240 Notes
of spontaneous dialogues regularly transcribe tone units with more than two
accented syllables. Crystal and Davy, like this book, record prominent syllables
other than the onset and tonic which are stepped up in pitch. Even if one accepts
that Brazil’s notion of prominence is somehow different from the notion of
accenting Brazil’s argument does not appear to describe what is occurring in
the corpus. In three of the four cases the speaker chose to make prominent a
closed-class lexical item and pitch it signifi cantly higher than the previous onset
syllable.- // REALize that ↑OUR determi\NAtion // [T1-Bs-19]
- // Even in our ↑OWN MUslim co\MUNities // [T2-Bc-57]
- // to USE ↑ANy means at \↑ALL // [T2-Tr-45]
(^) The selection of prominence on items such as our, own and any does not seem to
be an automatic process. It appears that by making these items prominent
the readers are projecting contexts where our, own and any realize existential
selections from closed lexical paradigmatic sets. The co-selection of a high pitch
level particularizes these selections. For example, Bs projects a binary opposition
where our opposes all other relevant lexical senses such as his, her, my, your and
their and instantiates the meaning our and not anybody else’s.
(^12) There were 125 examples of high key/termination in the corpus which have not
been included in this fi gure.
(^13) The example was originally taken from Crystal and Davy (1975: 36) with the
grammar coding added by O’Grady.
(^14) The hearer is of course very unlikely to fi nd the news that Muslims are free to
worship in America to be either new information or surprising. For the purposes
of his own rhetorical purposes and effect Bs has manipulated the context by
projecting that the information is both newsworthy and likely to be surprising.
(^15) High termination is described by Brazil (1997) as anticipating adjudication or
more loosely here as seeking an active hearer intervention. In the data studied
here active hearer intervention was precluded. The only apparent method of
investigating whether a high termination value is present appears to be to ask the
speakers whether their use of high pitch on a tonic syllable was in fact intended
to invite an active hearer intervention. Accordingly the high termination value is
assumed to be the default.
(^16) In Text 1 there were also 22 (56.4 per cent) increment fi nal low key/terminations
which were immediately followed by a high key and one (2.6 per cent) low key/
termination immediately followed by a high key/termination. In Text 2 there
were 15 (25 per cent) low key/terminations immediately followed by a high key
and 5 (8.3 per cent) low key/terminations followed by a high key/termination.
The communicative value of low key/termination is discussed in Section 7.3.1.
(^17) Falling tone is, as previously discussed, a necessary but not suffi cient condition in
indentifying increments. For a stretch of speech to form an increment it must
also satisfy the grammatical chaining rules and in the context in which it was
produced realize an act of telling.