janralfl~ ELSEVIER
Journal of Pragmatics 24 (1995) 283-294
Discussion note
On 'new/non-native' Englishes: A quartet Rajendra Singh a, Jean D'souza b, Karuvannur Puthanveettil Mohanan c, Nagur Sheshagiri Prabhu d aFacult~ des arts et des sciences, D~partement de linguistique et de traduction, Universit~ de Montreal, C.P. 6128, succursale A, Montreal, Quebec, Canada H3C 3J7 b37 Shardaram Park, Sasoon Road, Pune, India ~Department of English, National University of Singapore, 10 Kent Ridge Crescent, singapore 0511 d302 Sovereign Park, Bangalore, India 560005
Received March 1994
1. Introduction This paper is a somewhat stylized reconstruction of a discussion amongst its authors regarding the so-called 'new/non-native' varieties of English. In our view, it brings out some of the central issues involved in the study of varieties such as Indian English (IE) and Singapore English (SE). Although sections 2-6 were written individually, they reflect the essence of our dialogic interaction. We offer this reconstruction as an invitation to others interested in these varieties to help us all to arrive at a better understanding of the issues involved.
2. Initiation (Rajendra Singh) Discussion of World Englishes frequently makes a division of the different Englishes (or varieties of English) into two groups, labelled 'Old/Native/Inner Circle' Englishes/Varieties, on the one hand, and 'New/Non-native/Outer Circle' Englishes/Varieties on the other. What justifies this division? In particular, is there any linguistic justification for it? English, like any other language, has several varieties, each defined and delimited by the praxis of those who can be said to share it; and English, like various other languages (including Latin and Sanskrit in the past) has varieties in more than one national or geographical unit. Why should this fact give warrant to a two-way classification of the varieties involved? 0378-2166/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 037 8 - 2 1 6 6 ( 9 4 ) 0 0 0 4 6 - H
284
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
The actual classes set up, and the labelling employed, suggest three possible notions behind the act: (a) 'transplanted' varieties (such as American and Australian Englishes) continue to belong with those in their historical homes while 'indigenized' ones (such as IE and SE) do not; (b) mother tongue/monolingual speakers' Englishes are qualitatively different from those of 'other tongue'/bilingual speakers; and (c) historically older Englishes are more truly 'English' than historically younger ones. It may be noted in passing that there is a conflict between the first and the last notion, in that some 'indigenized' varieties have a longer history than some 'transplanted' ones. For any of these notions to have empirical substance, it is necessary to produce evidence such as that 'transplanted' varieties follow a path of diachronic development different from that of 'indigenized' and 'stayed-back-home' ones; or that bilingual speakers' Englishes all share some structural properties not shared by any monolingual speakers' English; or that some defining characteristics of an entity called 'English' are consistently present in one class of Englishes and absent in the other. Showing that a particular English in one class (say, IE) differs in some lexical, syntactic, morphological or phonological matters from a particular English in the other (say British English (BE)) can only be an act of assuming and exploiting the classification, not one of justifying it. To be an act of justification, the demonstration needed is that (a) all Englishes in one of the classes differ in similar ways from all Englishes in the other; and, equally, (b) no two Englishes within the same class (in particular, within the class of 'transplanted' and 'stayed-back-home' Englishes) differ between them in those ways. In the absence of this latter demonstration, any differences shown to exist between IE and BE have neither more nor less weight than any differences between IE and Nigerian English, American English and Australian English, or SE and American English. All such differences lead only to the following picture: English American
Nigerian
Singapore
Australian British etc.
Fig. 1. not to the following:
English 'Old Native'
'New/Non-Native'
American Australian British etc.
Indian Singapore Nigerian etc. Fig. 2.
All the evidence available in the literature (a) compares one of the Englishes in the 'new/non-native' class with one of those in the 'old/native' class, thus assuming Fig. 2, and (b) shows only that the two Englishes differ in certain ways, thus justifying Fig. 1. The enterprise as a whole seems to be based on this fallacy in reasoning.
