Misrepresenting Halliday On Recognising Elements Of Structure

 Fawcett (2010: 278):

A modern theory of SF syntax is — or should be — an explicitly functional theory of language, so that the criteria for recognising an element of structure are — or should be — functional and semantic rather than formal and positional. 

Thus the elements of a unit are those that are required to realise the meanings that have been selected in the system networks for realisation in this unit — ultimately, of course, as items (see below).

Halliday has surprisingly little to say in "Categories" (or indeed in any later writings) about the criteria for recognising elements of structure (especially the elements of groups).

Blogger Comments:

[1] To be clear, given that this is Fawcett's belief, it is surprising that he has, nowhere in this publication, provided "functional and semantic" criteria for recognising his elements of structure, and has, instead, foregrounded form (classes of unit) and position (place).

[2] To be clear, in SFL Theory, functional elements of grammatical structure (syntagmatic axis) are specified by realisation rules (such as 'insert Subject') in the grammatical system networks (paradigmatic axis) that the grammatical structures realise. In the absence of grammatical metaphor, grammar (wording) and semantics (meaning) are in agreement (congruent).

[3] This is misleading, because it is untrue. Halliday ± Matthiessen (1985, 1994, 2004, 2014) devote three chapters to providing criteria for recognising elements of clause structure, one chapter for each metafunction, and one chapter to providing criteria for recognising elements of group and phrase structure. For example, Halliday & Matthiessen (2014: 240, 249) provide the recognition criteria for Scope and Senser as follows:

…the Scope of a ‘material’ clause is not in any way affected by the performance of the process. Rather it either (i) construes the domain over which the process takes place … or (ii) construes the process itself, either in general or specific terms…

In a clause of ‘mental’ process, there is always one participant who is human; this is the Senser: the one that ‘senses’ – feels, thinks, wants or perceives… . More accurately, we should say human-like; the significant feature of the Senser is that of being ‘endowed with consciousness’. Expressed in grammatical terms, the participant that is engaged in the mental process is one that is referred to pronominally as he or she, not as it.

It is the fact that, in a functional theory, such criteria are 'from above' — rather than 'from below' — that may explain why Fawcett is unable to recognise them as criteria.