Fawcett (2010: 284):
One can, of course, still find within the present theory traces of the phenomena that suggested the original 'rank scale' concept. Firstly, there is the fact that the clause (or "clause complex" when co-ordination occurs) is the only unit that can occur as the highest unit in a tree representation of a text-sentence. Secondly, we can say — for what it is worth — that, for each word that the analyst identifies in a text, there is a fairly good possibility that it will be functioning as an element of a group, and that if one has successfully identified a group there is a fairly good possibility that it is filling an element of a clause. But there is also quite a strong possibility in each case that the supposed generalisation will not hold. (See the fuller discussion in Section 10.2.1 of Chapter 10).
Blogger Comments:
[1] To be clear, as this demonstrates explicitly, Fawcett's approach to formal composition is indeed that of ranked constituent analysis, since it ranks constituents on a scale from sentence to clause to group (and cluster) to item.
[2] To be clear, this is the case whichever rank scale is used: that of SFL Theory or the Cardiff Grammar. SFL Theory identifies these exceptions as instances of rankshift.
[3] To be clear, Section 10.2.1 (pp193-5) is titled The General Concept of Class of Unit. The posts that examine that section are:
No comments:
Post a Comment