R, Singh et al. / Journal of Pragmatics 24 (1995) 283-294
285
Though no justification is offered for Fig. 2, conclusions based on an assumption of Fig. 2 are readily drawn, such as the following: (a) Speakers of some Englishes (e.g. British, Australian) are native speakers of 'English' while those of others (e.g. Indian, Nigerian, Singapore) are not. (Given the only synchronically sustainable concept of a native speaker as one who shares with others in the relevant speech community relatively stable wellformedness judgements on expressions used or usable in the community, the only conclusions justified by Fig. 1 are of the form: speakers of BE are not native speakers of IE; speakers of American English are not native speakers of Australian English; speakers of IE are not native speakers of Nigerian English, etc.) (b) Some Englishes (American, Australian) can/do have their own internal standards, distinct from that of their historical home, while other Englishes (Nigerian, Indian) cannot/do not. (The unargued 'can/cannot' declaration has the effect of clouding the more factual 'do/do not' question.) (c) Bilingualism or bidialectalism in speakers of some Englishes (British, Australian) does not affect their native-speaker status in those Englishes, while bilingualism or bidialectalism in speakers of other Englishes (Singapore, Indian) deprives them of that status. (The further concept that a single language system in the monolingual's mind is somehow superior to any of the different language systems in the bilingual's mind can be matched with the equally possible, and equally untested, concept that the monolingual mind's capacity for further language systems is in an undeveloped state). Since (a) all such conclusions have the common import of projecting some Englishes to be more equal than others, (b) all of them crucially assume Fig. 2 above, and (c) Fig. 2 is unsupported by any linguistic evidence or argument, it is reasonable to conclude that the only motivation for Fig. 2 is a wish to distribute equality unequally. It looks as though all it takes to propagate a prejudice is a convenient graphic representation and clever labelling.
3. First response (Jean D'souza) I believe that all researchers in the field of World Englishes would accept the tree in Fig. 1 and not that in Fig. 2. However, the fact that some branches of the tree have traditionally been treated as less equal than others has led to these 'poor relations' being grouped together as 'new' varieties in the attempt to assert their equal status. It may be noted that all labels are problematic though some are less value-laden than others; 'new', 'outer circle' are less judgmental than 'non-native' (whatever that means in the context of English). All varieties, other than BE, are transplanted and all are nativized. The attempt to assert the equality of all branches, while necessary, has not met with much success so far because most studies of the new varieties of English (NVEs) have tended to use difference as a means of defining these varieties. Difference has inevitably meant difference from BE, thus tacitly accepting BE as some kind of yardstick. As stated in section 2, this is a fallacy in reasoning.
286
R. Singh et al. /Journal of Pragmatics 24 (1995) 283-294
To show that Fig. 1 is indeed correct, one would have to study and describe each branch of the tree as a system in its own right and not with reference to any other branch. Each variety should be seen as a totality and not as a set of differences. After holistic descriptions become available, the crucial step in the NVE contexts is the preparation of reference materials based on the description of the given variety. These reference materials will in effect be a manifestation of the standards in place within the community because, as with any reference materials, they will highlight the contrast between written/formal usage and spoken/informal usage. Such materials are crucial because until internal standards are acknowledged, English in the NVE context will always be 'their' language, not 'ours'. Speakers of English in the NEVE contexts often do not accept that they are native speakers of their varieties and the availability of reference materials based on their own usage may go a long way to counter the notion that BE is the only possible norm. The term World English will only make sense when all branches of the tree in Fig. 1 make an equal contribution to any description of English.
4. Second response (Karuvannur Puthanveettil Mohanan) Sections 2 and 3 correctly point out that the collection of varieties grouped under 'new/non-native' does not have any structural properties that distinguish it from what is grouped under 'old/native'. Granted that there is no structural justification for the grouping in Fig. 2, we may still ask if there is some other justification for a grouping that includes IE and SE but excludes Australian English and American Black English. The discussion in section 3 makes a case for new/non-native on the grounds of social injustice. I would like to make a case for the grouping as constituting a domain of inquiry, on par with domains of enquiry for phenomena such as second language acquisition, bilingualism, pidgins, and creoles. In Quebec, there is a community of bilingual speakers who use English and French as their first language. In Kerala, there is a community of bilingual speakers who use Malayalam as their first language and English as their second language. The major difference between the two bilingual systems is in the order of acquisition. In Quebec, some individuals acquire English and French simultaneously, while in Kerala, individuals acquire Malayalam first and English later. Now, a community of researchers may decide to focus their research on bilingual systems of the second type, where one system is acquired after the other. There is nothing illegitimate about the decision to investigate this domain. I take it that there is no disagreement about the proposition that entities like SE and IE constitue varieties of English (Fig. 1). The disagreement is about the label 'new/non-native' (Fig. 2). One possible way of providing substance to this term is in terms of the order of acquisition: (i) For a given speaker, a non-native system is one that (s)he has acquired after the acquisition and stabilization of some other linguistic system. (ii) A variety V of a language L is a non-native variety iff L is a non-native system for the majority of speakers of that variety.
