"The Focus In The Present Book"

Fawcett (2010: 158):

However, the focus in the present book is not on text analysis, but on establishing the concepts that are required in a theory of syntax for SFL. We shall therefore not discuss such representations further, except in so far as they affect the theory of syntax that is required in a modern SFL. Clearly, the focus from now on must be on the nature of such a theory — and this is, of course, the topic of Part 2. 

Up to this point, Part 1 has been principally concerned with establishing the antecedents to a modern theory of SF syntax in terms of the Sydney Grammar, principally through the writings of Halliday himself. However, there is a third set of post-"Categories" concepts that has had some influence on a number of current descriptions of systemic syntax, and it is this body of work that is the direct antecedent of the framework of concepts to be set out in Part 2. It is to this work that we turn in Chapter 8.

Its genesis is significantly different from that of either "Systemic theory" or IFG, in that it was formulated as a theory of syntax in the 1970s as an explicit response to Halliday's still recent proposal that the system networks should regarded as modelling the meaning potential of a language. It is not impossible that Halliday might have produced a theory of syntax that would have been something like the one to be considered in the next chapter, if in the 1970s he had chosen to develop SFL in the way indicated in Sections 4.3 and 4.9 of Chapter 4, i.e., with the system networks of TRANSITIVITY, MOOD, THEME and so on treated as explicitly modelling semantics.

Blogger Comments:

[1] To be clear, as Halliday (1985 & 1994: xiv) pointed out, a model in which syntax is opposed to semantics, like Fawcett's, is inconsistent with the theoretical assumptions and method of inquiry of SFL Theory:

[2] As previously pointed out, Halliday's 'meaning potential' refers to language as system, as opposed to instance (text). Fawcett repeatedly misunderstands the term as referring just to the semantic stratum.

The Implications Of The Argument For A Theory Of Syntax For SFL [2]

Fawcett (2010: 157-8):

However, since the current situation is that there is not yet a published work in any version of SFL that demonstrates how to use the features in system networks for the systemic analysis of texts, IFG-style analyses may still have a role to play for a while. The system networks for the Sydney Grammar have of course now been published (in Matthiessen 1995), but these networks are not in fact designed for use in text analysis, and they are often hard for the reader to interpret. Moreover there are no published guidelines as to how to use system networks for the task of making analyses of real-life texts, and no published examples of such analyses, other than those in Fawcett (1999) and in Figure 10 of Section 7.9. The publication of Fawcett (in press) and Fawcett (forthcoming a), as described at the end of Section 7.8, make available the Cardiff Grammar versions of (a) the functional structure of text-sentences and (b) system networks that are equivalent in their coverage (but more explicitly semantic) to the Sydney Grammar networks found in Matthiessen (1995). Moreover Fawcett (forthcoming a) contains guidelines that show the analyst how to go about making such analyses, and it contains many such examples.

Blogger Comments:

[1] This is still misleading, because it is still untrue. As previously noted, Halliday & Matthiessen (2004) provides, for each metafunction, clause analyses which display the features selected.

[2] To be clear, the structures exemplified in IFG represent an integral dimension of SFL Theory.

[3] To be clear, in SFL Theory, system networks are the formal means by which language is modelled, and which specify the structures that Fawcett equates with text analysis.

[4] To be clear, here Fawcett presents his own difficulty in interpreting system networks as a problem with the networks themselves.

[5] This is misleading, because it is untrue. As previously noted, Figure 10 does not include the networks from which its features are derived, and misrepresents paradigmatic features of the entire clause as individual elements of syntagmatic structures ("strands of meaning").

[6] As previously noted, these promised works — Fawcett (in press) and Fawcett (forthcoming a) — remain unpublished, 20 years after the first edition of this publication.

The Implications Of The Argument For A Theory Of Syntax For SFL [1]

Fawcett (2010: 157):

The main purpose of this chapter has been to clarify the status of the representations of clauses given in IFG and the derived works. We have also seen, in Section 7.9, that there is an alternative and more truly systemic means of representing the multifunctional nature of language, i.e., a representa[t]ion of the systemic features. In systemic functional grammar such a representation is inherently more revealing than any structural representation. As Halliday himself says, "the system takes priority; the most abstract representation [...] is in paradigmatic terms. Syntagmatic organisation is interpreted as the 'realisation' of paradigmatic features" (Halliday 1993:4505).

The consequence of having both a representation at the level of meaning and a representation of the single integrated structure — something that is required, as we have seen, in the Sydney model as well as the Cardiff model — is that the many lines of structure in an IFG analysis are, from both the theoretical-generative and the text-descriptive viewpoints, redundant. They are redundant because the analysis in terms of the features chosen in the system networks already displays clearly the different strands of meaning, as in the example of such an analysis in Figure 10.

Blogger Comments:

[1] We have also seen, in the examination of Section 7.9, that Fawcett's "alternative and more truly systemic means of representing the multifunctional nature of language" — Figure 10 — confuses systemic features with structural elements, thereby attributing features of the whole clause to elements of clause structure. These theoretical inconsistencies, alone, invalidate Fawcett's model.

[2] This is true, and Fawcett has repeatedly demonstrated that he does not understand what Halliday means by it. Despite the system taking priority in SFL Theory, Fawcett has focused instead on structure, not just in this chapter, not just in this entire book, but in his reworking of Halliday's theory — even to the extent of misrepresenting systemic features as structural elements (Figure 10).

