David Rose On Focus Structures

David Rose wrote to sysfling on 14 Apr 2024, at 22:17:

The argument in IFG 6.2.5 is elegantly designed to explain nom gps with two Heads, ‘a phenomenon that from one point of view appears as a single entity and from another point of view as two’.

It begins by positing a dual experiential/logical structure at group rank, in order to then posit a post-modifier function in the logical structure, in which the experiential Thing is embedded in the Complement of a downranked prep phrase. This model is then applied to a series of nom gp structures, culminating with this one.

What if we were to argue instead from this example backwards? As facet is just one of the semantic options for this type of structure (richly captured in IFG Table 6-11), we could generalise its function name, e.g. as Focus. While the Deictic and Thing in this nom gp are realised by words, the Focus is realised by a downranked nom gp, that is linked to the hills by the structure marker ‘of’, whose function we could label LK for Linker.

As the Thing in the Focus nom gp is downranked, do we need an additional logical structure in which the primary Thing is downranked?

Working backwards, similar structures are analogous with this one, such as those classified as Classifier/Head in IFG.

Or Epithet/Head

This last one is ‘a more restricted type’ which seems to have evolved as an idiomatic realisation of attitude, on the model of the more common Focus structures.

Blogger Comments:

[1] This is misleading because it is not true. To be clear, a nominal group has only one logical Head. Rose's misunderstanding is resolved by the part of the quote he omitted. Halliday & Matthiessen (2014: 396): 

In all such nominal expressions where Head and Thing are not conflated although both are clearly present, what is being construed is a phenomenon that from one point of view appears as a single entity and from another point of view as two.

[2] This is misleading because it is not true. To be clear, the experiential Thing is not embedded. It is the nominal group serving as Facet that is embedded in the experiential analysis. The Thing in the experiential structure conflates with the Head of an embedded nominal group in the logical structure.

[3] To be clear, this merely describes the general function of the Extended Numerative in Halliday & Matthiessen (2014: 395), with Matthiessen's 'Extended Numerative' rebranded as Martin's 'Focus'.

[4] To be clear, this creates a theoretical inconsistency, since 'linker' is a class of conjunction that marks a paratactic relation, whereas of is not a conjunction and does not mark a paratactic relation.

[5] To be clear, there is no Thing function in the logical structure of a nominal group. Moreover, the logical structure is necessary in such instances to show that the Head and Thing are dissociated.

[6] To be clear, these analyses fail to (i) recognise the dissociation of Head and Thing, (ii) distinguish between Classifier and Epithet, and (iii) distinguish these structures from Extended Numeratives. As such, they have reduced explanatory power, and can be discarded on that basis.

[7] This is a bare assertion since it is unsupported by evidence: the logical fallacy known as ipse dixit.