R. Singh et al. / Journal of Pragrnatics 24 (1995) 283-294
287
Consider SE. For most, though not all, Singaporean speakers of English, English is acquired at a later stage in their development. Hence, by definition (i) these individuals have non-native systems of English. By definition (ii), SE is a non-native variety. If it turns out that, in fifty years, most speakers of SE acquire it by the age of, say, five, then SE would have become a native variety. If we take this position, 'non-native variety' (Fig. 2) is a perfectly legitimate concept in linguistics. A linguist may decide to specialize in non-native varieties because (s)he wants to investigate the acquisitional or historical problems in these varieties, with the hope that they may provide types of evidence on the human language faculty, not typically available in other domains. It is not the case that "the only motivation for Fig. 2 is a wish to distribute equality unequally" (section 2). The Collins Cobuild Dictionary of English defines a native speaker of a language as "someone who speaks that language as their first language rather than having learnt it as a foreign language". It is this meaning of '(non-)native' that is present in the definitions above. The discussion in section 2 deviates from the traditional meaning of the term and proposes that a 'native speaker' is "one who shares with others in the relevant speech community relatively stable well-formedness judgments on expressions used or usable in the community". Shared stable judgements may be a necessary condition for calling something a 'variety of L', or a 'stable variety of L', but this is hardly the sense of 'non-native' either in ordinary discourse or in the technical discourse of linguists. To turn to a different matter, the issue of some variety being a native variety should be distinguished from the issue of that variety being a standard variety. By definitions (i) and (ii), we should treat SE as a non-native variety; the social facts demand that we treat it as a non-standard variety. The justification given in section 3 for the label 'new variety' is that it is needed to rectify social injustice: some branches of the tree in Fig. 1 have been treated as poor relations. Implicit in this argument is the idea that new varieties are those which have less social prestige, and are socially disadvantaged. If so, New York Black English and Cockney English are also new varieties of English. The term 'new variety' then means the same as 'nonstandard' variety. There is no need to introduce a new term for a concept for which a traditional term already exists. A related issue is the plea for local standards (section 3). In a perfectly uniform speech community, where all speakers have identical internal linguistic systems, the notion 'standard' is meaningless, since there is no single variety that can be considered socially more prestigious than others. In other words, the existence of a standard variety entails a social asymmetry, which can result in injustice. If so, the best way of ensuring social equality is to abolish standards altogether, not to institute local standards. Let us conduct the following thought experiment. In the year 2050, Californian English becomes the standard variety of English for the whole world, either because Californians become socially and economically powerful, or because some international academic body decides to promote Californian English. With respect to the English-speaking community of the world, such a standard would be endo-normative, for what is chosen as standard is one of the varieties within the community. With respect to speakers of BE and IE, however, the standard would be exo-normative,
288
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
because Californian English is not one of the varieties of BE or IE. The speakers of BE, IE, etc. will be at a disadvantage. In order to prevent this injustice, let us replace the notion 'Standard English' by a set of local standard Englishes such as Standard American English, Standard BE, Standard Australian English, and Standard IE. Suppose we choose one of the varieties of IE, say Malayalee English, as Standard IE. Then, with respect to the community of 1E, Standard IE will be endo-normative. However, it will still be exo-normative with respect to the sub-communities within IE, say, the community of speakers of Gujarathi English. Choosing Malayalee English as the standard will be an unfair disadvantage to speakers of Gujarathi English. The social reformer's logic of local standards will therefore require us to recognize standards such as Standard Malayalee English, Standard Gujarathi English, and so on. Even this move does not prevent injustice, since there are sub-varieties within varieties such as Malayalee English. The next step will, therefore, be the recongition of Standard Southern Malayalee English, Standard Central Malayalee English, and so on. Extending this logic further, we arrive at the conclusion that the only guaranteed way of ensuring endo-normative standards as a means of preventing injustice is to recognize a standard for each individual, which is the same as abolishing standards. The conclusion that emerges now is the following. Any notion of standard, whether it is for the language community as a whole, or for particular sub-communities within the community, entails an unfair prestige asymmetry. We can either accept the need for standards and live with the social injustice, or refuse to live with social injustice and reject all standards. The intermediate step of calling for local standards in section 3 is at best a plea for a small subset of the disadvantaged group to be admitted to the inner circle, letting the other subsets continue with their disadvantaged position; it is "the manifestation of a wish to distribute equality unequally" (section 2). To conclude: (a) The term 'non-native variety' refers to a perfectly legitimate concept in linguistics as a domain of inquiry. The term 'new variety' is either the same as 'non-native variety', or does not refer to any coherent concept. (b) The plea for 'endo-normative standards' as a means of preventing social injustice contains a logical contradiction. We should be willing either to abolish all standards, or to accept exo-normative standards.
