A very substantial body of work on classroom interation has been
published since the late 1960s. There is no room to review this work here,
and it is in any case fairly well known. The work varies according to how it
draws on linguistic, sociological, anthropological and psychological methods,
but broadly speaking there have been three influential kinds of study of
classroom language. Type 1 could be called i~sightful observation. This
involves detailed study and commentary on recorded lessons. It is valuable
in that it demands close attention to be paid to the details of real language,
but is inevitably limited since it is restricted to impressionistic and selective
commentary. The best known British work is probably that of Barnes et al.
( 1969). There is no doubt that many teachers find Barnes' work very helpful:
it has made them aware of all kinds of things they had never noticed before,
and Barnes is a very sensitive observer. This is precisely one of the problems;
there is no method or guiding principle for those of us who are not as
sensitive and full of insight as Barnes. Such work can be made more
principled and theoretically secure by using fieldwork methods developed in sociology and anthropology, and by relating the observational data to an
explicit theoretical framework such as symbolic interactionism. Type 2
involves the use of coding schemes: that is, sets of categories designed to
code or classify large amounts oflanguage, usually as it happens in real time.
This may be valuable in allowing broad trends to become visible and in
making gross comparisons between different teachers, different school
subjects, even groups of teachers in different countries, and so on.
However, it inevitably means that close attention is no longer paid to the
actual language used. This approach derives from work done in the 1950s by
the American social psychologist Robert Bales. The best known work on
classroom language is by Flanders (1970). Such coding schemes are often
frequently used in micro-teaching when this is used as a teacher-training
technique. Type 3 could simply be called discourse analysis. The aim here is
to describe spoken discourse as a linguistic system in its own right: to
discover what the units of analysis are, and how these units relate into
sequences.