A Review of A Grammar of Speech 23
Upon completion of the suspension the speaker proceeds from the point
reached in the chain prior to the suspensive element(s). Extensions have
the potential to achieve target state. The intermediate state after an exten-
sion is identical to that which would have been precipitated had the V' been
a V in a simple chain. Production of the fi rst element of an extension
commits the speaker to a second run through of the chaining rules. If an
extension fails to achieve target state, speakers are obliged to produce
further extensions until target state has been achieved.
2.2.4 The coding of lexical elements in chains
Brazil claims that a grammar which aims to describe the reality of observed
used language, does not need to include higher level constituents such as
nominal groups, verbal groups, etc. Instead, he argues that higher-level
constituents are products of constituency analyses which are useful in the
post-hoc analysis of complete texts but not in the descriptive analysis of
speech as a happening. He argues that what he calls ‘the facts of piecemeal
encoding and decoding’ of speech are not to be denied (1987: 147).
He says:
It is important to stress that the real-time presentation of speech we make
central to our account of grammar is an observable and incontrovertible
fact, not a theory. People just do utter one element and then follow it with
another. (p. 229)The expository examples presented to this point, which have described
speech in terms of N, V, V', A and E elements, are in Brazil’s full description
broken down into smaller elements. The following examples illustrate the
full descriptive notation.
Simplifi ed expository description Full description(26) (The little red book) The little red book
(........... N......... .) d e e NThe N element the little red book is decomposed into a string of words
commencing with a determiner (d) followed by two e elements, little and red,
and ends with the N element book. All elements before the fi nal N are
notated in lowercase, analogous to suspensions, because once speakers
produce d or e elements they must produce a following N element. In (27)