DeKeyser, input is seen as more important now than in the past. Elizabeth
Lanza concurs:
In thefield of bilingualfirst language acquisition, psycholinguists have in
recent years acknowledged the need to take into account the impact of
input on bilingual language acquisition. This was previously a neglected
aspect of study. However, what is currently in vogue is not a conception
of input as interactional, that is, qualitative; rather, a quantifiable notion
of input, for example, hours exposed to a language.The next step was the development of the interaction hypothesis, which
was presentedfirst in Long’s 1996 article on the role of the linguistic envi-
ronment in second language acquisition. While this hypothesis was popular
for some time in the 1990s, few informants mention it as a major trend.
7.1.2 Transfer and cross-linguistic influence
From its beginnings, applied linguists have looked at the impact of thefirst
language on the language to be learned. Work by Terence Odlin and Eric
Kellerman has continued the tradition of contrastive analysis, though in
updated form. Kellerman’s“psychotypology”concept has changed views on
transfer as a purely mechanical copying of rules and elements from thefirst
language into the second. Robert DeKeyser notices still some remnants of
the old contrastive analysis tradition, but also less of an obsession with
errors:“Language learning is more than avoiding errors.”
A typical European development has been the study of learner varieties.
Wolfgang Klein and Clive Perdue were most active in this research, which
has led to a rethinking of what constitutes learner language. In a sense it is a
move away from the interlanguage concept as developed by Selinker,
because it is not focused on the development in the direction of a standard
native speaker but has its own dynamic. The basic variety emerged from
interaction between non-natives of a particular language, similar to pidgin
and creole languages. In this context, Chris Candlin observes that the Klein
and Perdue work never went well with British applied linguists, who look
more at the social side than the psycholinguistic one.
Though not specific for research on input and interaction, Martin Bygate
sees a tendency that
people seem too ready to move on to something new, rather than stick
with a particular problem and use empirical evidence in order to test
and if necessary refine the formulation. For example, people working on the
interaction hypothesis invariably produce results demonstrating that
the input hypothesis is correct. Rare are those who bother to examine
the limits of the hypothesis–whether in terms of the minimum levels of
interaction needed for acquisition to take place, nor the maximum that74 Trends II