed in the students' L1, they miss a valuable opportunity for experiencing the L2 as a genuine means of communication. In a recent call for a role of students' native language in ESL teaching, Auerbach (1993) proposed that classroom management is one of the activities that could be carried out in students' L1 rather than the L2. Auerbach argues that using minority students' native language for classroom management is one way of validating the students' ethnolinguistic identity in an ESL classroom. In my view, Auerbach's call against English Only classrooms in ESL settings for immigrant minorities is valid and necessary, but I want to caution against extending it to EFL situations or any other foreign language classrooms, for that matter. For students of English in Continental Europe or Asia, or students of Japanese and French in the US, the FL classroom may be the only regular opportunity for using the FL for communication. These opportunities should not be curtailed, and certainly not when it comes to routinized activities such as classroom management discourse. In a recent study of his learning of Japanese as a Foreign Language, Cohen (1997) reports:

"Classroom talk was focused primarily on completing a series of planned transactions, such as making introductions, buying stamps or postcards at a post office, buying clothes in a department store, telling the doctor about our illness, and the like. There was little non-transactional social conversation in class, other than asides in English. In addition, spoken language tended to be focused on structures that we were to learn (...). Toward the end of the second month, we would start the class off with teacher-directed questions and answers, usually inquiring about what we had done the previous day or weekend, or what we intended to do - usually with the purpose of practicing some structure or other."

Because little genuinly communicative interchange was conducted in Japanese, students had not much exposure to authentic input in this classroom.

From the studies reviewed earlier and from other theory and research of SL learning, we can distill a number of activities that are useful for pragmatic development. Such activities can be classified into two main types: activities aiming at raising students' pragmatic awareness, and activities offering opportunities for communicative practice.

Awareness-raising

Through awareness-raising activities, students acquire sociopragmatic and pragmalinguistic information - for instance, what function complimenting has in mainstream American culture, what appropriate topics for complimenting are, and by what linguistic formulae compliments are given and received. Students can observe particular pragmatic features in various sources of oral or written 'data', ranging from native speaker 'classroom guests' (Bardovi-Harlig, et al., 1991) to videos of authentic interaction, feature films (Rose, 1997), and other fictional and non-fictional written and audiovisual sources.

Observation tasks
Especially in a second language context, students can be given a variety of observation assignments outside the classroom. Such observation tasks can focus on sociopragmatic or pragmalinguistic features.

A sociopragmatic task could be to observe under what conditions native speakers of American English express gratitude - when, for what kinds of goods or services, and to whom (cf. Eisenstein & Bodman, 1993). Depending on the student population and available time, such observations may be open or structured. Open observations leave it to the students to detect what the important context factors may be. For structured observations, students are provided with an observation sheet which specifies the categories to look out for - for instance, speaker's and hearer's status and familiarity, the cost of the good or service to the giver, and the degree to which the giver is obliged to provide the good or service. A useful model for such an observation sheet is the one proposed by Rose (1994) for requests.

A pragmalinguistic task focuses on the strategies and linguistic means by which thanking is accomplished - what formulae are used, and what additional means of expressing appreciation are employed, such as expressing pleasure about the giver's thoughtfulness or the received gift, asking questions about it, and so forth. Finally, by examining in which contexts the various ways of expressing gratitude are used, sociopragmatic and pragmalinguistic aspects are combined. By focusing students' attention on relevant features of the input, such observation tasks help students make connections between linguistic forms, pragmatic functions, their occurrence in different social contexts, and their cultural meanings. Students are thus guided to notice the information they need in order to develop their pragmatic competence in L2 (Schmidt, 1993). The observations made outside the classroom will be reported back to class, compared with those of other students, and perhaps commented and explained by the teacher. These discussion can take on any kind of small group of whole class format.

Whether gathered through out-of-class observation or brought into the classroom through audiovisual media, authentic native speaker input is indispensible for pragmatic learning. This is not because students should imitate native speakers' action patterns but in order to build their own pragmatic knowledge on the right kind of input. Comparisons of textbook dialogues and authentic discourse show that there is often a mismatch between the two. For instance, Bardovi-Harlig, et al. (1991) examined conversational closings in 20 textbooks for American English and found that few of them represented closing phases accurately. Myers-Scotton and Bernstein (1988) discovered similar discrepancies between the representation of many other conversational features in authentic discourse and textbook dialogues. The reason for such inaccurate textbook representations is that native speakers are only partially aware of their pragmatic competence (the same is true of their language competence generally). As Wolfson (1989) noted, most of native speakers' pragmatic knowledge is tacit, or implicit knowledge: it underlies their communicative action, but they cannot describe it. Even the most proficient conversationalist has little conscious awareness about turn-taking procedures and politeness marking. Miscommunication or pragmatic failure is often vaguely diagnosed as 'impolite' behavior on the part of the other person, whereas the specific source of the irritation remains unclear. Because native speaker intuition is a notoriously unreliable source of information about the communicative practices of their own community, it is vital that teaching materials on L2 pragmatics are research-based (Myers-Scotton & Bernstein, 1988; Wolfson, 1989; Olshtain & Cohen, 1991; Bardovi-Harlig, et al., 1991).

