Notes 241
Chapter 8
(^1) Coulthard (1985: 134) discusses the diffi culty in describing extended speech
such as a two-minute teacher monologue in terms of exchanges.
(^2) It should be noted that Halliday’s concept of ‘learning to mean’ derived from the
fi eld of fi rst and not second language acquisition.
(^3) Ellis(1994: 643–5) provides a useful summary of how consciousness-raising
exercises can be a valuable classroom practice in helping learners develop explicit
knowledge of grammatical structures prior to being asked to produce the
structures. Similarly, before being asked to produce contextually appropriate
chains, learners could be explicitly instructed in how to realize chains which obey
chaining rules.
Appendix 3
Text 1
(^1) The (#) diacritic indicates throughout a possible and not an actual increment
ending.
(^2) The coding W refers to what Brazil labels an open selector: an item which serves the
purpose of indicating that the making of a particular selection is relevant to the
achievement of target state but that the item itself does not make the selection
which is in fact made later (Brazil 1995: 251). Examples in the corpus are that
(when functioning as a relative pronoun), what, which, why, how, because and who.
(^3) The elements try and tell you etc. have been coded in a manner analogous to try to
tell you as an expansion. An alternate coding would be try and tell you
(^) V c Ø V' N
(^) with the Ø diacritic indicating elided nominal and verbal elements
(^4) Had ellipsis not been coded increments 8, 9, 10 and 11 would all have been
coded as being part of the same increment.
(^5) Strict application of the chaining rules would lead to the coding of the A element
later as suspensive but this seems counterintuitive as the A element does not seem
to be out of place. Consider I will return later where later does not suspend.
(^6) The Ø coding notates an elided NV projecting clause.
(^7) The subchain those engaged in terrorism is suspensive.
(^8) The initial elements of increment 9 are missing. Ellipsis has not been coded as it
seems as if the increment commences in mid thought. An alternative analysis
would be to attempt to reconstruct semantically what is unsaid and code the
ellipsis. Had this procedure being followed the elided elements would appear to
realise a semantic value approximate to the problem/fact/matter etc.
(^9) An alternate analysis would have been to code I mean as N V elements. Had this
been done increment 10 would have formed two increments:
(^) I mean in Algeria for example (#) tens and tens...
(^) N V P N PHR (#) num c num...