[3] This is misleading, because it is untrue, as has been demonstrated previously, over and over. In the "Sydney model" (Halliday's version of his own theory), there is no need for an "integrated structure", since it is the realisation of the metafunctional clause structures in a syntagm of clause constituents, groups and phrases, that provides the skeleton onto which the functions are mapped. As previously explained, Fawcett's misunderstanding arises from his confusing formal constituency (the rank scale) with function structure. As will be seen, Fawcett argues against the theoretical value of a rank scale.

[4] This is misleading, because it is untrue. Fawcett's alternative analysis (Figure 10) does not make metafunctional clause structures redundant, because his semantic analysis, which displays paradigmatic features, does not present an analysis of syntagmatic structure:

An Addendum To The Argument For A Theory Of Syntax For SFL

Fawcett (2010: 156-7):

As an addendum to the line of argument set out here, we should note that there are important theoretical differences between the Cardiff Grammar and the Sydney Grammar (within the general framework summarised in Figure 6 in Section 4.9 of Chapter 4). The most important difference is in way in which the various 'strands of meaning' are represented. While the Cardiff Grammar represents the meanings systemically — i.e., in terms of the features that have been chosen in the system networks that constitute the 'meaning potential' of the language — Halliday represents the meaning by 'functions'. His use of this concept for representing meanings can be traced back to the stage in the development of the theory in which it was believed that the realisation of the features in the system networks necessarily took place in two stages: in the first the grammar would use the 'selection expression' of features that were chosen on a traversal of the system network to create an unordered string of 'functions' (and of conflated functions), and in the second these 'functions' would be ordered in sequence.²⁶

²⁶ See Berry (1977:29-31) for an introductory account of this version of the model. The fullest implementation by far of the concept of realisation in two stages is that described in Hudson (1971), which had 76 such 'functions' (these being explicitly at the level of form, however). In a later work Hudson reduced them to just three (Hudson 1976), and this constitutes a little subsidiary evidence for the position taken here, i.e., that it is unnecessary for a systemic grammar first to generate 'functions' and then to order them.

Blogger Comments:

[1] This is true, as the careful examination on this blog demonstrates. For example, the "Sydney Grammar" is the theory that Halliday painstakingly created, whereas the "Cardiff Grammar" is merely Fawcett's alteration of the theory that Halliday painstakingly created. More importantly, as demonstrated over and over again, Fawcett repeatedly misrepresents Halliday's model, while his own model, Figure 4, is invalidated by its own internal inconsistencies.

[2] This is very misleading indeed. The discussion in Section 4.9 is not concerned with Halliday's "Sydney Grammar", but with an earlier development (c1970) that was soon superseded, as previously explained. Figure 6, in which Fawcett has added the label 'COMBINED' to Halliday's diagram, otherwise displays the earlier model:

[3] To be clear, as this blog demonstrates, this is not the most important difference between the Cardiff Grammar and Halliday's version of his own theory, but it nicely illustrates the type of misunderstandings that invalidate Fawcett's version of Halliday's theory. As previously explained, Fawcett represents 'strands of meaning' — that is, structures — as systemic features.  As well as confusing the syntagmatic with the paradigmatic axis, this results in features of the clause being  represented as features of clause constituents.

Halliday's 'functions', on the other hand, are elements of function structure, with each structure of the clause being the relation between elements.

[4] To be clear, this "concept for representing meanings" is the syntagmatic dimension of the theory. As previously explained, Fawcett's model (Figure 4) misrepresents the relation between system and structure as a (stratified) relation between potential and instance:

[5] To be clear, other people's interpretations of earlier versions of Halliday's theory do not represent the "Sydney Grammar". That is, the inclusion of this in Fawcett's argument constitutes the logical fallacy known as a red herring.

[6] This is misleading, because it misrepresents Halliday's theory. To be clear, the grammar does not first generate "functions" — i.e. elements of function structure — and then order them. In SFL Theory, the grammar is modelled as a system of choices, with different choices specifying the different structures that realise them.

The Logical Conclusion Of The Argument For A Theory Of Syntax For SFL [3]

Fawcett (2010: 156):

16. As a final point, we might note that the Cardiff Grammar has a simpler architecture than the Sydney Grammar with respect to the relationship between the system networks of TRANSITIVITY, MOOD, THEME etc. and their realisations in form. The former is summarised in Figure 4 of Chapter 3 and the latter in Figure 8 of Section 7.4 3. The Cardiff model is also simpler in that its system networks have been semanticised to the point where there is no need to have a higher layer of networks, as there seems to be in the model described in Halliday (1994) and Matthiessen (1995).

Blogger Comments:

[1] To be clear, as previously demonstrated, the simpler architecture of the Cardiff Grammar (Figure 4) is riddled with internal inconsistencies that invalidate it as a model. These include positing realisation rules involving semantic features at the level of form, such that realisation rules realise system networks, and positing structure as an instance of realisation rules:

[2] To be clear, the figure in Section 7.4.3, Figure 9*, says nothing about "the relationship between the system networks of TRANSITIVITY, MOOD, THEME etc. and their realisations in form" in the "Sydney Grammar" since it is merely a metafunctional analysis of clause structure, without reference to the systems that specify each structure:

* Figure 8 presents a modification of Figure 4 to accommodate Fawcett's mistaken notion of structure conflation.

[3] To be clear, Fawcett's "semanticisation" of Halliday's grammatical grammatical systems is merely his relocation of them from Halliday's stratum of lexicogrammar to his level of meaning. Fawcett does not present any of his "semanticised" versions of these networks in the entire publication.

More importantly, as previously noted, the distinction between semantic networks (Halliday & Matthiessen 1999) and lexicogrammatical networks (Halliday & Matthiessen 2004, 2014) is warranted by the explanatory power it provides in modelling grammatical metaphor.