5. Third response (Nagur Sheshagiri Prabhu) The term 'variety' necessarily denotes a shared linguistic system in a speech community. Just as each individual's system can be considered a mental organ, a shared system in a speech community can be considered a social organ. Individual systems are both triggered, in their development, by inputs from the shared system, and constrained, in their operation, by their participation in the shared system. Being expressions of individuals' inherent linguisticality, they have the power to transgress the shared system, constrained only by their inherent linguisticality. Being participants in the shared system, they also have the opportunity to alter (or fail to alter) the shared system, through their own transgressions, or to alter themselves (or not to
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
289
alter themselves) in response to other individuals' transgressions. The act of sharing, which is an expression of human sociality, is at once a constraint on individual transgressions (in that the latter may fail to alter other individuals' systems) and an opportunity for them (in that they may succeed in doing so). The interaction between individual systems and shared systems (their mutual dependence, mutual influence and the resulting states of stability or instability in their relationship) can be conceived of as an 'organic' process (internally and spontaneously regulated, jointly by human linguisticality and human sociality - in a way not different from the 'organic' development of individual systems); hence the term 'social organ'. A speech community, in this sense, is not a collection of individuals with similar individual systems; it is a community that is operating a shared system as a social organ. When one is learning a language, one is attempting to develop an individual system which is a replication of the shared system that underlies the input received. The system one develops taps one's linguisticality and can therefore transgress the shared system, but in so far as one sees oneself (or is seen to be) a learner, not a participant, of the shared system, one's transgressions count as deviations from the shared system (failures to achieve replication), not contributions to/influences on it. There is also the fact that intermediate stages in the development of individual systems (constituting degrees of approximation to the target shared system) are likely to prompt relatively large transgressions, too large and unsettling for the shared system to adopt - hence best treated as learners' errors. Learners' systems, that is to say, are viewed as degrees of approximation to a shared system, not as constituents of it. Success in language learning is known to differ between first language learners and second language learners. First language learning is marked by uniform success (replication of the input system) while second language learning is marked by varied success (different degrees of approximation). It is therefore easy to accept learners in the former category, when they complete the learning process, as participants in the shared system, and exclude those in the latter category as unsuccessful aspirants for that position - hence, perpetually a class of learners, whose transgressions will always be (viewed as) deviations or errors. Suppose, however, a set of people who have learnt English as a second language, perhaps achieving less than a replication, begin to function as a speech community, sharing the individual systems they have developed, and operating the shared system as a social organ, as outlined at the beginning of this section. Their shared system is different from that which they once attempted to replicate but, unless one attributes to them a smaller measure of linguisticality and sociality, it is as truly a shared system as any other, and is as fully replicated by successive generations as any other. One can make reference to the phylogenetic fact that the shared system arose from individual approximations to another shared system but any notion that, because of that fact, the shared system itself is an approximation (i.e. failed replication) to the other system will need to be justified by evidence of the kind sought in section 2. One sense of the term 'native speaker' has to do with whether the language was learnt as a first language, and the implication of the term (in 'ordinary discourse', whether or not in what in section 4 is referred to as "the technical discourse of linguists"), is that the speaker concerned has a full replication of the shared system
290
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
