Confusing Composition With Realisation

Martin & Rose (2007: 96):

Nuclear relations below the clause
Below the clause, processes, participants and circumstances are themselves made up of groups of words, including lexical items. In Halliday’s 1994/2004 model, clause, group and word are different ranks in the grammar; a clause is realised by a configuration of word groups, each of which is realised by a configuration of words. As with the clause, nuclear relations also pertain between lexical words in groups. To describe these relations, we need to distinguish two kinds of word groups — nominal groups that realise things and people, and verbal groups that realise processes.

Blogger Comments:

[1] Here Martin & Rose confuse composition (the rank scale) with realisation (the relation between levels of symbolic abstraction).  In terms of composition, the rank unit, clause, is made up of units of the lower rank, groups and phrases.  In terms of realisation, clause functions, such as Process, are realised by forms, in this case, the verbal group.

In terms of expansion relations, composition is a type of extension, whereas realisation is a type of elaboration.  In terms of symbolic abstraction, the composition of the rank scale is of one level of abstraction, form, whereas realisation relates two distinct levels of abstraction, in this case, function and form.

[2] Here, as throughout this chapter, Martin & Rose confuse two distinct notions of 'word': 'word' as grammatical rank unit and 'word' as lexical item.  Halliday & Matthiessen (1999: 568):

The folk notion of the “word” is really a conflation of two different abstractions, one lexical and one grammatical.

[3] Here again Martin & Rose confuse composition (the rank scale) with realisation (the relation between levels of symbolic abstraction).  In SFL theory, a clause is composed of groups (± phrases), each of which is composed of words.  Moreover, each is composed of syntagms of lower rank forms, not configurations, since it is functions, not forms, that are configured.

[4] To be clear, Martin & Rose have not identified what it is that is scaled from nuclear to peripheral in groups, nor the basis on which it is nuclear or peripheral.  Without a clear statement of the underlying principles involved, this is merely an empty exercise in relabelling.

[5] It will be seen, in the discussion of verbal groups, that Martin & Rose mistake elements of clause structure, circumstantial Adjuncts, for elements of verbal group structure.