But an option not on offer so far (and, of course, a task impossible for a corpus called the British National Corpus) is to give these non-native speaker teachers access to a corpus capturing the successful use of English among non-native speakers, as a lingua franca, thus offering supremely relevant models for many learners wishing to use the language for similar purposes. So when Aston and Burnard refer to ?the political implications of changing the basis on which assessments of correctness or appropriateness of usage are made? what has changed about the "basis" is how it can be accessed, not how it is defined. There is also another problem that operates at a deeper and unrecognized level: the language attitudes of those who, paradoxically, are themselves re- commending the challenge to native-speaker norms. This is evident in the contradictory statements made by those such as van Els, who, in the same essay, claims on the one hand that the ownership of a lingua franca passes to its non-native speakers and on the other that the Dutch should not be satisfied about their English because ?simply very few are able to attain a level of proficiency that matches the native or native-like level? (2000: 29). Similarly, Hoffman (2000: 19) calls the English of European learners as straddling ?the whole spectrum from non-fluent to native-like, as if fluency in English were not an option for those whose speech does not imitate that of a native speaker. In other words, non-native speakers own the English which they speak, but unless it conforms to native speaker norms, it is un- acceptable. English as a world language is to be judged as if it were English as a native language. No change there, then. The abstract nature of the proposals put forward by Pennycook (above), for example, has done little to allay the sense of insecurity and unease among English language teachers about what is for them the most critical issue: that of the language norm which they teach - the main basis of their professional qualification, the hub around which their daily practices revolve. Widespread politically correct rhetoric is no effective antidote for this unsatisfactory situation, and so the familiar chip-on-the-shoulder syndrome among non-native teachers of English persists.

In a paper which addresses topics of linguistic as well as literary interest, it seems apposite - though admittedly unconventional - to point out parallels between the writings of two well-known authors in language teacher education and postcolonial literature:

We suffer from an inferiority the effect of the cultural bomb is to complex caused by glaring defects annihilate a people's belief in their names, in our knowledge of English. We in their languages, in their environment, are in constant distress as we [...] in their capacities and ultimately in realize how little we know about themselves. It makes them see their past as the language we are supposed to one wasteland of non-achievement and it teach. makes them want to distance themselves (Medgyes 1999: 40, our em from that wasteland. It makes them want phases) to identify with that which is furthest removed from themselves; for instance, with other people's languages rather than their own. (Ngugi 1981: 3, our emphases)

In an essay whose main point is the renegotiation of the customary distinction between the Outer and Expanding Circles, it is interesting to note that Medgyes comes from the latter and Ngugi from the former. Both, however, share the assumption of the uniformity of English and seem to deny the inherent flexibility of the language, its adaptability to change: English is English. The distress expressed in Medgyes' book, whose objective, after all, is to foreground the particular strengths of non-native language teachers, indicates that (in Ngugi's words) these teachers' ?belief [...] in themselves? has been "annihilated," and we think it is not an exaggeration to say that the ?inferiority complex? ascribed to these teachers on the basis of the ?glaring defects? In their ?knowledge of English? causes them to ?see their past as one wasteland of non-achievement.

We would argue that what is most likely to arrest this negative spiral is not giving non-native teachers pep-talks about their linguistic human rights, nor access to ever-larger native-speaker corpora Rather, what is required is a reconceptualization and concomitant description of ?the language they are supposed to teach? In terms of what it predominantly...

As long as no empirically-based description exists of how English is actually used as a global lingua franca, the merely, hence obvious, descriptive reality accessible when speaking about 'English' is ENL.
In the remainder of this essay, we will therefore argue that it is necessary and feasible to conduct a conceptual and empirical enquiry into the actual nature of ELF. We will summarize relevant work already undertaken, offer suggestions for further enquiry and point to some socio-psychological and pedagogical implications. Of course, no empirical investigation starts from scratch; there are always preconceptions and predecessors. Notions of how characteristics of English as an international language may be captured and how the language may be modified or simplified have some history which it would be foolish to ignore. There are points of reference in the past which will help us in our enquiry: important work, along conceptual as well as empirical dimensions, directed at identifying salient features of EL F. use has already been achieved. A first tradition of research on track from the early decades of the twentieth century however is now all but forgotten: notably Ogden's Basic English (eg, 1930), Palmer & Hornby's Thousand Word English (1937) as well as West's s empirically derived Service List (1953).

These anticipate, and offer many profound insights into, many of the EWL issues that we are confronted with today.4 in addition, many theoretical and methodological problems discussed in reference to varieties captured in the ICE corpus are also highly relevant for the description of ELF (Mair 1992). In recent years, a small number of descriptions and analyses of selected aspects of ELF use have been conducted, in particular in the area of the (intercultural) pragmatics of 'non-native - non-native' communication in English (House 1999 and 2002). James (2000) offers a conceptually rich discussion of the place of English in bi/multilingualism, making allusion to a project, at present in its pilot phase, called ?English as a lingua franca in the Alpine-Adriatic district. He also outlines hypotheses as to what findings the future analysis of this use of English by speakers of German, Italian, Slovene and Friulian might yield. Last (but not least, we think) is our own work, focusing on ELF phonology (Jenkins 1998, 2000, 2002) and lexico-grammar (Seidlhofer 2000, 2001a, 2001b, 2001c).

The work on phonology, culminating in the 'Lingua Franca Core', takes as its starting-point the need for empirical data drawn from interactions between L2 speakers of English in order to assess which phonological features are (and which are not) essential for intelligible pronunciation when English is spoken in lingua franca contexts. This data, then, replaces NS intuition and data drawn from NS-NS (or NS-NNS) interactions, both of which up to now have formed the basis of pedagogic pronunciation decisions (Benrabah's 1997 study of word stress is one of very many such studies). If the concern is with intelligibility among NN Ss of English, which is by definition the case in ELF, however, it makes no sense whatsoever to look to non- EL F. contexts for evidence of such intelligibility.

These items are typical of those occurring time and again in the different types of phonological data collected by Jenkins and they did not affect intelligibility for a NN S. listener regardless of the L. 1s of the interlocutors. Of course this interaction relies heavily on a shared context, which limits the potential for misunderstanding and conflict, and in many situations in which EL F. is used such favorable conditions will not apply. But this caveat does not invalidate the observation that, for the purpose at hand, the kind of English that is employed works and it serves the participants quite adequately for doing the job they have to do. The investigations we have carried out so far have confirmed that a great deal of ELF communication is conducted at comparable levels of proficiency, and that rather often it is features which are looked upon as 'the most characteristically English', for example third-person -s, tags, phrasal verbs and idioms, which happen to be non-essential for mutual intelligibility. Given the extent of such occurrences in lingua franca contexts, we might well ask whether we should not stop regarding them as 'errors', but this is a pedagogical question which cannot be dealt with by reference to linguistic observations alone, however fascinating they may be.

The point we hope to have made in the present essay is that it is high time that the legitimacy which has already been accorded…