The Logical Conclusion Of The Argument For A Theory Of Syntax For SFL [2]

Fawcett (2010: 155-6):

13. The 'multiple structures' showing strings of 'functions' for each strand of meaning are therefore redundant.

14. However, while they are not necessary, there might still be some reason why they are desirable. The reason that seems most likely to be justifiable is that they may help the student of language to picture some central aspect of language, such as the fact that clauses — or, more accurately, several elements of structure in a typical clause — express several different meanings at the same time. 

15. However, the concept of 'strands of meaning' in a clause can be modelled in an equally insightful (and more accurate) way as part of the representation of the semantics of the clause (as exemplified in Figure 10 of Section 7.8). When guidelines for analysing clauses in these terms (and also in setting out such representations) are available, there will be no logical reason to continue using 'multiple structures' for this task — especially if it is agreed, as proposed here, that the concept that a clause consists of a set of several different 'functional structures' should no longer be a part of the theory.

Blogger Comments:

[1] To be clear, the theoretical value of metafunctional structures of the clause lies in their explanatory power: modelling the functional relations between the elements of structure according to metafunction.

[2] This is misleading, because it is untrue. To be clear, Fawcett's model is neither "equally insightful" nor "more accurate". As Figure 10 (p148) demonstrates, Fawcett' syntax mixes experiential elements (Agent, Affected) with interpersonal elements (Subject, Operator, Complement, Adjunct) with formal constituents (Main verb). This theoretical inconsistency has the further defect of being inconsistent with the SFL notion of structure as the relation between elements.

Moreover, Fawcett's semantic structure mixes syntagmatic elements (e.g. agent, affected, subject, theme) with paradigmatic features (e.g. positive, unassessed). This theoretical inconsistency has the further defect of misattributing features of the entire clause to individual elements of the clause.


[3] As previously observed, these guidelines (Fawcett in press and Fawcett forthcoming) are still unavailable, 20 years after the first edition of the work under discussion.

The Logical Conclusion Of The Argument For A Theory Of Syntax For SFL [1]

Fawcett (2010: 155):

This is the point that the argument has reached so far. Let us now follow it through to its logical conclusion: 

10. Since it is agreed that, within Halliday's approach, the multiple structures must be integrated into a single structure as the final stage in generation, the next logical step is to ask whether it is in fact possible to first generate the type of multiple structure representation found in IFG as an intermediate stage between the systemic features and the single representation, and then to generate from the multiple structures the single integrating structure (which would require the architecture illustrated in Figure 8 in Section 7.4.1). If the answer were to be that it is possible (which I doubt), we would then need to ask whether this way of generating the final integrated structure is either necessary or desirable. The alternative that should be considered is to generate it directly by realisation rules that take as their input the semantic features chosen in the system network. 

11. This is precisely what is done in the very large computer-implemented Cardiff Grammar (as we saw in Section 7.8). The fact that it is possible to generate text-sentences directly from systemic features demonstrates that there is no need first to generate multiple structures of the type illustrated in IFG and then to integrate them into a single structure. 

12. We therefore know that it is not necessary to generate multiple structures of functions as an intermediate stage. This fact is particularly welcome, given that it is not possible to build a component that will convert seven or more different non-coterminous structures into a single structure (for the reasons given in Section 7.4.1).

Blogger Comments:

[1] This is misleading, because it is untrue. As previously explained, in Halliday's approach, there is no requirement that the metafunctional clause structures be "integrated into a single structure as the final stage in generation". Fawcett's claim arises from his misunderstandings of formal constituency and functional structure, and is motivated by the fact that his own model has a "single structure as the final stage in generation". That is, his own model is posited as solving a non-existent problem he wrongly attributes to Halliday's model.

[2] This is misleading, because the architecture of Figure 8 is a reworking of Fawcett's model (Figure 4), not Halliday's, and is inconsistent with the architecture of SFL Theory. See, for example, Fawcett Adjusting His Model To Accomodate His Misunderstanding Of Halliday.

[3] This is misleading. The metafunctional clause structures in IFG are "directly generated by" (realise) systemic features — those of clause rank, on the stratum of lexicogrammar. The theoretical motivation for distinguishing grammatical systems from semantic systems is the explanatory power that the distinction provides for understanding grammatical metaphor.

As previously explained, Fawcett's model, Figure 10, misrepresents systemic features as structural elements, thereby misattributing features of the whole clause to individual elements of clause structure.

The Need For A Theory Of Syntax For SFL: The Argument Summarised [5]

Fawcett (2010: 154-5):

8. Halliday recognises that the 'multiple structures' of IFG — which we shall assume for the moment to be capable of being generated by the grammar — must ultimately be merged into a single structure at the level of form. 

9. However, there are several serious problems in reconciling Halliday's multiple structure representations with this position. No advocate of it has yet stated what it actually means when one says that the various structures get conflated with each other — other than treating all such 'functions' as elements of a 'flat tree' representation of constituency, which would be the dismemberment of those structures rather than a 'conflation of them. Such a conflation has not even been achieved with simple clauses, either in the framework of a computer implementation or as a theoretical exercise (as I have shown in Section 7.4). On the basis of this evidence and my own failure to find a way to make this approach work, I make two proposals:

(1) 'multiple structures' with non-coterminous elements cannot be integrated in the manner proposed by Halliday, so that another solution to the problem of generating a single structural output must be sought; and 

(2) even if such structures could be integrated, something like an updated version of S&C-type syntax would be required to achieve this — i.e., a syntax such as that to be described in Part 2.

