30 A Grammar of Spoken English Discourse
In Brazil’s account of Discourse Intonation pitch sequences contract the
same relationships between themselves as tone units do, e.g.
(45) // R and my friend just PUT her FOOT down // P and ↑SPED OFF //
P as FAST as she ↓COULD // P ↓HAPpy to be ↓aLIVE //
There are two pitch sequences in (45) the fi rst of which ends with the word
could. The second pitch sequence, a single tone unit, has initial low-key
signalling that it is equivalent to the fi rst pitch sequence; the speaker
projects an understanding of the state of speaker/hearer convergence
that the friend’s happiness to be alive is equal to the expectations which were
previously generated by the discourse.
All examples discussed in this section have extended tonic segments: tone
units with more than one prominent syllable. Brazil (1997: 14) states that
tone units with only one prominent syllable have minimal tonic segments.
In minimal tonic segments there is no possibility of the independent
selection of key and termination: they are concomitantly selected on the
tonic syllable (ibid. 61). Brazil (ibid. 63) provides example (46):
(46) // he’s ↑LOST //and argues that:
In order to invite adjudication, he/she [a speaker] may attach unnecessary,
but harmless contrastive implications to lost by reason of the concomitant
high-key choice.He argues (ibid. 62 and 63) that the communicative purpose realized by a
mid-key selection is also usually realized by a high-key selection, but the
communicative purpose realized by a high-key selection is not realized by
a mid-key selection. Information that is contrary to expectations is always
additive but information that is additive is not always contrary to expecta-
tions. This suggests that speakers who wish to invite adjudication may
on occasion attach ‘unnecessary, but harmless contrastive implications’ to
their utterances. These contrastive implications are presumably harmless
because they are overridden by the interlocutors’ appreciation of the
existing speaker/hearer state of understanding. Speakers presume that
the implications generated by high key are tolerable in situations where
hearers are aware that they are inviting adjudication.