Misconstruing The Conflation Of Elements As The Dismemberment Of One Element

Fawcett (2010: 127, 127n):

Indeed, it is not even clear what it would mean to perform such a 'conflation'. For example, does it mean slicing the Rheme up into a number of pieces, and then conflating each of these with each of the Process, the Goal, and the Circumstance, thus: "Rheme/Process + Rheme/Goal + Rheme/Circumstance"? It is hard to imagine what other action could be performed that could conceivably be termed a 'conflation', but when this is done the effect is that the 'Rheme' is broken up into two or more bits, each of which matches the other 'functions' with which it is supposedly being 'conflated'. It would be a dismemberment of the 'Rheme' rather than its 'conflation' with something. 

If this type of dismemberment really is what is intended, we have to ask if any insights follow from it. I can see none. Moreover, even if some insight could be identified, there remains the problem that any such rules would increase in complexity exponentially, once they were expanded beyond the needs of a 'toy' grammar. ⁸ 

8. For example, it is not the case that there is just one 'function' of 'Circumstance'; there are many different types (possibly around forty in all in English), each of which would have to be provided for separately in the grammar, in each of its various possible positions in the clause.

Blogger Comments:

[1] To be clear, if Theme conflates with Actor, then, by default, Rheme ("not Theme") conflates with the following "not Actor", that is, with Process, Scope and Extent. The metafunctional structures are integrated by the syntagm of the rank below: nominal group ^ verbal group ^ nominal group ^ nominal group, as previously explained. It will be seen in later posts that Fawcett argues against the theoretical need of a rank scale, the means of integrating structures.

[2] To be clear, any "pieces" here are not pieces of Rheme, but the three group rank units that realise Rheme: verbal group ^ nominal group ^ nominal group.

[3] To be clear, here Fawcett is failing to find an insight from his own misunderstanding of SFL Theory.

[4] This is misleading, because it is untrue.  The types of circumstance are a matter of systemic selection, not structural configuration. There is no exponential increase in complexity, because if any such conflation rule were actually required, it would be located, as a realisation statement, in the network at a point more general than the sub-system that models the different types of circumstance. In Fawcett's model (Figure 4), on the other hand, where realisation rules are not located in systems, the solution may not be quite so simple and straightforward.

The "Problem" Of How To Conflate Non-Coterminous Elements

Fawcett (2010: 126-7):

The more serious of the two types is that of how to 'conflate' the non-coterminous elements, such as 'Rheme' with 'Process + Goal'. 

Let me now try to state precisely why this is a problem. The theory of generative Systemic Functional Grammar has always, since the mid-1960s, included conflation rules that 'fuse' two or more 'elements' into a single element, e.g., Halliday (1969/81) and Halliday (1970/76b). This is the core of the well-established 'element conflation' model, and we know how to implement computer models of language which incorporate such relationships. But there is no published description of how to map an element onto another element that is not coterminous with it. For example, there is no description of how the supposed 'element' of the 'Rheme' in the thematic structure comes to be mapped onto two or more elements in another structure, such as the Process, Goal and Circumstance in the experiential structure of Figure 7. And the total number of permutations of mappings of all such configurations of elements would be absolutely enormous.

Blogger Comments:

This is misleading, because there is no problem.  Rheme is the element that follows Theme.  For the corrected experiential structure of Figure 7, Actor ^ Process ^ Scope ^ Extent, if Theme is conflated with Actor, then Rheme is conflated with the experiential elements that follow. The structures are integrated by the syntagm that realises them: nominal group ^ verbal group ^ nominal group ^ nominal group.

We	would visit	Mrs Skinner	every Sunday
Theme	Rheme
Actor	Process: material	Scope	Extent
nominal group	verbal group	nominal group	nominal group

To be clear, 'Rheme' is the label for 'not Theme'. It is the labelling of an absence, like 'black' for the absence of light.

