It’s well known that the English language is changing all the time. In recent years Cambridge ESOL has followed with interest the vigorous debate about ‘world Englishes’ in different regions; we also make a valuable contribution to this debate through our own research and published articles*. For example, we monitor the approaches of other international providers of English language proficiency tests to compare their policy and practice with our own, and in 2005 we conducted a survey of perceptions, policy and practice among a selection of European language testing agencies representing seven other languages. The results show an interesting diversity of approach across different language testing agencies.
Dealing with issues of language variation
In addressing the issues raised by language variation in language assessment, and especially the challenge posed by ‘world Englishes’, our fundamental principles remain broadly the same as they have always been.
For example, when selecting test input, i.e. the content and linguistic features of reading and listening texts and the tasks designed around them, our guiding principle is the purpose of the test and the focus of what is being assessed.
The linguistic content of the test is as representative as we can make it of the broad context in which English will be used beyond the test. This has implications for the level of variation we consider acceptable or desirable across different modes (spoken and written) or codes (informal and formal) as well as for other types of variation such as regional accent.
Criteria for success
What is the criterion for success? How do we respond to features of a candidate’s written or spoken production (e.g. spelling/pronunciation) which may reflect a regional variety they have learned or grown up with?
We begin from a position of being clear about the focus of the assessment and the degree of precision required. With this in mind, we set standards accordingly. For example, we can be more flexible when evaluating candidates’ speaking ability (since variation is the norm in everyday spoken language); we may have more stringent requirements for their written production (since conformity to a written standard is more common and reasonable).
We make our marking criteria (e.g. in relation to spelling requirements) as transparent as possible not just for our examiners, but also for test candidates. Test takers are often aware of differences across language varieties and sometimes fear they will be penalised for using the ‘wrong’ lexical item or spelling convention. We address this through information provided in the test handbooks as well as by making publicly available the specified assessment criteria against which any task will be evaluated.
Native versus non-native speaker examiners?
There is sometimes a discussion about who is best qualified to make judgments about standards. In language testing this raises the interesting question of whether native (NS) or non-native speaker (NNS) examiners are better qualified to evaluate proficiency and it echoes a similar debate about the relative strengths and weaknesses of NS/NNS English teachers. Of course the reality is that all examiners
– whether NS or NNS – need to be suitably qualified, and to receive initial training and ongoing standardisation as they do at Cambridge ESOL.
Preparing for future change
Over the next 10 or 20 years, the newer ‘Englishes’ that are emerging in some parts of the world may well grow in status and take on a role as teaching and assessment models for English learners. Work currently being carried out at Cambridge ESOL will help us keep up to date with such developments so that they can inform our policy and practice and thus help to ensure that our exams remain useful measures of English language ability.
* Taylor, L (2006) The changing landscape of English: implications for assessment.
ELT Journal, 60, 51–60.
Taylor, L (2006) What to teach? What to test?
English Teaching Professional, 46, 25–27.