12 Forensic discourse analysis
Malcolm Coulthard
INTRODUCTION
Courts are increasingly calling on linguists to help in certain types of case
and one can already see the beginnings of a new discipline, forensic linguistics—
1991 saw the third British conference of Forensic Phoneticians and 1992
the first British conference of Forensic Linguists.
Until recently most of the forensic work has been in the area of substance,
i.e. comparisons of samples of handwriting (Davis 1986), and of tape-recorded
voices (Nolan 1983; Baldwin and French 1990), where the methodology is
already well developed. By contrast, forensic discourse analysis is a very
new area and the methodology is still being developed ad casum. This
chapter, which uses data from real cases,^1 is a contribution towards this new
discipline.
THE FACTS
Forensic discourse analysis is, in the main, concerned with two kinds of
text: handwritten contemporaneous records made by police officers of interviews
with witnesses and suspects, and statements dictated by witnesses and suspects
to police officers.
An interview is normally conducted by one, usually the more senior,
police officer and transcribed by a second. It proceeds very slowly, usually
at 20–5 words a minute, a pace which is basically governed by the officer’s
writing speed. Some time after the interview both police officers prepare a
typed version of the interview, based on the contemporaneous ‘notes’; these
versions are identical, except for the reporting clauses: ‘I then asked Power
“what did you do next?”’ as compared with ‘Sergeant Jones then asked
Power “what did you do next?”’
The text is supposed to be a complete record of what was said during the
interview and thus it includes the caution:
You are not obliged to say anything unless you wish to do so, but what
you say will be put into writing and given in evidence. Do you understand?