Blogger Comments:

[1] This is misleading, because it misrepresents Halliday's model. In SFL Theory, there is no merger of structures, no structures at the level of form, and no level of form. Instead, the three metafunctional clause structures are all realised by the same syntagm of units at the rank below the clause (group/phrase) — at the level (stratum) of lexicogrammar.

[2] This is misleading, because 'this position' is not Halliday's position; see [1].

[3] To be clear, the reason why no-one has tackled the problem of structure conflation is that structure conflation is not part of SFL Theory. The notion only arises through Fawcett's misunderstandings of Halliday's notions of structure and constituency, as previously demonstrated.

[4] Here Fawcett repeats his incomprehension of 'flat tree' constituency.  'Flat tree' constituency describes ranked constituent analysis (minimal bracketing) as opposed to immediate constituent analysis (maximal bracketing); see Halliday (1994: 20-4). That is, it refers to the rank scale of forms in SFL theory. Elements of function structure, on the other hand, are not formal constituents, but the functions (e.g. Sayer) that formal constituents (e.g. nominal group) realise.

[5] This is misleading, because it is untrue. See the earlier post Misconstruing The Conflation Of Elements As The Dismemberment Of One Element.

[6] To be clear, here Fawcett proposes solving a non-existent problem in SFL Theory by using an approach that is inconsistent with SFL Theory (syntax) that is derived from Halliday's superseded theory (Scale and Category Grammar).

The Need For A Theory Of Syntax For SFL: The Argument Summarised [4]

Fawcett (2010: 154):

7. Thus Halliday now presents the structure of a clause (and in principle other units) as multiple structures. In any such representation of a clause there is, in principle, a line of structure that corresponds to each 'metafunction' (but in practice several more, with seven for many text-sentences as analysed in IFG). Each line consists of a string of 'functional elements', or 'functions'. The problem is that the elements of these 'structures' are not necessarily coterminous (and nor are the structures themselves coterminous). In other words, if two or more elements in two or more strands of meaning are coterminous, they can be conflated with each other to form a compound element, but when they are not coterminous it is not logically possible to conflate them (as described in Section 7.4 of this chapter). Despite these serious theoretical-generative problems, it appears that Halliday may regard the notion that there are many different structures (and so many non-coterminous elements) in a single clause as one of the riches of language — at least from the text-descriptive viewpoint, as exemplified in the analyses in IFG.

Blogger Comments:

[1] This is misleading, because it is untrue. In IFG, what Fawcett refers to as "text-sentences" — his model, not SFL Theory — are the three metafunctional clause structures. The reason why Fawcett thinks there are "several more" is that he mistakes information structure as a clause structure, and counts mood and thematic structures twice because of their structural components (e.g. Subject of the Mood element, and topical Theme of the Theme element).

[2] To be clear, as previously explained, in SFL Theory, there is no need to conflate any elements. It is simply the case that, under certain circumstances, some elements do happen to conflate, as when Subject conflates with Behaver in behavioural clauses. The three metafunctional structures of the clause are integrated in the syntagm of units at the rank below, as when Subject and Behaver are realised by the same nominal group.

[3] To be clear, the conflation of two elements, such as Subject and Behaver, does not result in a "compound element", since each element is an element of a different structure, and, in SFL Theory, a structure is the relation between elements (e.g the relation of Subject to Finite, or the relation of Behaver to Process).

The Need For A Theory Of Syntax For SFL: The Argument Summarised [3]

Fawcett (2010: 153-4):

5. In two of the papers in which Halliday takes the position that the system networks of TRANSITIVITY, MOOD and THEME are at the level of semantics (1970/76b and 1977/78) he includes diagrams that show (1) several lines of analysis, each of which clearly corresponds to a major strand of meaning, and (2), below these, a single line of analysis that shows the 'Scale and Category' elements of the clause, such that this line represents the integration of the metafunctional representations above it. It is from this stage of Halliday's thinking that the Cardiff Grammar version of SFL has developed (as shown in Section 4.9 of Chapter 4).

6. However, these two papers represent just a brief period in the development of Halliday's concept of how structure should be represented, and almost immediately he dropped the integrative representation in his diagrams in favour of the type of representation shown throughout IFG and his subsequent descriptive works to date. Here the original S&C elements of Subject, Predicator, Complement and Adjunct are no longer shown as elements of the structure that integrates the various strands of meaning, but instead they are introduced as a structure within the representation of 'interpersonal' meaning (with the addition of the Finite). Indeed, Halliday presents them as merely the 'secondary' structure of 'interpersonal' meaning, the 'primary' structure being that of "Mood + Residue" (also called "Modal + Propositional", in earlier works). Halliday's reasons for including the elements of 'Predicator', 'Complement' and 'Adjunct' in the 'interpersonal' line of analysis (which were summarised in Section 7.2) are considered here to be unpersuasive — especially in comparison with the much more persuasive reason for assigning them the role described in Point 5 above, namely their role as elements of the structure that integrates the various strands of meaning (with others, as illustrated in Appendix B).

Blogger Comments:

[1] This is misleading, because it misrepresents Halliday's final theory, Systemic Functional Grammar, in terms of his earliest theory, Scale and Category Grammar (1961). The terms 'primary and 'secondary' — but not 'merely' — derive from Halliday originally applying the principle of delicacy to structure (Halliday 2002 [1961]: 48). This old view is inconsistent with SFL Theory, where delicacy is a type of elaboration, whereas structural composition — e.g. Mood composed of Subject and Finite — is a type of extension; see Halliday & Matthiessen (1999: 145-6).