concerned. Based on this, the term 'non-native speaker', understood as NOT (NATIVE SPEAKER) is used to refer to someone who, having learnt the language as a second/foreign language, has achieved less than a replication (though The Cobuild Dictionary is too polite to spell this out). The next step that is taken is to extend the less-than-a-replication sense to all shared systems/varieies which once arose form individual approximations to another shared system, by referring to them as 'non-native varieties', understood as NOT (NATIVE VARIETY). It is then easy to re-use the term 'non-native speaker' to refer to all those who speak these 'nonnative varieties', thus treating them (over successive generations, regardless of full replications of the shared system) as instances of failed replication. This latter sense of 'non-native speaker' (which is also a part of ordinary discourse) has two notable features: (i) it ignores the implications of a variety as a shared system, pointed out at the beginning of this section or, rather, it ignores them for some shared system and not for others, since ignoring them for what it regards as 'native' varieties would make all changes that take place in those systems deviations or failed replications; (ii) by using the further terms, 'native and non-native speakers of English' understood as (NATIVE SPEAKER) OF ENGLISH and NOT ((NATIVE SPEAKER) OF ENGLISH), it appropriates the terms 'English' for those shared systems which it regards as native varieties, denying it to those regarded as non-native, as pointed out in section 2. Anyone who does not wish to participate in this tendentious extension of terms is justified in refusing to use them and to speak instead of 'native speakers of variety X', 'native speakers of variety Y', etc., all of them constituting native speakers of English. The term 'native speaker of variety X' is then best seen as referring to a participant in the shared system concerned, sharing of wellformedness judgements being a good indicator of the participant role. The only nonnative speakers, then, are those who are not participants in the shared system, either because they are (viewed as) learners of the shared system concerned or because they are participants of some other shared system rather than of the one in question. The definitions in section 4 of native/non-native systems and native/non-native varieties do not seem to carry any implications of failed replication. Nor do they recognize the implications of a variety/shared system selectively for some shared system (though they seem to me to fail to recognize them fully for any shared system). They do not appropriate the term 'English' for some varieties, denying it to others: I take it that the terms 'non-native system', 'non-native variety' and 'nonnative speaker of English' mean (NOT NATIVE)SYSTEM, (NOT NATIVE) VARIETY and ((NOT NATIVE) SPEAKER) OF ENGLISH, respectively. Indeed, although the definitions are couched in terms of sequence of acquisition (first vs. second language), they disclaim any implications of the acquisitional sequence for the structural properties of the systems acquired. These therefore are definitions which are radically different from those signalled or implied in ordinary discourse on the subject. They are, of course, "perfectly legitimate concepts in linguistics" as claimed in section 4. What is less legitimate is for them to claim to be in line with ordinary discourse (e.g. to assume that The Cobuild Dictionary definition of a native speaker also disclaims any differences in structural properties from "having learnt it as a foreign language"), and indeed, to be in line with the discourse of section 2: the
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
291
fact that the terms, as defined in section 4, do not make Fig. 2 an unequal distribution of equality does not mean that the terms, as used in the ordinary discourse on the subject, which is what was being discussed, do not do so. In seeking support from ordinary discourse for its own definitions, section 4 seems to me to lend the weight of what it calls "the technical discourse of linguists" in the service of the very different implications of ordinary discourse. I might also add that the putative benefits said to be derivable from such a technical definition of the terms (study of different types of bilingual systems and their uses in society, investigation of the 'acquisitional or historical problems in these varieties', all without any expectation of structural differences in the systems themselves) seem to me to be too far-fetched to justify the act. I see the concept of standards as being related to that of varieties. Language as a social organ has both a centrifugal tendency, leading to the formation of sub-communities and sub-varieties, as expressions of group identities, and a centripetal tendency, which involves the perception of standards. Individuals in a speech community do not just conform to a shared system and alter the shared system through shared transgressions; they also attach differential social values to the conformity and transgression of different sets of individuals - or to those of different sub-communities. The social values attached may arise from a variety of factors, such as social class, economic or political power, level of education or scholarly stature, which may or may not be related to linguistic systems. But, whatever their source, differential values do have an effect on the shared system (or a cluster of shared systems): there tends to be a greater conformity to the more highly-valued individuals' systems, thus bringing about a convergence and countering the effect of group identities. The two opposite tendencies are in a state of tension and the balance can be in different states of stability. Both the centripetal tendency and the centrifugal one are controlled by individuals in the speech community, in the same way as changes in the system are. If group identity asserts itself to the point of negating conformity to the standard, the speech community breaks up into two or more, each then developing its own centrifugal tendency (i.e. standard). An exo-normative standard is one which operates as a standard in one speech community (where it is endo-normative) but is held up/asserted to be the standard for one or more other speech communities (where it is not). It is exo-normative for the latter because individuals in these communities do not control it and its operation is not held in tension by the normal, opposing tendencies. To point this out to those who seek to assert or promote an exonormative standard is as legitimate as it is to point out to those who seek to uphold some shared systems over others the nature of what they are doing.