Authentic L2 input is essential for pragmatic learning, but it does not secure successful pragmatic development. When students' observe L2 communicative practices, their minds don't simply record what they hear and see like a videocamera does. Students' experiences are interpretive rather than just registering. Cognitive psychology (e.g., Sanford & Garrod, 1981) as well as radical constructivism (e.g., von Glaserfeld, 1995) emphasize the importance of prior knowledge for comprehension and learning. In our attempt to understand the practices of an unfamiliar community, we tend to view such practices through the lenses of our own customs. We tend to classify experiences into 'familiar' and thus not requiring further reflection or analysis, and 'unfamiliar', i.e., peculiar, enigmatic, inviting explanation, and attracting evaluation. Müller (1981) referred to this interpretive strategy as cultural isomorphism. As a strategy for the acquisition of everyday knowledge, cultural isomorphism is a combination of assimilation and spot-the-difference. L2 practices are subjected to the same social evaluations as the apparently equivalent L1 practices. The resulting perspective is that of a tourist who sorts experiences in the visited country into 'just like home' and 'strange'. As Elbeshausen and Wagner (1985) comment, "Tourism is not educational but it dramatically increases our repertoire of anecdotes" (p. 49), and this is because through the assimilative and contrastive strategy of isomorphism, stereotypical evaluations of L2 practices emerge. Language teaching therefore has the important task to help students situate L2 communicative practices in their sociocultural context and appreciate their meanings and functions within the L2 community. The research literature on cross-cultural pragmatics documents the rich intracultural variation of communicative action patterns and thus offers compelling counter-evidence against unhelpful and often mutual stereotypes. For example, a stereotype held by some Japanese learners of English is that Americans have a very direct style of communication (Tanaka, 1988; Robinson, 1992); however, research on requests (Blum-Kulka & House, 1989; Blum-Kulka, 1991) and refusals (Beebe, Takahashi, & Uliss-Weltz,, 1990; Beebe & Cummings, 1996) provides evidence to the contrary.

Practicing L2 pragmatic abilities

Turning to students options for practicing their L2 pragmatic abilities, such practice requires student-centered interaction. In their books on tasks for language learning, Nunan (1989) and Crookes and Gass (1993a, b) explain the rationale underlying a task-based approach from the perspectives of second language acquisition and pedagogy. Most small group interaction requires that students take alternating discourse roles as speaker and hearer, yet different types of task may engage students in different speech events and communicative actions. It is therefore important to identify very specifically which pragmatic abilities are called upon by different tasks. A useful distinction can be made between referential and interpersonal communication tasks. In referential communication tasks (Yule, in press), students have to refer to concepts for which they lack necessary L2 words. Such tasks expand students' vocabulary and develop their strategic competence. Interpersonal communication tasks are more concerned with participants' social relationships and include such communicative acts as opening and closing conversations, expressing emotive responses as in thanking and apologizing, or influencing the other person's course of action as in requesting, suggesting, inviting, and offering. Activities such as roleplay, simulation, and drama engage students in different social roles and speech events. Such activities provide opportunities to practice the wide range of pragmatic and sociolinguistic abilities (Crookall & Saunders, 1989; Crookall & Oxford, 1990; Olshtain & Cohen, 1991) that students need in interpersonal encounters outside the classroom.

Reconsidering pragmatic ability as a teaching goal

The purpose of the proposed learning activities is to help students become more effective and successful communicators in L2. But what exactly does 'effective' and 'successful' mean? In conclusion of this paper, I will briefly re-examine the goals that instruction in pragmatics should aim for.

First, it may be useful to remind ourselves that NS are no ideal communicators. As Coupland, Wiemann, and Giles, (1991, p. 3) comment, "language use and communication are (...) pervasively and even intrinsically flawed, partial, and problematic". And yet, by and large NS communication succeeds more than it fails - not because it is perfect but because it is good enough for the purpose at hand. It would be unreasonable and unrealistic to place higher demands on L2 learners' communicative abilities than on those of NS. Therefore, there is a continued need for studies examining how NS and NNS communicate effectively in different contexts.

Secondly, there often appears to be an implicit understanding that effective and successful NNSs have the same or very similar pragmatic ability as NS. On this view, pragmatic competence as a learning objective should be based on a NS model. However, as Siegal (1996) points out, "Second language learners do not merely model native speakers with a desire to emulate, but rather actively create both a new interlanguage and an accompanying identity in the learning process" (1996, p. 362ff) Second language learners' desire for convergence with NS pragmatics or divergence from NS practices is shaped by learners' views of themselves, their social position in the target community and in different contexts within the wider L2 environment, and by their experience with NS in various encounters.

Thirdly, members of the target community may perceive NNS's total convergence to L2 pragmatics as intrusive and inconsistent with the NNS's role as outsider to the L2 community, whereas they may appreciate some measure of divergence as a disclaimer to membership. Giles, Coupland, and Coupland (1991) documented that in many ethnolinguistic contact situations, successful communication is a matter of optimal rather than total convergence. Optimal convergence is a dynamic, negotiable construct that defies hard-and-fast definition. It refers to pragmatic and sociolinguistic choices which are consistent with participants' subjectivities and social claims, and recognizes that such claims may be in conflict between participants.

Fourthly, as Peirce (1995) noted, language classrooms provide an ideal arena for exploring the relationship between learners' subjectivity and L2 use. Classrooms afford second language learners the opportunity to reflect on their communicative encounters and to experiment with different pragmatic options. For foreign language learners, the classroom may be the only available environment where they can try out what using the L2 feels like, and how more or less comfortable they are with different aspects of L2 pragmatics. The sheltered environment of the L2 classroom will thus prepare and support learners to communicate effectively in L2. But more than that, by encouraging students to explore and reflect their experiences, observations, and interpretations of L2 communicative practices and their own stances towards them, L2 teaching will expand its role from that of language instruction to that of language education.

Go to References.