[2] See Fawcett's previous misunderstandings and bare assertions on this matter here. To be clear, labelling Halliday's reasons 'unpersuasive' and Fawcett's reason 'persuasive' is not reasoned argumentation based on evidence. Importantly, in SFL Theory, the integration of 'the various strands  of meaning' of the clause occurs in the form that realises the elements of clause structure: the syntagm of units on the rank below the clause.

Fawcett, on the other hand, argues against the value of a rank scale, and uses a blending of Halliday's interpersonal and experiential functions as his means of integrating 'the various strands of meaning' at his level of form, as his Figure 10 (p148) demonstrates:

The Need For A Theory Of Syntax For SFL: The Argument Summarised [2]

Fawcett (2010: 153):

4. However, Halliday's writings in the 1970s and 1980s have explored two approaches to modelling meaning in language. One of these involves treating the system networks for TRANSITIVITY, MOOD, THEME and so on as the level of semantics. In the second model these networks are considered to be, in some sense, at the level of form — though it is a 'form' that has been 'pushed in the direction of the semantics' (Halliday 1994:xix) — and there is (in principle) a higher stratum of even more 'semantic' system networks that correspond to each of the networks of TRANSITIVITY, MOOD, THEME, etc. I prefer the first (and simpler) of these two positions, and I consider that the phenomena for which Halliday sets up the second level of 'meaning' can be more appropriately accounted for in other ways (as argued in Sections 4.6 to 4.8 of Chapter 4).

Blogger Comments:

[1] To be clear, in Halliday's superseded model, written between 1972 and 1976, structures produced by semantic systems were mapped onto units of the grammatical rank scale. By the 1980s, these semantic systems were reconstrued as lexicogrammatical systems, so as to account for grammatical metaphor.

[2] This is misleading on two counts. Firstly, in the second model — the current model — these systems are located on the stratum of lexicogrammar, not form; 'form' is a level in Fawcett's model only. In Systemic Functional Grammar, form is modelled as a rank scale of units, each of which serves as the entry condition to functional systems. Secondly, the quote from Halliday (1994: xix) is explicitly concerned with a functional grammar, not form:

[3] To be clear, 'I prefer' and 'I consider that…' are not reasoned argumentation based on evidence.

[4] This very misleading indeed. Fawcett provides no argument whatsoever in Sections 4.6 to 4.8 (pp53-70) for the "other ways" that grammatical metaphor can be "more appropriately accounted for". The only statement in this regard, in the entire three sections, is yet another bare assertion, and the promise of a still unpublished work, in footnote 7 on page 57:

I consider that the phenomena that have led Halliday to adopt his latest position are to be explained in other ways (one being to further semanticise some of Halliday's networks, e.g., that for MOOD). See Fawcett (forthcoming a) for a set of such semantic system networks.

The Need For A Theory Of Syntax For SFL: The Argument Summarised [1]

Fawcett (2010: 152-3):

We are now in a position to bring together the main steps of the argument that is presented this book. 

1. Scale and Category Grammar (S&C) was essentially a theory of syntax, and so an account of the structure of language at the level of form (as we saw in Chapter 2). 

2. A modern, computer-implementable theory of language must have levels of form and meaning, providing for both a 'potential' and 'instances' of the potential at both levels (as described in Chapter 3). 

3. Halliday's development of the theory into Systemic Functional Grammar involved (1) placing the concept of 'system' at the centre of the theory, (2) presenting the resulting system networks as choices between meanings, and (3) seeing each clause as the realisation of several different 'strands of meaning' (as outlined in the first half of Chapter 4).

Blogger Comments:

[1] This is misleading, because it is untrue, as previously demonstrated here. Even Halliday's initial theory went well beyond mere syntax. Moreover, any argument that includes Scale and Category Grammar is an instance of the Red Herring logical fallacy, since this was a distinct theory from Systemic Functional Grammar, and hence irrelevant to considerations of the latter.

[2] This is misleading, because it merely restates Fawcett's model (Figure 4), giving the false impression that its theoretical architecture has been established by reasoned argument. Moreover, Fawcett's model is invalidated by multiple internal inconsistencies, such as the confusion of axis with instantiation, as previously demonstrated here.

[3] This is misleading, because it couches Halliday's theory in terms of Fawcett's model, which only distinguishes meaning and form, and only has systems at the level of meaning. Halliday distinguishes three levels of language — meaning, wording and sounding — with lexicogrammatical systems as systems of wording. It is only in the absence of grammatical metaphor that meaning and wording are in accord ('congruent').

[4] This is misleading, because it misrepresents Fawcett's theoretical priority, the syntagmatic axis, as Halliday's theoretical priority. In Systemic Functional Grammar, priority is given to the paradigmatic axis, such that a clause is modelled as conjoint metafunctional systems that specify structural realisations.

Fawcett's Claim That IFG-Style Analyses Are Misleading

Fawcett (2010: 152):

However, Halliday also occasionally states that it is desirable to represent the clause in terms of the features selected in the system networks that show the language's 'meaning potential', i.e., the features from which it has been generated. The problem is that he provides no examples of descriptions of this sort. It is hard to understand why this should be so. It may be that the need to provide a systemic representation seems less urgent to those working in the Sydney Grammar, because an IFG-style analysis already expresses (though in a misleading way, I would of course say) the multifunctional nature of language. 

