Misrepresenting Halliday On The Development Of Systemic Theory

Fawcett (2010: 45-6):

I have already suggested that the first three changes to the "Categories" model introduced by Halliday in the late 1960s and early 1970s were revolutionary. Interestingly, however, Halliday himself writes about these momentous developments in the theory as if they were, in large measure, simply additions to it — rather than changes that might involve re-assessing the existing concepts. Thus he writes (1993:4507) that "systemic work [...] has tended to expand by moving into new spheres of activity, rather than by re-working earlier positions". The difference between expanding a theory and changing it is an important one. The term "expand" typically implies additions rather than alterations, so that the "expansion" of a theory does not necessarily require one to rethink the concepts of the earlier version. But any changes to the existing concepts in a theory should be followed by a thorough check to discover whether they lead to the need for any further changes. In a theory of language, as in language itself, tout se tient (Meillet 1937). It is certainly true that the theory has expanded greatly, in the sense that it now covers many additional aspects of language and additional languages, and that is has been used in additional areas of application. But many of the innovations — including the three to be summarised here — have had an effect that is ultimately revolutionary. And such changes do indeed demand the "re-working [of] earlier positions". It is a nice irony that Halliday should have written the words cited above in his 1993 paper "Systemic theory", because it is there that he spells out most clearly the revolutionary effect of the changes from "Categories" — as we shall see in due course. (Perhaps this is part of the general phenomenon that it is often easier for others to see the significance of a new idea than it is for the innovator.)

Blogger Comments:

[1] This is misleading.  Halliday (1993) is termed 'Systemic Theory', and the quote that Fawcett cites is concerned with theoretical developments since the inception of Systemic Theory — as the words 'systemic work' makes clear.  Halliday acknowledged significant difference between his earlier model, 'Scale–and–Category Grammar', and his later model, 'Systemic Grammar', by the change of name.

[2] Halliday uses 'expand' as a technical term that subsumes three subtypes:

• elaboration (exposition vs exemplification vs clarification)
• extension (addition vs variation vs alternation)
• enhancement (temporal, spatial, manner, causal, conditional).

[3] Here Fawcett identifies precisely what his alterations of Halliday's theory demand, and which he himself has not done, while implying that Halliday has failed in this regard.

[4] These "revolutionary innovations" were introduced at the beginning of Systemic Theory, not in the course of its development; see [1].

[5] The "earlier positions" that Halliday "reworked" are those of 'Scale–and–Category Grammar', and the outcome of that reworking is the new theory 'Systemic Grammar'; see [1].

[6] This might have alerted a more careful reader that he had misunderstood the quote from Halliday (1993).

[7] Halliday did see the significance, and changed the name of his theory to reflect this, and outlined the significant changes in his retrospective (Halliday 1993) that Fawcett quotes here.

Misrepresenting Halliday On Stratification

Fawcett (2010: 45):

Then in the second half of the chapter I shall describe how, in the 1970s, Halliday tentatively explored two contrasting approaches to meaning — one of which adds a second level of meaning — and how in the 1990s he finally decided in favour of what we may call the 'two-level model of meaning'. Although I shall not present here the full set of arguments against his decision (which deserve a paper or even a book of their own) I shall show why, even if you accept Halliday's position, it does not seriously affect my claim that the model of language presented in Figure 4 of Chapter 3 is common to all systemic functional grammars. 

Blogger Comments:

[1] This is misleading on two counts.  On the one hand, Halliday has never proposed two levels of meaning in any of his models, and on the other hand, the two levels of content that Halliday does propose, meaning and wording, are labelled as such as early as Halliday & Hasan (1976: 5):

[2] As shown in numerous previous posts, the model of language presented in Figure 4 of Chapter 3 cannot be "common" to any systemic functional grammars, not least because of its serious internal inconsistencies.  Moreover, Fawcett's model conceives of human language as a Fordian production line in which operations are performed in separate modules, producing outputs — largely because it is designed, instead, for the purpose of text generation by computers.

The Place Of Syntax In Fawcett's Model

Fawcett (2010: 43, 43n):

We can summarise this chapter so far by saying that, in terms of Figure 4, the place of syntax in a model of language is in the syntagmatic relations at the level of form. At various points in the rest of this book, therefore, we shall find ourselves thinking in terms of either (1) instances of syntax or (2) the syntax potential that specifies those outputs from the grammar.*

* The other parts of the 'form potential' are the 'lexis potential', the 'intonation potential' and the 'punctuation potential'.

Blogger Comments:

This again refers to Figure 4:

[1] To be clear, since it is concerned with syntagmatic relations only, and not paradigmatic relations, the place of syntax in Fawcett's model of language is on the syntagmatic axis.

[2] As Figure 4 shows, Fawcett regards an instance of syntax as a structure, and syntax potential as realisation statements.  That is, on this model, structures are instances of realisation statements.  Moreover, instances are specified by realisation rules.

[3] To be clear, in SFL theory, Fawcett's 'lexis potential' is modelled as the most delicate systems on the stratum of lexicogrammar; Fawcett's 'intonation potential' is modelled as systems at the rank of tone group on the stratum of phonology; and Fawcett's 'punctuation potential' would be modelled as systems on the stratum of graphology.  In SFL theory, lexicogrammar and phonology/graphology are distinguished as different levels of symbolic abstraction (strata), with lexicogrammar as a level of content, and phonology/graphology as (parallel) levels of expression.  In Fawcett's model, then, content and expression are modelled as being of the same level of symbolic abstraction.

Fawcett's Two Aspects Of Syntagmatic Relations

Fawcett (2010: 43):

This book, then, focuses on syntagmatic relations. There are in fact two aspects to syntagmatic relations in language: part-whole relations and sequential relations. The more fundamental concept is that of part-whole relations, and while syntagmatic relations are usually thought of in terms of the level of form, part-whole relations are found at both the level of semantics and at the level of form. But it is the level of form that we shall focus on here, i.e., as shown in Figure 4 in Section 3.2.

Blogger Comments:

[1] To be clear, in SFL theory (Halliday & Matthiessen 2014: 60, 83) there is an important distinction between:

• a structure: a configuration of functions, such as Senser ^ Process ^ Phenomenon

and  

• a syntagm: a sequences of classes (of form), such as nominal group ^ verbal group ^ nominal group.

[2] To be clear, in SFL theory, part-whole relations are organised as a rank scale.  Halliday & Matthiessen (2014: 21):

Structure (syntagmatic order) … is the compositional aspect of language, referred to in linguistic terminology as ‘constituency’. The ordering principle, as defined in systemic theory, is that of rank: compositional layers, rather few in number, organised by the relationship of ‘is a part of’.