aims at a larger range of languages, though as will become evident later
English is also dominant in AAAL.
A second demarcation problem exists with the large group of researchers
doing experimental work on multilingual processing, such as Albert Costa,
Ton Dijkstra, Annette de Groot, Janet van Hell and Judith Kroll. Though
their work has been influential, and some of it is core AL, they have not
been included in the list because they would not define themselves primarily
as applied linguists. Since the work of this large group of researchers is pri-
marily based in psychology, they would not be able to contribute to this
study when it comes to bringing in a historical perspective. In addition, co-
citation data gathered by Paul Meara (2012, 2014, to be discussed in more
detail in Chapter 9) show that, at least for research on vocabulary, the psy-
cholinguists form a cluster that is clearly separated from the mainstream AL
vocabulary group.
The third demarcation problem is with the researchers doing neuro-
linguistic research, which includes both work on language pathology (bilin-
gual aphasia, bilingual aspects of neuro-degeneration, including aging) and
work on neuro-imaging. Again, this is a world of its own, and certainly there
are researchers like Lorraine Obler and Michel Paradis who have been
influential, but whose work is not essentially AL.
I realize that these demarcations are arbitrary to a certain point and that it
would be easier if researchers in these traditions would state that they do not
consider themselves as applied linguists (as Marianne Celce-Murcia did),
rather than me saying this more or less arbitrarily. Other demarcations are
easier because the definition used is clear in at least one respect: the research
has to do with more than one language. This means thatfirst language
acquisition, conversational analysis, discourse analysis, forensic linguistics,
text linguistics and stylistics, to name just a few topics, will not be discussed
here unless the research includes more than one language.
2.2 Representativeness
Research like this should be based on a survey of the population that covers
all the various subgroups and communities. The sample should in that sense
be representative of the population. As some of the informants pointed out,
that is not the case here. Like the definition of what constitutes AL, the
selection of informants was based on my own network and knowledge of the
field. This means that it tends to be western Europe and North America
biased and to overlap to a considerable extent with the people attending the
AAAL annual conferences. I asked all my informants to give me names of
people I should contact and that certainly helped to make the sample more
representative. Nevertheless, as the lists of informants interviewed and sur-
veyed in Tables 2.1 and 2.2 show, the bias is still there. The overwhelming
majority is from the United States, there is a sizable number of British and
Canadian informants, but small numbers from other countries like Sweden,
The informants 9