In this paper I have reviewed some of the major aspects that typify the language use and
learning trajectories of second language speakers of English and contrasted them with those
of first language speakers. Many of these aspects are well-evidenced in the research
literature, although it has to be said that the literature on second language acquisition is rather
more substantial than that on adult first language speakers.
Taken individually, the research evidence on aspects such as vocabulary learning, grammar,
speaking, listening and reading demonstrates that the government-funded tests, standards and
ESOL curriculum, as they are at present, do not reflect adequately the learning trajectory of
second language speakers. Taken cumulatively, it has been striking to see how many aspects
of learning are affected, primarily the underpinning skills such as vocabulary, grammar,
underpinning skills on the macro skills of speaking, listening, reading and writing is also
marked. The evidence, as it presents itself, is that the standards and exams were designed to
capture the needs of first language speakers (even if with hardly any evidence to
19
speakers, ie that test scores give an accurate reflection of the candidate’s true level of
language skills, really is problematic. This is because the learning development of second
language speakers differs in many respects from that of first language speakers, both in the
processing of language, cognitive development and the skills of speaking, listening, reading
and writing to be learnt.
The question is whether it matters that there is a bias against second language speakers. After
all, these tests have been applied since Key Skills were introduced in 2000. The evidence
presented here suggests that the practice of having the same standards and exams for first and
second language speakers should be reconsidered for multiple reasons. The first is the
principle of fairness. Candidates are entitled to tests which do not discriminate against
individuals or groups of candidates. Employers are entitled to have reliable information on
the candidate’s language skills. The websites of organisations such as the Department for
Business Education and Skills and the Qualifications Curriculum Development Agency show
that they cover aspects such as race, gender and disability equality but that they have not yet
addressed the concept of fairness in relation to language testing. It would, in this context, be
interesting to see how native English speakers (and their parents) would react if they were
asked to take language exams aimed at second language speakers. This does not mean
necessarily that standards and tests are inadequate; but rather that tests designed for one group
should not be applied automatically to another and without regard for unfair consequences.
Secondly, there is the question of reliability. Unfortunately, government departments do not
collect data on achievement by target group, so it is impossible to analyse the results of first
and second language speakers for reliability of outcome. However, as we saw, several aspects
of the national literacy test affect second language speakers’ scores disproportionally. In
addition, there has been substantial and unacceptable variation between awarding bodies in
terms of the complexity of tasks and achievement outcomes. Reliable standards and exams
which match the development of language skills should increase the reliability of tests across
the awarding bodies.
The third criterion is that of effectiveness. We have seen that the subskills approach does not
provide an effective framework to test and teach English to second language speakers. Yet it
dominates the teaching of reading in the classroom. Other skills, such as listening, are taught
as comprehension. This may be suitable for native English speakers who can already
understand English but, as we have seen, this is not the right approach for second language
speakers. The latest OFSTED thematic report on ESOL (2008) reported that, ‘while now
satisfactory, the proportion of provision that is good or outstanding [ ...] remains too low’.
Inspectors still find too many examples of classroom practice that is not effective. This is not
just de-motivating for the learners, it is also not effective use of resources, largely paid for by
the tax payer.
20
So what is needed to develop a coherent framework for the testing of English language skills?
The government is advised to take on board the implications of research on language
acquisition; and to design standards and tests which are fit for purpose. In short they need to
produce standards and tests which:
reflect the learning trajectory of second language speakers
are well-constructed and reliable
promote effective learning
deliver value for the tax payer’s money
give an accurate picture of the skills of the individual and which is meaningful to
employers, schools, colleges and not least the learner him or herself.
Dr Philida Schellekens
August 2011