Misrepresenting Fawcett's Misunderstandings As Halliday's Problems

Fawcett (2010: 126):

These, then, are two serious, practical problems that arise in the 'structure conflation' model that is presented in IFG. It is interesting that Halliday himself identifies them in his first theoretical-generative paper about the then new model of systemic functional grammar. Yet this is as far as he goes, and there is no hint in his later writings of how the problem of formalising this general proposal is to be solved. Thus, while he identifies the phenomena that cause the problems, he does not ident[i]fy them specifically as problems, and he does not show how he proposes to resolve them.

Blogger Comments:

[1] This is misleading, because it is untrue. As previously demonstrated, Fawcett's notion of 'structure conflation' is a misunderstanding of Halliday's theory, and the problems that Fawcett thinks arise only arise from his misunderstanding of the theory, and viewing it though his own model.

[2] This is misleading.  The reason Halliday does not identify Fawcett's misunderstandings as problems is because they are not problems, but misunderstandings. Similarly, the reason Halliday does not propose how the resolve what Fawcett mistakes as problems, is because they are not problems, but misunderstandings. See previous posts for details of the misunderstandings involved.

On The Non-Problems Created By Fawcett's Mistaken Notion Of "Structure Conflation"

Fawcett (2010: 126):

The next question is that of what the rules in the "structure conflation rules" component in Figure 8 might be like. (Again, there is no guidance on this matter in the writings of those who work in the framework of the Sydney Grammar.) Let us therefore consider for a moment the nature of the task that such a component would be required to perform. 

Many problems would need to be overcome. In fact, there are points in Halliday's writings when he shows that he is aware of some aspects of them. He hints at one problem when he writes in "Options and functions in the English clause" that "not every clause constituent occupies a role in respect of all three [of TRANSITIVITY, MOOD and THEME] — a Modal Adjunct, for example, has no TRANSITIVITY role" (1969/81:143). In other words, the structures that are to be conflated with each other are not necessarily coterminous, in that they may contain gaps at either end (or indeed in the middle). This raises problems for making explicit statements in the putative 'structure conflation rules'. It seems probable that such rules would need to include many conditional operations, if they are to provide for all possible eventualities, since the conflation rules cannot simply refer to the boundaries of the clause. 

Halliday then goes on to make a second point about the nature of such conflations — and it is one that raises horrendous problems for the theoretical-generative strand of the theory. This is the fact that, in the approach to structure that it entails, "a role may extend over more than one element, for example Rheme over Process and Goal" (1969/81:143). Thus the model that is being proposed in the Sydney Grammar does not simply involve the conflation of non-coterminous structures. In addition, many of the elements of which those structures are composed are also non-coterminous. 

Blogger Comments:

[1] To be clear, there is "no guidance on this matter" because there is no need for "structure conflation rules" in Halliday's model, because there is no structure conflation in Halliday's model, as previously explained. Figure 8 is Fawcett's modification of his own model (Figure) based on his own misunderstandings, as previously explained.

[2] This is misleading, because it is untrue. There is no problem here because the metafunctional clause structures are not conflated, but integrated by the realising syntagm of units at the rank below, as in:

we	usually	visit	Mrs Skinner	on Sundays
Actor		Process	Scope	Location
Subject	modal Adjunct	Finite/Predicator	Complement	Adjunct
nominal group	adverbial group	verbal group	nominal group	prepositional phrase

[3] This misleading, because it is untrue. Halliday does not make a second point about the nature of such conflations because there are no structure conflations in his model, and the "horrendous problems" only arise from Fawcett's misunderstandings of Halliday's model.

[4] This is misleading, because it is untrue. There is no problem here because the metafunctional clause structures are not conflated, but integrated by the realising syntagm of units at the rank below, as in:

we	usually	visit	Mrs Skinner	on Sundays
Actor		Process	Scope	Location
Subject	modal Adjunct	Finite/Predicator	Complement	Adjunct
Theme	Rheme
nominal group	adverbial group	verbal group	nominal group	prepositional phrase