6. Peroration (Rajendra Singh) If the praxis of researchers in 'the field of World Englishes' shows that Standard BE is still used as a yardstick, in what sense, one might ask, can these researchers be said to accept Fig. 1 ? The malignant face of the posture of grouping some 'poor relations' together is the leaving out of other poor relations, including some back ' h o m e ' . Is it possible that that exclusion is what reduces the "attempt to assert their
292
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
equal status" to the plea that they be recognized as belonging to the 'outer circle'? It is interesting to note that section 4 counterfactually arrives at a fairly accurate description of the actual practice of most researchers in 'the field' in question: they in fact use 'native' and 'standard' interchangeably and treat 'non-standard' back on 'native grounds' and 'new' on 'alien grounds' in a similar fashion. Most of them, including some advocates of new varieties, actually don't even show any awareness of systems other than what is seen as the standard on 'native' grounds. Given this time and this place, the step towards full description is useful, welcome, and perhaps even crucial, but Sanskrit was certainly something to write home about even before Panini, one of its last native speakers. Making sense of a language with its speakers is a joyful activity, and it is a sad comment on our times that some must inventory what is inalienably theirs. Is it possible that the praxis of some scholars can be made sense of only as a counterfactual exercise? The insistence is not, as section 4 suggests, so much on structural classification as on structural consequences of whatever classificatory criteria one uses, a synchronic sine qua non. In the absence of such consequences, 'acquired above/below the Equator' is as good a criterion as the familiar mother-tongue criterion refined and asserted anew in section 4. Although it correctly draws attention to the natural domain of sequential bilingualism, it ignores what is said in sections 2 and 5 and the fact that the actual research on the varieties in question is generally not placed in that context. And even if it were to be placed there, the synchronic interpretation of the criterion remains problematic. The trouble with it is that the class of non-primary ('nonnative') grammars (languages) is known to be non-distinguishable from the class of primary ('native'=mother-tongue) grammars (languages). How grammars arrived at differently manage to generate equally possible languages is interesting but irrelevant for the issue at hand. When the learners' unstable and unshared systems become shared and stable, they become theirs, and they their native speakers. If the innovations in these systems (languages) are guided not by the secure guard-rail of Universal Grammar (UG) but by other 'things', one would expect them to belong to another class. But they don't. Even within the context of 'mother-tongue' acquisition, it is well-known that the primary source of what is called 'historical change' is 'error'. The fact that this error is seen as a constituent of creative replication is, as pointed out in section 5, irrelevant. To insist on a particular acquisitional path in the absence of any criteria that could distinguish between the two classes of grammars in question is an act of faith, not very different from the faith that climatic variation is one of the main causes of linguistic variation. In addition, the licence to practice that faith includes, gratis, the licence to preach the 'deficit hypothesis': if the context and the input matter that much, poor input must lead to poor grammars (languages). Although the naturalness of the domains of monolingualism, sequential bilingualism, and non-sequential bilingualism is not in doubt, monolingualism cannot be made an essential condition for 'nativeness' except in the circular acquisitional sense. Even if the licence correctly denied by section 5 is granted, the legitimacy provided by the definitions offered in section 4 can acquire linguistic meaning only when the
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
293