In the Cardiff Grammar, on the other hand, a systemic representation in terms of the features that have been chosen in generating the text is regarded as a vital second level of representation in any attempt to represent the meanings as well as the forms of language. As the lower half of Figure 10 shows, the systemic representation takes the form of a display of the key features from the selection expression, arranged in separate lines, such that each feature is entered in a column below the element to whose generation it contributes. Since it is at this level that meanings are modelled, it is natural to show the strands of meaning at this level too — and there is therefore no need to show them at the level of form as well. The representation of functional structure in the Cardiff framework is therefore much simpler than its equivalent in IFG, in that there is only a single structure at this level — not the five to seven (or even sometimes eight or more) structures that are shown in the analyses in IFG. To see the difference, compare the structural analysis in Figure 10 with that in Figure 7.

Blogger Comments:

[1] This is misleading, because it is untrue. Halliday & Matthiessen (2004) provides examples for theme (p101), mood (p158-9, 164-7) and transitivity (p304-5).

[2] To be clear, this is easier to understand if Halliday's own explanation in the Introduction to the first two editions of IFG is taken into account. Halliday (1985 and 1994: xv):

[3] As previously demonstrated, this is misleading because it is untrue.

[4] As previously explained, the Cardiff Grammar analysis (Figure 10) confuses the syntagmatic axis (structure at the level of syntax) with the paradigmatic axis (features at the level of semantics) and, in doing so, misattributes features of the clause to individual elements of clause structure.

[5] As this wording makes clear, Fawcett mistakes systemic features (paradigmatic axis) for syntagmatic structures ('strands of meaning').

[6] As previously explained, the Cardiff Grammar analysis (Figure 10) violates this stipulation by including metafunctional meanings at the level of form "as well". The interpersonal meanings are Subject, Complement and Adjunct; the experiential meanings are Agent and Affected (Medium).

[7] To be clear, this is misleading, because the Cardiff Grammar analysis (Figure 10) is not simpler,  when taken as a whole, since it requires more levels and more types of meaning to provide an analysis equivalent to that of the "Sydney Grammar" (Figure 7). Consider how little Figure 10 actually analyses, if the semantic level is not included.

[8] To be clear, this is misleading, because it is untrue. As previously explained, the analyses of clause structure in IFG feature only three lines of meaning, one for each metafunction: theme (textual), mood (interpersonal) and transitivity (experiential). Fawcett inflates this number by misconstruing information as a system of the clause, and counting mood and theme structures twice. Fawcett's own model (Figure 10) involves seven levels analysis.

Two Further Problems With Halliday's Approach To Modelling Structure

Fawcett (2010: 151-2):

Two further problems with Halliday's current approach to modelling structure are that we are not shown (1) what the final, integrated structure would be like, or (2) what the component that performs this integration on structures with non-coterminous elements would be like. It seems likely that the final structure would have to be something like the structure shown in the upper half of Figure 10, but after Halliday (1970/76b) and (1977/78), Halliday appears to have abandoned that position, as we saw in Section 4.9 of Chapter 4.

So far as I am aware, the present book is the first publication in which the point is made that IFG-style representations do not play a role in the theoretical-generative version of the theory — and indeed that they are incapable of doing so, because of the inherent problem that it is only possible to conflate elements (or indeed units) when they are coterminous.

Blogger Comments:

[1] This is misleading, because it misrepresents SFL Theory. To be clear, the reason why "we are not shown" "the final integrated structure" or "the component that performs the operation" is simply that a "final integrated structure" is inconsistent with SFL Theory, as previously explained. In SFL Theory, the three metafunctional structures of the clause are integrated in the formal syntagm of groups and phrases that realises clause structure.

[2] To be clear, the reason why the present book is the first publication to make these claims is that the claims are invalid, since they are based on Fawcett's misunderstandings of SFL Theory, as demonstrated in many preceding posts.

Misrepresenting Halliday As Misrepresenting His Own Theory

Fawcett (2010: 151):

The problem for the text analyst who is using SF grammar is that of how best to represent all three of (1) its functional syntax, (2) its meanings and (3) the multifunctional nature of language. 

As we have seen, the Sydney Grammar's approach is to present each clause as having several different functional structures, one for each of the five to eight lines of analysis that are recognised in that approach. (There are not just three or four, as many of the writings of Halliday and others imply). In this view every clause is a conflation of several different structures, each roughly of clause length (give or take an element or two). Any one element in such a structure is said to be capable of being conflated with other elements in other structures — including ones that are not coterminous. This approach produces attractive diagrams that illustrate the 'strands of meaning' metaphor of language. However, such diagrams are essentially a misrepresentation of the real position, because it is only some functional elements in some of the functional structures that are recognised in this version of the theory that can be conflated in this manner, and there are long stretches in several of the strands of meaning in which there is nothing to be said about the meaning (i.e., the boxes labelled 'Rheme', 'Given' and 'Residue', in Sydney Grammar analyses).

Blogger Comments:

[1] This is misleading, because it misrepresents Systemic Functional Grammar. On the one hand, 'functional syntax' is a component of Fawcett's Cardiff Grammar only, and, as Halliday (1985/1994: xiv) has explained, is inconsistent with the theoretical approach of Systemic Functional Linguistic Theory. On the other hand, the 'multifunctional nature of language' is not distinct from meaning, but its metafunctional differentiation.

[2] This is misleading, because it is untrue, as previously explained. In the "Sydney Grammar" (Halliday's version of the theory he himself created), clause structure involves three strands of meaning: theme (textual), mood (interpersonal) and transitivity (experiential). Fawcett inflates this by counting mood and theme twice, and by mistaking information as a structure of the clause.

[3] This is misleading, because it is untrue, as previously explained. In Halliday's theory, there is no "structure conflation", and given that structure is the relation between elements, structure conflation is nonsensical.

