nationalism’, and presenting a whole range of new communicative dynamics
and processes”. Similar views are expressed by Durk Gorter, who sees as a
major trend:“The internet and technology and its consequences for inter-
connectedness, for using (and learning) languages, for publications, for
research, or, in general, for how we do our work and live our life.”
Another aspect of technology is the emergence of social media and their
impact on communication and linguistic variation. This is a fairly recent
development and the average age of the informants is such that they are not
typically the forerunners in this respect, but the potential impact is generally
seen as relevant.
There has been an enormous growth of what is generally labeled as Com-
puter Assisted Language Learning (CALL). Though some people such as
Joan Kelly Hall argue that the same type of meaningless exercises continue to
be used, but now with a computer rather than a teacher, others feel that the
new technologies have led to better teaching, e.g. through the role of corpora
and online resources.
6.2.7 Ethics in testing
A number of informants talk about the issue of ethics in testing. Tim
McNamara remembers that when Elana Shohamyfirst raised, in a provoca-
tive way, the political character of language testing (already raised by Bernard
Spolsky, but more moderately), he atfirst rejected her argument and felt
wounded by her attack, but now he realizes that the liberal approach to
ethics in language testing as exemplified by the work of Alan Davies and
Anthony Kunnan is inadequate.
6.2.8 Meta-analyses and overview studies
Alan Juffs comments on the role of meta-analyses, which he sees as an
important new development since they could lead to conclusive results for
specific controversial issues. The most (in)famous meta-analysis in ourfield
is probably the Baker and De Kanter study from 1983 on the effectiveness of
bilingual education. Their conclusion is that the studies available do not
show a positive effect for bilingual education. Baker and De Kanter have
been criticized on various aspects of the study, including the random
assignment to conditions and the selection of studies in the meta-analysis.
Willig (1985) re-analyzed the dataset of the Baker and De Kanter study and
found a positive effect of the use of the children’s mother tongue in educa-
tion. This shows the complexity of meta-analyses: they may create the kind
of problems they intend to solve. Even for meta-analyses, questions of
selection of studies and their interpretation arise. The goal of most meta-
analyses is to end a long discussion on specific problems and to integrate all
research that meets certain criteria and adds to the knowledge about the
problem. In thefield of AL, the best-known example of a meta-analysis is
68 Main trends I