Chapter 6
Increments and Tone
Brazil (1995), as we have seen, claims that only the presence of end-falling
tone projects that the speaker has altered the pre-existing state of conver-
gence between the speaker and the hearer. End-rising tone regardless
of where it occurs in the increment signals that the tone unit it is contained
in does not alter the pre-existing state of speaker/hearer convergence.
Increments unfold in a linear manner with target state only achieved after
the production of the fi nal elements in the increment. A potential incre-
ment must contain at least one falling tone and in the unmarked cases
we would expect that the fi nal tone unit in an increment would contain a
falling tone. Crystal (1975: 34) claims that around 80 per cent of tones
are neutral and that it is only what he labels unpredictable occurrences
of tone: tones which are out of their expected place in an utterance which
merit attention. If we follow this line of argument we can see that incre-
ment fi nal end-rising tone is marked and prima facie appears to be doing
more than signalling a non-telling. Table 6.1 illustrates that around 73%
of increments in Text 1 and 78% of increments in Text 2 contained an
end-falling tone.
While it is clear that increments which have end-falling tone are unmarked it
is also clear that a signifi cant subgroup of increments have a fi nal non-falling
tone. Of the increment fi nal non-end-falling tones the majority are fall-rises
followed by rises – see Table 6.2 for the actual numbers. Unexpectedly three
instances of level tone were located in increment fi nal position.
Table 6.1 Tone in increment fi nal position*
End-falling tone Non-end-falling tone
Text 1 160 (73.1%) 59 (26.9%)
Text 2 557 (78.3%) 153 (21.7%)
* The two incomplete increments presented as example (11)
in Chapter 5 have not been included in the count.