English varieties in question are studied in conjunction with comparable varieties of, let us say, Dutch and French (and Latin and Sanskrit in the past). The fact that these obvious linguistic connections are not made only underlies the tendentious appropriation highlighted in section 2. Although there is plenty of philological justification for studying any or all of these varieties in question, the linguistic justification for studying only these together remains somewhat opaque, if not mysterious. It is possible to argue that only future research can tell whether the investigation of the domain section 4 seeks to create anew will offer some insights into what is at issue; but the available evidence in that domain makes that possibility much too farfetched indeed. It is also possible to argue that there is no reliable evidence that the Englishspeaking communities in, for example, India and Singapore do not in fact subscribe to the exo-normative Anglo-American norm. Although the issue is an empirical one, one of the things for which the 'new varieties enterprise' must be given credit for is the unearthing of that evidence. It in fact seems to suggest that for these contexts we need to redefine interference variety as 'IE or SE with Anglo-American interference'. The acquisitional tables, arranged neatly in section 4, seem to have been turned, along the well-known lines of 'dialect borrowing'. I am happy to conclude that the trial balloon launched in section 2 can indeed continue to float, largely because the only difficulty encountered is the warning that there is a remote possibility that the elasticity of 'sharedness', the material with which it was built, may exceed the specifications provided by future research in sequential bilingualism. The only way to force the balloon down, it seems, is to show that there is no such thing as IE or SE. If that is the case, one should talk not of IE or SE but of English in India and English in Singapore. But as long as IE and SE can be said to exist, one cannot, I submit, talk about them in ways different from the ways in which one talks about BE or American English. If IE and SE exist, those who, as section 5 puts it, "operate them" operate them exactly as speakers of American English and BE operate their respective systems, i.e. as native speakers. The evidence unearthed by the current new-varieties enterprise, fortunately, leaves no doubt about the fact that IE and SE do in fact exist.
7. Conclusion The reflexions and arguments presented above seem to have been inspired by three different points of view. The author of section 3 (JD) speaks for the 'new varieties' enterprise as it is currently constituted, examines the assertion that varieties such as IE and SE are not varieties of English, and tries to argue that, in a manner of speaking, they are. Its starting point is a counterpoint, its logic the logic of liberalism, and its manner of speaking consistently ambivalent. The author of section 4 (KPM) takes the position that although there may not be any structural differences between varieties such as American English, BE, and Canadian English on the one hand and IE and SE etc. on the other, research in sequential bilingualism may still
294
R. Singh et al. / Journal of Pragmatics 24 (1995) 283-294
provide some justification for grouping the 'new' varieties, together. The other two authors (NSP and RS) take the position that varieties such as IE and SE have, both linguistically and acquisitionally, the same status as varieties such as AE and BE; linguistically, because there are no structural differences; and acquisitionally, because once a non-primary system is shared it ceases to be a non-primary one. Despite disagreements amongst us, particularly regarding 'standards', we seem to have convinced each other at least of the following: (a) Hypotheses of the following sort are empirically interesting: (i) There is no structural feature ct such that all 'non-native' varieties of English have ct and no 'native' variety does. (ii) There is no structural 13 such that no 'native' variety of English has 13 but some 'non-native' varities of English do. (b) Given what we know, the stronger of the two hypotheses above, namely (ai), seems correct. (c) Although the current new-varieties enterprise has un-earthed some interesting facts, it has not been able to turn those facts into evidence for what they may in fact constitute evidence for. It should, we feel, situate some of those facts in the context of diachrony and phylogenesis and others in the context of hypotheses such as (i) and (ii) above. (d) The reason that that enterprise stops where it does is that it takes the 'transplanted/indigenized' dichotomy for granted and treats, unreflectively, 'stayedback-home' or 'transplanted' 'standard' to mean 'native', shortchanging what it claims to defend as well as other 'stayed-back-home' varieties. (e) Unfortunately, it also refuses to explore the implications of sharing though it does pay lip-service to it. (f) It is only by problematizing its findings within the parameters involved in (a), (c) and (e) that the enterprise in question can gain some legitimacy.