THE PROBLEM OF ETHNOCENTRIC ~IAS IN SPEECH ACT STUDIES: IMPLICATIONS FOR LANGUAGE TEACHING

Most language teaching specialists today hold the view that the aim of second language teaching should be to facilitate learners' acquisition of so-called "communicative competence". Leaving aside the many questions concerning the meaning and use of this term that are being hotly debated in the literature, I will use the term "communicative competence" to refer to the system(s) of knowledge that underlie the ability to use a language both accurately, that is, in a grammatically correct way, and appropriately in different social and situational contexts.1 It is with the latter aspect of communicative competence in particular, viz. the knowledge underlying the ability to use a language appropriately in context, that this paper will be concerned. Let us call this aspect of communicative competence "pragmatic competence".2


Introduction
Most language teaching specialists today hold the view that the aim of second language teaching should be to facilitate learners' acquisition of so-called "communicative competence".
Leaving aside the many questions concerning the meaning and use of this term that are being hotly debated in the literature, I will use the term "communicative competence" to refer to the system(s) of knowledge that underlie the ability to use a language both accurately, that is, in a grammatically correct way, and appropriately in different social and situational contexts.1 It is with the latter aspect of communicative competence in particular, viz. the knowledge underlying the ability to use a language appropriately in context, that this paper will be concerned.
Let us call this aspect of communicative competence "pragmatic competence".2 Teachers, teacher trainers, curriculum designers and materials writers faced with the task of producing not only grammatically competent, but also pragmatically competent second language speakers, need answers to ~uestions such as the following: (1) What does it mean to be pragmatically competent in a language?
(2) What aspect(s) of pragmatic competence can be assumed to be universal and can therefore be expected to carry over from the learner's mother tongue?
(3) How can the development of pragmatic competence in a second language be facilitated?
Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 Providing answers to questions such as (1) and (2) in particular is a concern of linguistics, with linguistics being taken in a broad sense to include disciplines such as pragmatics, sociolinguistics and psycholinguistics.The third question, being a question about teaching practice, is perhaps not first and foremost a linguistic question.However, given that the answer to this question is, at least to a certain extent, dependent on the answers given to the first two questions, it is also, partly, a linguistic question.
The aim of this paper is to highlight the contribution that a field of linguistic research known as cross-cultural pragmatics has made and could potentially make to answering questions such as (1 )-( 3) above.The focus will be on question (2), the question of what aspects of pragmatic competence, if any, can be taken to be universal or non-Ianguage-specific.Some rather strong claims have been made in the literature regarding the putative universality of particular aspects of pragmatic competence.
A number of these claims will be presented in section 3 below.However, the fact that these initial universality claims were based almost exclusively on evidence from English and languages closely related to English has given rise to the criticism that they reflect an anglocentric bias.
As will be shown in section 3, this criticism is supported by the findings of a growing number of studies that compare the ways in which particular speech acts are performed in different languages and cultures.The results of these studies and the insights they offer into the way in which cultural differences are encoded in speech act performance, has important implications for first and second language teaching in linguistically and culturally diverse societies.
A brief look at some of these implications in section 4 should give an indication of the direction in which answers to question (3) must eventually be sought.Section 2 will deal, very bri~fly, with question  3 A recent answer to this question is the one given in (Bachman 1990).
Sociolinguistic competence, according to Bachman (1990:90), is "knowledge of the sociolinguistic conventions for performing language functions appropriately in a given context".Returning to our example, then, we may say that the knowledge This question can now be made more specific, given Bachman's view of pragmatic competence as comprising two kinds of competence, viz.illocutionary competence and sociolinguistic competence.
Firstly, with regard to illocutionary competence, the more specific questions in (4) arise: (4 ) i.
Do all languages allow the same speech acts, or at least the same types of speech acts, to be performed?
For example, "do all languages have representative speech acts, such as asserting, claiming, saying, reporting, etc.; directives, such as ordering, requesting, suggesting, etc.; commissives, such as promising and threatening; and a number of other types that have been proposed in the literature?Secondly, as far as sociolinguistic competence is concerned, the general question (2) above gives rise to more specific questions such as those in (5): (5 ) i.