[4] This is misleading, because it is untrue, as previously explained. In Halliday's theory, element conflation is a realisation rule that specifies structural realisations of systemic choices. This is not specified for all elements, and it only applies to elements that are realised by the same syntagmatic unit ("coterminous elements").

[5] This is misleading, because 'strands of meaning' is a congruent realisation of the theory, not a metaphorical rendering that stands for something else.

[6] This is misleading, because it is untrue. Halliday's structural diagrams are faithful representations of his theory. The only misrepresentations here are Fawcett's, as multiple lines of evidence on this blog have demonstrated.

[7] This is misleading, because it misrepresents a statement that is consistent with Halliday's model as inconsistent with it.

[8] This is misleading, because it is untrue. On the one hand, Rheme and Residue (clause) and Given (information unit) are single elements. The notion of "long stretches" betrays the fact that Fawcett view is of formal constituents, not functional elements. On the other hand, and more importantly, the meaning of each of Rheme and Residue (clause) and Given (information unit) involves what might be termed 'negative structural complementarity':

• Rheme ("not Theme") structurally complements the Theme element of the clause,
• Residue ("not Mood") structurally complements Mood element of the clause,
• Given ("not New") structurally complements New element of the information unit.

Fawcett (in press) and Fawcett (forthcoming a)

Fawcett (2010: 150-1):

Very large quantities of text analysis in terms of the Cardiff Grammar have been undertaken over the last twenty-five years as part of its development — most at the level of functional syntax (e.g., in the major text-analysis project described in Fawcett & Perkins 1980), but also in the last decade at the level of meaning (a good example of the latter being Ball 1995). Tucker (1998) provides a very fine introduction to the treatment of adjectives and of all of the structures into which they may enter in Cardiff Grammar terms, but since Fawcett (1974-6 /81) there has unfortunately not been a publication that offers overall coverage of either level of description.

However the publication of Fawcett (in press) will make publicly available both a broad coverage grammar of functional syntax and full guidelines for analysing texts in these terms. Then Fawcett (forthcoming a) will make available, in a form that is usable for text analysis, the major system networks for English, in a systemic functional grammar in which the networks constitute the level of semantics. This volume will also provide guidelines for using the system networks for the semantic description of texts.

Blogger Comments:

[1] This promised 'in press' publication

• Fawcett, Robin, in press. Functional Syntax Handbook: Analysing English at the Level of Form. London: Continuum.

remains unpublished, 20 years after the first edition of the publication under review. It is current publication date is listed as November 2022, having previously been promised for January 2022, November 2021 and January 2020.

[2] This promised 'forthcoming' publication

• Fawcett, Robin, forthcoming a.  Functional Semantics Handbook: Analysing English at the Level of Meaning. London: Continuum .

also remains unpublished, 20 years after the first edition of the publication under review. No expected publication date is currently available.

Misrepresenting A Disadvantage Of The Cardiff Grammar As A 'Great Advantage'

Fawcett (2010: 150n):

One great advantage of a model in which the system network is explicitly semantic is that it often avoids the need to introduce the 'double analysis' of a single 'strand of meaning', which involves what Halliday calls 'grammatical metaphor'. (For an introduction to this set of concepts see Halliday 1994:340-67.) 

Consider the case of Could you read that again? In the Cardiff Grammar this is analysed as directly expressing the meaning of 'request' (a choice in the semantic network for MOOD), but in IFG there is no such feature at this level of analysis and it would be analysed as a 'polar interrogative' — but one which the Addressee is expected to re-interp[r]et as the 'imperative' Read than again\. 

A similar example of "interpersonal metaphor" occurs in IFG (p. 378), where Halliday first analyses you have to be beautiful with it as a 'declarative', and then re-expresses it as be beautiful with it and analyses it as an 'imperative'. In the Cardiff Grammar, it is treated as a direct realisation of an option in the semantic MOOD network (specifically, as a sub-type of 'directive' with the feature 'requirement').

Blogger Comments:

[1] This is misleading, because it misrepresents — in a footnote — a significant disadvantage of the Cardiff Grammar as "one great advantage". In Halliday's model, grammatical metaphor involves making two meanings at once, and this is unaccounted for in a model (Figure 4) whose levels of analysis are merely meaning and form.

[2] To be clear, the wording Could you read that again? could realise either:

• a command: Could you read that again (please)?, or
• a question: Could you read that again (if you had to)?

Fawcett offers no principled way of differentiating the two meanings from the same wording in his analysis.

[3] This is misleading, because it misrepresents how Could you read that again? would be analysed using IFG (Halliday 1994). At 'this level of analysis' — semantics — this instance would be analysed as a command (the SPEECH FUNCTION features 'demand' and 'goods-&-services'), whereas 'polar interrogative' (MOOD) would be the analysis at the level of lexicogrammar. It is the mismatch between meaning and wording that constitutes the grammatical metaphor: a command is congruently realised by imperative mood, whereas interrogative mood is the congruent realisation of a question.

[4] This is misleading, because it underplays the significance of Halliday's analysis, which shows both the metaphorical wording (MOOD (1)) and the congruent wording (MOOD (2)), thereby also demonstrating the systemic agnation of imperatives and modulated declaratives. (Halliday 1994: 378):

[5] To be clear, the Cardiff Grammar analysis merely relabels 'imperative' MOOD as 'directive' and 'modulation' as 'requirement' and provides no explanation as to why such features should be realised as a declarative clause instead of an imperative. (The systems from which these features derive are not provided.)

