The Corpus and its Coding 121
Table 5.2 Tone choices in Texts 1 and 2
Text 1 Text 2
Number of tones Number of tonesReaders \ / – \/ /\ \ / – \/ /\
Bc 55 20 1 14 2 117 25 13 38 3
Bs 62 13 1 15 1 149 38 5 17 4
Dc 58 13 2 13 3 130 34 4 30 3
Dmc 46 9 2 20 2 102 30 4 30 13
Emi 40 18 6 18 2 114 30 9 40 4
Gc 61 9 2 12 2 135 11 6 29 8
Jt 64 11 2 2 2 111 20 5 19 8
Mh 48 10 2 14 1 96 14 8 32 7
Rf 40 6 4 16 1 98 20 10 22 4
Sn 51 9 1 16 3 115 21 5 29 9
Tr 56 6 5 11 0 113 20 23 13 5
Table 5.2 illustrates that there was individual variation in the tones
selected by the eleven readers; the readers selected differently from the
meaning making resource of the tone system in order to construe their
intended readings.^7 Chapter 6 discusses how different tone selections add
communicative value to the target state achieved within and between
increments.
5.3 Coding the Corpus
In order to code the corpus into increments, the orthographic versions
of Blair’s readings of Texts 1 and 2 were coded. These coded versions
were then used as templates and individual adjustments were made to the
codings of the 22 readings to refl ect any differences between the readers’
readings and the printed text. The coding of the orthographic texts
into increments was done without reference to intonation and so all the
increments identifi ed must be understood to be no more than possible
increments dependent on the assumed presence of a tone unit containing
a falling tone. The coding of the orthographic text into possible increments
is presented in Appendix 3.
5.3.1 Identifying increment boundaries in Texts 1 and 2
In order to identify the readers’ actual increment boundaries or the achieve-
ment of target state within an increment two formal conditions must be