Is the relationship between contextual factors
and the choice of specific speech act strategies the same across languages and cultures?For example, do speakers across languages and cultures choose more polite strategies when addressing requests to older people, people of higher status, strangers, etc.? ii.
Is the relationship between social norms and the choice of particular speech act strategies the same across languages and cultures?For example, are speakers across languages and cultures motivated by a desire to be polite in choosing indirect rather than direct strategies to realize directive speech acts such as requests?
Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 In the next section we will con~ider some of the claims and counterclaims that have been made in the literature in response to questions such as those in (4) and ( 5) concerning the possible universality of aspects of pragmatic competence. 3 The question of universality

Illocutionary competence
As was pointed out in section 2, a first set of questions that bear on the issue of universality in the domain of pragmatic competence are questions about aspects of illocutionary competence, viz.knowledge of what speech acts can be performed and of the pragmatic and linguistic means available for performing them.
The first question, formulated as (4i) above, is whether all languages allow the same speech acts, or at least the same types of speech acts, to be performed.According to Schmidt and Richards (1980:138), most researchers assu~e that the same basic types of speech acts (representative, directive, commissive, expressive, etc.) occur in all languages and cultures.
In an often quoted paper, Fraser, Rintell and Walters (1980:78-79) go even further, claiming that every language makes available to the user the same basic individual speech acts, such as requesting, apologizing, declaring, and promising.They do make provision for the existence, outside the "basic set of speech acts", of acts such as baptizing, excommunicating, doubling at bridge, etc.
that they take to be culture-specific and often highly ritualized.
In a recently published monograph, Anna Wierzbicka takes issue with Fraser et al.'s claim. She (1991 :1S0ff) ~oints out that "English words such as question.command or blessing identify concepts which are languagespecific.
They embody an English folk taxonomy, which, like all folk taxonomies, is culturespecific".She (1991:1S2ff)  These components include "an assumption that the speaker has authority over the addressee, the intention of protecting the addressee from evil, and good feelings towards the addressee".The concepts of authority, responsibility and care do not form part of the concept encoded by the English word warning, according to Wierzbicka (1991:153).
A second example comes from an Australian Aboriginal language.
In the language of the Yolngu people, according to Wierzbicka (1991 :158) Examples such as these, according to Wierzbicka (1991:151), provide clear evidence that speech acts are not necessarily language-and culture-independent natural conceptual kinds, to which different languages merely attach different labels.
As to the question whether the assumption holds that all languages at least have speech acts belonging to all the proposed basic types, viz.representatives, directives, commissives, etc., the answer still has to be the one given by Schmidt and (1980:138):" in fact there has been no ethnographic research carried out to confirm or disprove the assumption".

Richards
Let us turn to the second question, (4ii) above, which is whether the pragmatic strategies available for realizing a seem particularly odd and amusing -from a Polish point of view " It would seem, then, that even if it could be maintained that, in very general terms, the same kinds of strategies for realizing a request are available in all languages, it is still the case that the specific realization of these strategies differs from language to language.So, too, does the subset of conventionally indirect strategies which are considered to be the standard or preferred ones for performing requests indirectly in a particular language.
According to Wierzbicka (1991 :26), the claim that all languages share exactly the same strategies for realizing speech acts indirectly is just one more example of the mistaken assumption that Anglo-Saxon conventions hold for human behaviour in general.
Having said that, we have in fact, also partly answered the third of our questions concerning the universality of illocu-tionary competence (cf.(4iii) above), i.e. the question of whether all languages make available the same linguistic options for encoding the various pragmatic strategies by which a given speech act may be realized.
As we have just seen, the answer to this question must be negative, at least as far as conventionally indirect strategies are concerned.
There is abundant evidence in the literature that, even when closely related languages share an indirect pragmatic strategy, it may be the case that they encode this strategy differently.
A comparison of the different linguistic forms by which the indirect request strategies of questioning the addressee's ability or willingness to perform the desired act are encoded in English and Hebrew, according to Blum-Kulka (1982:34-35), will serve to illustrate this point.
a. Ability questions ii. --  In (7) above a dash ("_-.") in a given box means that an utterance with the linguistic form concerned cannot be used to realize a request in that particular language.It should not be taken to mean that such an utterance is not a possible utterance in the language.
Will you be able to do X? (cf. (7 a iii)), for example, is a possible utterance of English, but it will not be interpreted by speakers of English as a request to do X.Rather, it will be interpreted as a genuine question concerning the addressee's ability to do X, illustrating that an utterance with a particular conventional illocutionary force in one language may lose this force when translated into another language.
No linguist, to my knowledge, has defended the claim that if one language uses a particular syntactic structure to encode

