Misunderstanding Rank Scale Constituency And 'Accountability At All Ranks'

Fawcett (2010: 309):

Huddleston has rightly described the concept of 'rank' as both "a salient and controversial feature of Halliday's model" (1988:140). Over the last forty years this concept has attracted more criticisms than any other aspect of Systemic Functional Linguistics (SFL) — and additional reasons for abandoning it are given in Sections 11.1 and 11.2 of Chapter 11, and also in Fawcett (2000) and (forthcoming b). (For a summary of these two see Section 4 of this appendix.) 

As we consider the question of whether a grammar should or should not include a 'rank scale of units', we should be clear that if it does so this logically entails having 'total accountability at all ranks' — or, in a weaker version of the same principle, 'accountability at all ranks, with a few clearly justifiable exceptions'. As we have seen, the concept of the 'rank scale' entails a 'consists of' relationship between its 'units', and if in a description a unit on the 'rank scale' can be omitted the concept of 'consists of is lost. Thus 'accountability at all ranks' is simply a more formal statement of the 'consists of relationship.

Blogger Comments:

[1] As demonstrated in the examination of Sections 11.1 and 11.2, none of Fawcett's reasons for abandoning the rank scale withstand close scrutiny, largely because Fawcett confuses formal constituency with form-functions relations. See the posts on this blog dated 1–18 June 2021.

[2] For an examination of this summary, see the upcoming relevant posts on this blog.

[3] To be clear, epistemologically, it is not a matter of whether a grammar should or should not include a rank scale, but a matter of whether a rank scale provides an explanatory advantage over other models of constituency within the architecture of a specific theory.

[4] To be clear, on the one hand, even taken at face value, this is misleading, because it is untrue. A clause consists of words, whether or not it also consists of groups. On the other hand, this is one of the serious misunderstandings that invalidates Fawcett's argument. Fawcett will claim that the limiting case of a group, a single word, is not a group. Halliday & Matthiessen (1999: 178):

Grammatically speaking, participants are realised by nominal groups, which are groups of nouns; and processes are realised by verbal groups, which are groups of verbs. The limiting case of a group is a single word.

[5] To be clear, on the one hand, the 'thus' here is misleading because it introduces a non-sequitur. That is, the omission of a unit on the rank scale does not entail that accountability at all ranks is a formal statement of the 'consists of' relationship.

On the other hand, the non-sequitur is misleading because it untrue. Total accountability at all ranks is not a more formal statement of the 'consists of' relationship. In SFL Theory, total accountability at all ranks is known as the principle of exhaustiveness (Halliday & Matthiessen 2014: 84):

The general principle of exhaustiveness means that everything in the wording has some function at every rank (cf. Halliday, 1961, 1966c).

whereas the 'consists of' relationship is known as the principle of hierarchy (ibid.):

The general principle of hierarchy means that an element of any given rank is constructed out of elements of the rank next below. This is a feature of the constituent hierarchy made up of units and their classes: clause, verbal group, and so on.

That is, exhaustiveness is concerned with form-function relations whereas hierarchy is concerned with formal constituency. It is Fawcett's confusion of these two — see [1] above — that undermines his argument against the rank scale, both in the main text and this Appendix.