Forensic discourse analysis 245and then I said...
and then she said...However, this is just not true. Experiments show that it is impossible for
participants successfully to remember, even immediately afterwards, what
was said to them in a two-party conversation that lasted for as little as five
minutes (Hjelmquist 1984). Participants can typically reproduce only 25–30
per cent of the ideas contained in their interlocutor’s contributions and then
only in a paraphrased form. Their ability to reproduce verbatim can be as
low as 1 per cent.
Secondly, perhaps more surprisingly, speakers do not remember accurately
even their own contributions; verbally transmitted information is not, in the
main, stored in a verbal form in the brain. Thus, when they need to say
again what they have already said, speakers encode the content linguistically
anew, and, as a consequence, there tend to be small but important differences
between the two utterings. These psycholinguistic facts can be of great
significance when we come to examine disputed texts.
One of the functions of the traditional police interview record has been
to provide the Court with unambiguous evidence of what was said—this
left it to the Court to make the interpretative decisions about the meaning
of what was said—a memorable example is the phrase ‘let him have it’ in
the police evidence at the Craig and Bentley trial, where the prosecution
alleged that the phrase meant ‘shoot him’ and the defence ‘give it [the gun]
to him’.
Although interviews used to be, in the vast majority of cases, recorded
contemporaneously in handwritten form, there were occasions when, for a
series of reasons, police officers wrote up a record of an interview afterwards
from memory, in a ‘verbatim’ form; that is they claimed to remember the
exact words spoken and presented their record as direct speech within quotation
marks. Indeed, in one case, where the notebook shows that the ‘notes’ were
written up after the event, a fact never disputed at trial, the sergeant reports
himself as saying, at the beginning of the interview, ‘What we intend to do
now is to interview you by way of a contemporaneous note.’ It would
appear that, in this context at least, ‘contemporaneous’ meant ‘recorded
later as direct speech’ —unless, of course, he was ‘remembering’ what he
would normally have said.
Such after-the-event verbatim records have, apparently, almost always
been accepted by courts as reliable records of what was said. Indeed,
there is no doubt that in most cases such records were made in good
faith. The problem facing the discourse analyst is to demolish the
common man’s assumption and to demonstrate that verbatim recollection
is impossible.
One obvious way of doing so would be to ask the policemen involved
to repeat their feat under controlled conditions but, so far as I know, this
has not yet been allowed by any court. So, in such situations there are two