Sociolinguistic competence
We turn now to the second set of questions that were raised in section 2: questions concerning the universality of aspects of sociolinguistic competence, or knowledge of the relationship between what and how on the one hand, and when and to whom on the other hand.
The first question to be considered (cf.(Si) above) is whether the relationship between contextual factors and the choice and linguistic realization of speech act strategies is the same across languages and cultures.To make this question more concrete, consider the options available to a speaker of English who wants to make a request in a given situation.
The speaker first of all has to make a choice from among nine different pragmatic strategies ranging on a scale of directness from direct and explicit, through conventionally indirect, to highly indirect, as shown in ( 6) above.
Having chosen a strategy, the speaker then has to decide on the precise linguistic form by which the strategy is to be encoded.For example, having chosen to realize the request by means of an ability question, the speaker has to decide whether the question should be phrased by means of can you or could you, whether to address the hearer as sir, or old chap, whether or not to use slang, etc.
The question, then, is whether and to what extent the relationship between pragmatic choices such as those outlined and aspects of the context within which a speech act is performed can be assumed to be constant across languages and cultures.This question is perhaps the easiest one to answer: no speech act theorist that I know of has been prepared to deny that languages and cultures differ significantly with respect to both what speech acts ought to be, ought not to be, or may be performed in what contexts, and how a given speech act is to be performed in a given context.
Factors such as the sex, age, status and authority of the speaker and addressee, their familiarity with each other, whether the speech act is performed publicly or privately, orally or in writing, the topic and the actual setting all influence the ways in which speech acts are realized.But the precise way in which each of these factors influences the realization of a given speech act differs from society to society, and from one culture to the next.
For example, in a comparison of the requests of speakers of British English and those of Spanish speakers, Rintell (1981: 15) found that Spanish speakers, but not English speakers, Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 were significantly more deferential when making requests of addressees of the opposite sex than when making requests of addressees of the same sex.
A study by Beebe (1985) of refusals in Japanese and American English, respectively, has shown that the status of the addressee has a much stronger influence on the form of refusals in Japanese than in American English.8 Examples of studies showing that different pragmatic choices reflect the assignment of different weights to the same social and situational variables in different languages and cultures can be proliferated.However, I can do no more here than to refer the interested reader to the extensive overview provided in (Wolfson 1989:ch. 4, 7).
The fact that the conventions determining the choice of strategies and forms for the realization of particular speech acts in particular situations are undoubtedly language-and culture-specific have not deterred linguists from hypothesizing that, underlying these surface differences, there may be universal norms or motivating principles to which particular pragmatic choices are systematically related across languages and cultures.This, of course, brings us to question (5ii) above: Is the relationship between social norms, or principles, and the choice of particular speech act strategies the same across languages and cultures?Let us consider one particular norm, or principle, that has been hypothesized to be universal and therefore capable of explaining aspects of the speech act performance of speakers cross-linguistically and cross-culturally, viz. the principle of politeness.The content of the notion 'politeness' is not as clear as it would seem at first blush.However, as the content of the notion is highly theory-dependent, a full clarification would take us far beyond the scope of this paper.9 I will therefore concentrate on one particular account of politeness.
In this account, Brown and Levinson (1987)  .
desire of the individual to be liked and approved of, and negatively as the individual's desire not to be imposed upon.
Some speech acts, such as directives, are considered to be intrinsically imposing and therefore threatening to the face of the addressee.
The seriousness of a face-threatening act is determined by the interplay of three independent and culture-sensitive variables: (i) the social distance between the speaker and hearer, i.e. their degree of familiarity and solidarity, (ii) the relative power of the speaker with respect to the hearer, i.e. the degree to which the speaker can impose his or her will on the hearer, and (iii) the ranking of the size of the imposition, i.e. the degree of the hearer's conventionally recognized obligation to provide the goods or services, or to perform the actions concerned, the right of the speaker to impose, and the degrE!e to which the hearer welcomes the imposition.
The choicE! of certain strategies rather than others to perform potentially imposing, and therefore face-threatening, speech acts is seen, then, as an attempt by the speaker to reduce the threat to the hearer's face, or to "soften" the imposition on the hearer.For example, and English speaker saying I would appreciate it if you would shut the door.rather than Shut the door!. implicates not only a request, but also the desire to be polite.
The question, now, is whether there is a systematic relationship between the choice of specific speech act strategies and the desire to be polite, and whE!ther this relationship holds universally.Searle (1975:641 maintained that " ordinary conversational requirements of polite: ness normally make it awkward to issue flat imperative sentences (e.g.Leave the room) or .explicitperformatives (e.g ..I order you to leave the room), and we seek therefore to find indirect means to our illocutionary ends (e.g.I wonder if you would mind leaving the room).
In directives, politeness is the chief motivation for indirectness." The implication of Searle's claim is that there is a systematic and universally stable relationship between a speaker's desire to be polite (and a hearer's recognition of this desire), on the one hand, and the degree of (in)directness of the strategy chosen to realize the speech act.In the Polish culture, by contrast, attributes such as warmth, sincerity and affection are more highly valued than personal autonomy.Therefore, the choice of speech act strategie~by speakers of Polish can never be adequately explained with reference to a norm such as politeness.Rather, a different norm must be used: one which reflects Polish cultural values rather than Anglo-Saxon ones.

