Fawcett's Different "Ground Rules"

Fawcett (2010: 9):

It is important to emphasise that the 'ground rules' that guide my work on syntax differ from Halliday's in one important way. This is that the aim is to show only the minimal necessary structure at the level of form, and to provide for the explicit representation of meanings — and so also the representation of the broad types of meaning corresponding to Halliday's 'metafunctions' — at a second level of representation, i.e., at the 'systemic-semantic' level of representation.

Blogger Comments:

These different "ground rules", in fact, represent the proposal for a significantly different architecture for SFL theory.  Where SFL involves system–structure relations (so-called "cycles") on both strata of the content plane, semantics and lexicogrammar, Fawcett's proposal is to have just one system–structure relation for the content plane, with system (paradigmatic axis) as the semantic stratum, and structure (syntagmatic axis) as the lexicogrammatical stratum.

That is, the Cardiff model of grammatical structure depends on a different theoretical architecture. This raises the stakes considerably, since adopting the Cardiff model of syntax, as a better alternative, entails adopting the Cardiff theoretical architecture, as a better alternative, as well as all its ramifications for the rest of the theory.

As already noted, the most immediate disadvantage of this architecture is its inability to model grammatical metaphor systematically — as a junctional construct involving the meanings of both the incongruent and congruent grammatical realisations.

If this significantly different model is to replace 'the standard model', it must be better, and it must be shown to be so.

Misjudging Halliday And Misrepresenting The Orientation Of SFL

Fawcett (2010: 8-9):

What we need, clearly, is a recent statement by Halliday in which he summarises his current theory of language, in the way that "Categories" did for Scale and Category Grammar. Fortunately, his contribution on "Systemic Theory" to the Encyclopaedia of Languages and Linguistics (Asher 1993) goes a long way to providing this, and it can be usefully supplemented by his "On grammar and grammatics" (1996). However, the orientation of "Systemic theory" is 'theoretical-generative' rather than 'text-descriptive' … and the perhaps surprising result is that it has rather little to say about the syntactic structure of texts. We shall therefore also need to make use of Halliday's major recent descriptive work, An Introduction to Functional Grammar, and this provides, as we shall see, a significantly different picture of language. I shall also draw occasionally on Matthiessen (1995), a work that complements IFG invaluably by providing the system networks that are largely missing from that work, and which also sometimes provide a hint of an interestingly different perspective on the Sydney Grammar. 

However, for reasons which will be explained at the relevant points, the fact is that Halliday has nowhere made a comprehensive statement as to the nature of syntax in a modern SF grammar that is comparable in its scope with that in "Categories" — and nor has Matthiessen or any other exponent of the Sydney Grammar. One would expect that the enormous changes made to the model as it was developed from the Scale and Category Grammar of the 1960s into the Systemic Functional Grammar of the 1990s would have led to changes in the representation of structure at the level of form. And indeed they have, as Chapter 7 will clearly demonstrate. But Halliday has provided only the most general of justifications for the immense changes that he has made in the way that formal structure is represented in his model (e.g., in Chapters 1 and 2 of IFG).  The only reasonably full statement by a systemic functional grammarian whose purpose is to reflect the major changes in the theory referred to above has been that of Fawcett (1974-6/81) — this being probably best known through the summary provided in Butler (1985:94-102).

Blogger Comments:

[1] Here Fawcett provides four negative judgements criticising Halliday's behaviour, juxtaposed with a positive judgement admiring his own behaviour:

Token	Appraised	Appraisal
		Polarity	Attitude
What we need, clearly, is a recent statement by Halliday in which he summarises his current theory of language	Halliday	negative	judgement: normality
and the perhaps surprising result is that it has rather little to say about the syntactic structure of texts	Halliday	negative	judgement: normality
the fact is that Halliday has nowhere made a comprehensive statement as to the nature of syntax in a modern SF grammar	Halliday	negative	judgement: normality
But Halliday has provided only the most general of justifications for the immense changes that he has made in the way that formal structure is represented in his model	Halliday	negative	judgement: normality
The only reasonably full statement by a systemic functional grammarian whose purpose is to reflect the major changes in the theory referred to above has been that of Fawcett (1974-6/81)	Fawcett	positive	judgement: normality

However, since the criticisms are entirely unjustified, this is very misleading. As previously pointed out, Halliday has

• been explicit about the place of formal syntagms in systemic functional grammar, 
• explained why a functional grammar views grammar 'from above', and
• explained why the metafunctions are included in the model of grammatical structure.