Fawcett's Summary Of His Answers To His First Two Difficult Questions For The Cardiff Grammar

Fawcett (2010: 150):

We can summarise this section by answering Questions la to lc with respect to the Cardiff Grammar. Firstly, in a full account of the analysis or generation of a text, we need to show representations at the levels of both meaning and form. Secondly, the multifunctional nature of language is most appropriately shown at the level of meaning. Thirdly, conflation occurs between individual elements, as exemplified in the relations between the Subject and Agent in Figure 10, and not between clause-length structures.

But what about Question 2? How complete are the Cardiff Grammar's descriptions of English (or any other language) at each of these two levels? And how available are they for general use in analysing texts? The first part of the answer is that the description and computer implementation of the lexicogrammar of English is very large indeed. Moreover, it has the advantage over other SF models that its system networks have been developed to function explicitly at the level of semantics, so that there is no need for a further level of analysis, e.g., as suggested in Halliday & Matthiessen (1999).

Blogger Comments:

[1] To be clear, the Cardiff Grammar's levels of meaning and form are inconsistent with SFL Theory, which operates with three levels of language: meaning, wording and sounding/writing.

[2] To be clear, the Cardiff Grammar violates this principle by including "the multifunctional nature of language" at the level of form (Agent and Affected are experiential, Subject and Complement are interpersonal).

[3] This is misleading. As previously seen, Fawcett has repeatedly misrepresented Halliday's version of his own theory as involving the conflation of structures, rather than elements.

[4] This is misleading. The size of the Cardiff Grammar's description of the lexicogrammar of English, even if substantiated, is not a measure of the completeness of its description — just as the size of a beheaded statue is not a measure of its completeness.

[5] To be clear, these systems do not appear in this publication, and the only system that does, Figure 1 of Appendix A, violates the principles of a system network, as previously demonstrated. Moreover, the promised work:

• Fawcett, Robin, forthcoming a. Functional Semantics Handbook: Analysing English at the Level of Meaning. London: Continuum

remains unpublished 20 years after the first edition of this publication (2000).

[6] To be clear, this is a non-sequitur. Even if Fawcett has developed semantic networks, this does not entail that the model of semantics in Halliday & Matthiessen (1999) is unnecessary.

[7] To be clear, the model of semantics in Halliday & Matthiessen (1999) provides the means of accounting for grammatical metaphor in a systematic fashion, and grammatical metaphor is the major means by which meaning potential is expanded.

Inferring Meaning From Syntax

Fawcett (2010: 149-50, 150n):

The representation of the syntax is sufficiently rich for all aspects of its multifunctional meaning to be shown in it — or to be directly inferable from it. Thus a text analyst who is equipped with an appropriate handbook can work out, from the evidence of the items and the functional structure, what the associated semantic features are. For example, even though the structure does not show that there is a 'Theme' that is conflated with the Subject and Agent, we can infer directly from the structural analysis that the item We is the type of Theme called here a "Subject Theme". (If it had been an 'empty Subject', as in It's nice to see you, it would not have a Participant Role conflated with it.)

In a computer model of text analysis there is a broadly equivalent component (called the 'semantic interpreter'). In a similar way this reconstructs, on the evidence of the analysis of the functional syntax, the features that must have been chosen in generating that particular structure.²⁴

²⁴ See O'Donoghue (1994) for a description of the semantic interpreter that has been developed for the Cardiff Grammar, its task being to turn a syntactically analysed clause into its set of selection expressions.

Blogger Comments:

[1] Again, this demonstrates that the Cardiff Grammar's theoretical orientation is the opposite of Systemic Functional Linguistic Theory, since it proceeds from form to meaning, rather than from meaning to form. Halliday (1985, 1994: xiv):

[2] Fawcett's Functional Syntax Handbook (forthcoming 2011a) is still unpublished, the latest due date being November 2021. Fawcett's Functional Semantics Handbook (forthcoming 2011b) is still unpublished.

How The Cardiff Grammar Operates In Its Full, Generative Version

Fawcett (2010: 149):

Let us now look for a moment at how the Cardiff Grammar operates in its full, generative version. The first stage is the selection of the features in the system network, i.e., the creation of a selection expression. The 'key' features that have been selected in the present example are displayed in the lower half of Figure 10. The realisation rules then integrate the various partial 'strands of meaning' that are represented by these features into a single functional structure (e.g., as shown in the top half of Figure 10). This involves the conflation of some elements with each other and with participant roles, such as the conflation of the Agent ("Ag") with the Subject ("S"). Notice that, in contrast with the 'structure conflation' version of the Sydney Grammar, there is no expectation that every element — or almost every element — will be conflated with an element in another strand of meaning.

Blogger Comments:

[1] To be clear, as the wording above demonstrates, the Cardiff Grammar is not a model of language, but a sequenced set of procedures for computer-mediated text generation, as represented in Figure 4 (p36):

[2] To be clear, despite this being "how the Cardiff Grammar operates in its full generative version", Fawcett does not provide either (a) the system networks of features from which selection expressions are derived or (b) the realisation rules that specify precisely how paradigmatic semantic features are realised syntagmatically at the level of form. Figure 10 (p148) merely presents an end product to be taken on trust:

[3] This is misleading, because it is the opposite of what is true, and as such, misrepresents Halliday's version of his own theory ("the Sydney Grammar"). As previously explained, there is no "structure conflation" in SFL Theory, not least because structure conflation is nonsensical in a theory that views structure as the relation between elements. As previously explained, Fawcett's misunderstanding derives from his confusion of clause constituents (the units of the rank scale) with elements of clause function structures, and from his view of structure as a sequence of isolated elements.