Conclusion
I have tried to identify very briefly some of the claims that have been made regarding the universality of aspects of pragmatic competence.10 I have also trief to show that claims such as these may not be correct.
In doing so, I focused on one particular line of argumentation against these claims and on the kind of evidence on which the argumentation is based.
It was not my aim to be complete or balanced in my overview.
This would have been impossible, given the limited scope of Their views, if correct, have implications for language teaching in a multilingual and multicultural society such as ours.Some of these implications are considered in section 4 directly below.
4 Implications for language teaching In section 3 we saw that many aspects of pragmatic competence which were initially hypothesized to be universal have since been argued to be language-specific or culture-specific.The question that now arises is why the issue of universality, a linguistic issue, should be of interest to language teachers.
To answer this question, let us consider what consequences it would have for second language learners if teachers wrongly assumed aspects of pragmatic competence to be universal when they were in fact language-or culture-specific.
A first possible consequence of wrongly assuming to be universal, aspects of pragmatic competence that are in fact language-or culture-specific, is that the task of the second language learner may be seriously underestimated.It may be assumed, for instance, that a second language learner of English already knows what it means to request, to insist, to hint, to suggest, etc., whereas this may not be the case.
Rather, it may be that these speech acts are not conceptualized in the same way in the learner's language or culture as they are in the target language.
To take another example: it may be assumed that the learner already knows the basic strategies In fact, however, there may be considerable differences between the ways in which speech acts are realized in the learner's mother tongue and the ways in which they can be realized in the target language.
As for the sociolinguistic aspects of pragmatic knowledge, learners may wrongly be believed to operate with the same assumptions concerning how to realize what speech acts when and to whom as target language speakers.In fact, however, there may be significant differences between the learner's mother tongue and the target language, reflecting differences in the social 12 and cultural norms of the two groups of speakers. A

"
if there is anything universal about rules of speaking, it is the tendency of members of one speech community to judge the speech behavior of others by their own standards.It is exactly this lack of knowledge about sociolinguistic diversity which lies at the root of most intercultural misunderstanding." In a linguistically and culturally diverse society all speakers need to be made aware of the diversity of social and cultural value systems and in the ways in which they are expressed through language.In Thomas's (1983:110) words, "Helping students to understand the way pragmatic principles operate in other cultures, encouraging them to look for the different pragmatic or discoursal norms which may underlie national and ethnlc stereotyping, is to go some way towards eliminating simplistic and ungenerous interpretations of people whose linguistic behaviour is superficially different from their own."This is the task of those i~volved in language teaching.The linguist's task is to undertake the research that is necessary to ensure that those involved in language teaching are as wellinformed about pragmatic aspects of language as they are about grammar.
An excellent overview of different interpretations and uses of the term "communicative competence" is given in (Taylor 1988).

2.
Throughout this paper a terminological distinction will be made between "competence" and "performance".The term "competence" will be used to refer to knowledge of (various aspects of) language, i.e. to what speakers know about language.The term "performance", by contrast, will be used to refer to what speakers do when they use language, i.e. to the observable result of the application of knowledge of language.
It is of course legitimate to ask whether grammatical competence and pragmatic competence are cognitive capacities of the same sort.This issue will not be addressed here.The interested reader is referred to (Chametsky 1992) for a discussion of different views on this issue.

3.
I realize that, in practice, there may be differences between the varieties of English used by South Africans of different linguistic and cultural backgrounds and, hence, that the use of a term such as "South African English" to denote anyone variety may be objectionable.
However, not wanting to become entangled in this particular controversy here, and for ease of exposition, I will continue to use the term "South African English" (SAE) to refer to the variety of English which is currently assumed, rightly or wrongly, to be the standard and hence the target for English instruction in South African schools.

