The Shape of Tests to come

This paper looks at the future of English language testing, in the light of the massive growth of English as a lingua franca. Today most interaction in English in the world is probably between non native speakers; in Europe, as elsewhere, English is the default language of choice between speakers who do not share the same mother tongue. Language tests, however, continue to refer exclusively to native speaker (NS) norms, and tests of spoken interaction continue to be premised on a model of native – non native communication. This is particularly evident in internationally known exams, in which native speakers assess the performances of non native speakers.\ud \ud In this paper I make the case that (in some contexts at least) the time seems to have come to abandon native speaker standards in order to provide valid and meaningful assessments of the use of English in an international ambit, in which the language and strategies of native speakers may actually hinder communication. Such assessments are likely to draw on the legacy of communicative language testing, and be defined in terms of a successful communicative outcome. \ud \ud However, the abandonment of a standard version of the language, for the fluid, context-dependent, norms of ELF, is problematic. What sort of language competences can be inferred from a ‘one-off’ performance which is in the nature of ELF interaction? And what are the implications of this for the language teacher wishing to prepare students to negotiate meaning in a lingua franca context, and to assess their students’ abilities to do so?\ud \ud I argue that rather than a ‘test of ELF’, teachers and testers (and ultimately international examining boards) will need to develop ‘ELF aware tests’ in which, for example, in a receptive skills test students are exposed to texts produced by NNSs, or in a test of spoken interaction they are assessed for strategies (such as lexical creativity and accommodation) which may be quite distant from the linguistic behavior of native speakers.\ud \ud In such a scenario the non native teacher/tester may well be in a better position than native speakers to be able to make predictions about a speaker’s performances in ELF contexts, in keeping with House’s (2003) suggestion that competence in ELF should be assessed by ‘an expert ELF user ... a stable multilingual speaker’. More than a decade later, with the seemingly unstoppable development of English as the preferred means of international communication, and as our understanding of ELF grows from research findings, the challenge for teachers and testers to engage with the phenomenon is there to be taken up