[2] This misrepresents the orientation of SFL.  Halliday & Matthiessen (2014: 56, 57):

Systemic functional theory also differs from many other functional theories in its emphasis on comprehensive, text-based descriptions — descriptions that can be used in text analysis; …

The description of English grammar presented here is not designed as a reference grammar. However, unlike the recent reference grammars — or all previous ones for that matter, this description has been designed as one that can be used in text analysis — a task that imposes quite stringent demands on the description.

Misrepresenting Theoretical Differences Within SFL

Fawcett (2010: 6):

Within the broad family of systemic functional theories of language, there are what we may term (1) the "Sydney Grammar" (with two 'sub-dialects' associated with Hasan and Martin concerning differences in their models relating to the higher levels of '(discourse) semantics', 'register' and 'genre'), (2) the "Cardiff Grammar", (3) the "Nottingham Grammar", (4) the "Leuven Grammar", and perhaps others. Halliday has made the interesting suggestion that we should think of these alternative versions of SFL as being related to each other in the way that the dialects and registers of a language are.⁷

7. It was Michael Halliday who first suggested the metaphor of 'the Cardiff dialect', 'the Nottingham dialect' etc, during the International Systemic Functional Congress held in Beijing in 1995. However, he has also suggested the metaphor of 'register variation' — originally for thinking about the differences between Martin's and Hasan's different approaches to genre and register. He calls the difference between those two models a "kind of variation in 'metaregister'", saying that this is "one of the ways in which systemic theory appears as a metaphor for language itself' (Halliday 1993:4507). In some ways the concept of 'register variation' provides a more insightful metaphor than that of 'dialectal variation'.

Blogger Comments:

[1] Presenting the Cardiff Grammar as a "dialect" or "register" of SFL theory is inconsistent with Fawcett's already stated intentions.  Fawcett (2010: xviii, xxi) presents his "dialect" or "register" as a replacement for Halliday's "dialect" or "register":

This book makes clear proposals for a (partly) new theory of syntax, and in particular for the replacement of the method of representing structure that is used in Halliday's Introduction to Functional Grammar (1994) by a simpler method.  … 

In what I have said so far, I have been writing as if the theory of syntax to be presented here is an alternative to Halliday's approach to structure. And this is indeed what it is, in that the method of representing the syntax of a text-sentence to be described here is ultimately an alternative to his 'multiple structure' method rather than a complement to it.

[2] This misrepresents the difference between Hasan's and Martin's models.  Hasan's work is largely self-consistent and consistent with the architecture of SFL theory, whereas Martin's work is neither self-consistent nor consistent with the architecture of SFL theory — as demonstrated at length and in detail here.  The notion of discourse semantics, register and genre as 'higher levels' is Martin's theoretical misunderstanding alone.

Promising An Argument Against Metafunctional Clause Structures

Fawcett (2010: 4):

We shall find that there are a number of theoretical problems with IFG which need to be addressed if we are to develop an adequate theory of syntax for a modern SF grammar. Part 1 of the present book therefore also functions as a friendly critique of that work from within a framework of shared basic assumptions. However, the argument that I shall present here concludes with a demonstration that the representations in IFG cannot serve as the 'final' representation at the level of form, and that this fact requires us to reconsider the theoretical status of the 'multiple structure' representations in IFG itself — and so in the many derived works.

Blogger Comments:

[1] This again repeats the still undemonstrated claim that the metafunctional clause structures of SFL theory are 'intermediate', and need to be integrated into a single formal syntactic structure, which is thus another instance of the logical fallacy known as proof by (repeated) assertion.  The reason for noting all these repetitions is the interpersonal function they serve in the discourse, which will become clear when the promised "demonstration" finally appears.

[2] It is the lack of 'shared basic assumptions' that makes Fawcett's work inconsistent with SFL theory, as already demonstrated for the notions of structure and syntax.

[3] To be clear, the metafunctions are, perhaps, the major innovation to Halliday's theorising since first setting out his 'categories for a theory of grammar' in 1961, and they are integral to the theory.  Halliday & Matthiessen (2014: 84):

… these three kinds of meaning run throughout the whole of language, and in a fundamental respect they determine the way language has evolved.  They are referred to in systemic accounts of grammar as metafunctions, and the concept of ‘metafunction’ is one of the basic concepts around which the theory is constructed.

A major advantage of theorising metafunctional structures on the stratum of lexicogrammar is that it enables the systematic examination of grammatical metaphor.