( 1 )
above, i.e. the question of what linguists are talking about when they use the term "pragmatic competence".Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 2 Pragmatic competence Although different answers have been given to the question of what kinds of knowledge constitute pragmatic competence some more detailed than others there is broad consensus among linguists that pragmatic competence includes knowledge of what speech acts can be performed in the language, what linguistic means and forms are available for encoding a given speech act, and what the social and situational conditions are for its appropriate performance.
determining the choice of the expression Good evening.sir! Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 -------------~~-~--~~~~-------,-----_ .. '---------------64 .rather than one of the available alternatives, depending on who is being greeted by whom and in what circumstances, is part of sociolinguistic competence.Far example, the speaker has to know that in SAE the expression Good evening.sir! may be used to greet someone whom one encounters late in the afternoon, in the evening, or even late at night, but that it cannot be used to greet someone whom one encounters for the second time in the course of the same evening.This is knowledge concerning the relationship between linguistic expressions and situational factors such as the time and circumstances of the encounter.The speaker also h.as to have knowledge of how social factors such as the relationship between him-or herself and the addressee, their respective ages, rights, obligations, etc. influence the choice of an utterance.Thus the utterance Good evening.sir! would normally be judged inappropriate if used by an adult native speaker of SAE to greet a lover, a friend or a child, whereas Hi! would be judged quite appropriate.As was mentioned earlier, an important question from the point of view of second language acquisition is whether and to what extent various aspects of pragmatic competence are universal.
define politeness as the manifestation of respect for and consideration of another's face."Face" is defined both positively as the Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 circuited and the hearer is saved the trouble of having to work out the intended meaning as he or she would have to do in the case of nonconventionally indirect strategies.But, at the same time, the speaker has indicated a desire to be polite by being indirect.Reporting on the results of a study conducted within the framework of the CCSARP project, does caution that the nature of the relationship may differ across cultures.BothThomas (1983) andWierzbicka (1991) have questioned the validity of claims such as those that we have been examining.They argue that claims such as those made by Searle and ch. 2 and 3) argues on the basis of empirical evidence from languages such as Polish that direct strategies (such as the use of imperatives or speech act indicating verbs) are more polite in some languages and cultures Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0Thomas (1983:106ff) argue that norms other than politeness may be the chief motivation for the choice of particular speech act strategies in other languages and cultures: norms such as cordiality, truthfulness or sincerity.The emphasis on politeness, defined as respect for another's face, reflects the high value placed on the autonomy of the indivi- general theory of speech acts proposed by Austin and Searle, to the exclusion of valuable work done within other theoretical frameworks.The choice was motivated by the fact that the Austin-Searle theory has generated such an immense body of research and is still considered a point of departure for work on speech acts even by those who have adopted a different framework.11 The general tenor of the line of criticism that I have been focusing on is that claims about the putatively universal Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 nature of aspects of pragmatic competence may reflect no way intimating that I agree with it, or that the arguments offered are valid arguments, or that it is the only possible line of criticism.I have focused on this line of criticism because it is important that language teachers and others involved in language teaching take note of views such as those expressed by Thomas, Wierzbicka and others.
for realizing speech acts and that he or she merely needs to Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 learn how these strategies are linguistically encoded in the target language.
Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 Could you •.• ?.•.Would you be so good as to ••• ?.•.Would you be so kind/gracious as to •.. ?But ••• pseudoquestions which ostensibly enquire about the addressee's desire and which in fact are to be interpreted as requests (Would you like to.Do you want Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 given speech act are the same across languages.ization of two speech acts, viz.requests and apologies.5Theyassumethatarequestcanbe realized by producing an utterance of one of the following types:(6 )i.an utterance in which the grammatical mood of the verb (viz.theimperativemood)signals the illocutionary force, e.g.Leave me alone!; ii. an utterance in which the illocutionary force is explicitly named, e.g.I am asking you to clean up this mess; iii.an utterance in which the naming of the illocutionary force is modified by hedging expressions, e.g.I want to ask you to give your presentation a week earlier; iv. an utterance in which the hearer's obligation to carry out the act is stated, e.g.You'll have to move your car; v. an utterance in which the speaker's desire for the ~ct to be carried out is stated, e.g.I really wish you'd stop bothering me; vi.. an utterance in which it is suggested that the hearer carry out the act, e.g.How about cleaning up this mess?; vii.an utterance containing reference to the preparatory conditions (such as the hearer's ability or willingness to do the act) for the successful per-formance of a request, e.g.Can you clear up the kitchen for me?Would you mind moving your car?;Fraser, Rintell and Walters (1980:78-79)have made the strongest claim, hypothesizing that all languages make available the same Stellenbosch Papers in Linguistics, Vol. 25, 1992, 61-88 doi: 10.5774/25-0-76 set of strategies for performing a given speech act.Moreover, the conditions that have to be' satisfied for an utterance to count as a request are claimed to be essentially the same across all languages.An example of such a condition is the one that stipulates that an imperative utterance can count as a valid request only if the hearer is in fact able to perform the desired act (so that, e.g., Come here!, but not Drop dead!, would count as a valid request in English)."onecould perform requests, or acts closely related to requests, by ostensibly 'asking' about the addressee's ability to do something, or about his goodness (or kindness): .•.