120 A Grammar of Spoken English Discourse
was to notate the tone movements off the tonic syllables and once this was
completed the key and termination selections were notated. Thus, each
individual recording was carefully listened to on fi ve separate occasions
before an auditory intonation transcript was produced. Once the auditory
transcriptions had been completed there were 22 transcriptions represent-
ing each reader’s reading of Texts 1 and 2.
Brazil, when doing his own intonation transcriptions, relied solely on the
auditory method of transcription which, however, is as Wichmann (2000: 2)
points out, impressionistic. Pickering, Williams and Knowles (1996), in an
analysis of transcriber differences in the compiling of the SEC corpus,
illustrate the subjectiveness of auditory transcriptions. They report that
two highly experienced transcribers differed in their marking of complex,
simple and level tones with a 33% level of disagreement between the tran-
scribers in how they marked level tone (ibid. 79). Pickering et al. speculate
that the differences between the transcriptions resulted from the fact that:
... the transcribers make use of different thresholds in determining
whether or not a change in fundamental frequency between syllables is
signifi cant. (ibid. 83)
In order to reduce the impressionistic element from the intonation
analysis, all 22 recordings were analysed instrumentally using PRAAT
version 5.1.^6 This resulted in some changes to the transcriptions: namely
whether a syllable was prominent, tonic syllable placements, tone move-
ments and the notation of syllables as high or low key or as high or low
termination selections.
Brazil (1997: 6) identifi es tone unit boundaries through the presence or
absence of a pause. Where there is no tonic syllable present before the
pause he labels the tone unit as an incomplete one. This book, however, fol-
lows numerous other scholars such as Crystal and Davy (1975) and Halliday
and Greaves (2008) in recognizing a difference between hesitation pauses
and junctural pauses. Hence a tone unit boundary was marked only if there
was a preceding tonic syllable e.g. from Bc’s reading of Text 1.
(1) that the MEEting should... con\TINue in my absence //Bc paused for 0.211 of a second between the words should and continue. As
there is no tonic syllable within the tone unit prior to the pause it is notated
by the three dots as a tone unit internal pause. Table 5.2 summarizes the
readers’ tone choices in Texts 1 and 2.