Fawcett Adjusting His Model To Accomodate His Misunderstanding Of Halliday

 Fawcett (2010: 125):

However, the fact is that there is no description in the literature of the Sydney Grammar of the way in which the conflation of the various structures would actually be achieved. Let us therefore now try to fill this gap. 

Clearly, the first requirement is that the architecture of the model described in Figure 4 of Chapter 3 should be considerably extended. It would need to generate a new type of output that integrated the various structures, and for this it would require a new component that would transform the second 'multiple structure' output into the third 'integrated' one. Thus the relatively simple model in Figure 4 in Section 3.2 of Chapter 3 would have to be modified in the way shown in Figure 8, i.e., to give it two 'levels' within 'form'.

Blogger Comments:

[1] This is still misleading, because it is still untrue. To be clear, this has been described, it is merely the case that Fawcett does not understand that the three metafunctional clause structures are integrated by the syntagm of group/phrase units that realises them.

[2] This is misleading, because it misrepresents Fawcett's model as the architecture of SFL theory. Here Fawcett is simply modifying his own model (Figure 4) to accommodate his own misunderstanding of SFL Theory, as a way of arguing against Halliday's model.

Misrepresenting The Architecture Of SFL Theory As Modular

Fawcett (2010: 124-5):

In the 'structure conflation' model several sets of structures, one corresponding to each metafunction (or more accurately each 'strand of meaning') are first generated as separate structures, and then, by an additional stage in the process of generation which has not so far been described by any SF linguist — they are mapped onto each other to form a single, integrated structure. Halliday is clearly assuming some such model as this when he writes, as in the passage cited above, that "it is the function of the lexicogrammar to map the structures onto each other". And Matthiessen strikes a similar note when he talks of "the structural unification of the metafunctional strands" (1995:613). Indeed, the use of 'multiple structure' diagrams throughout IFG can be said to presuppose the 'structure conflation' model. 

Thus a model of language that posits a set of several different structures for a single clause must also provide a way to integrate them into a single output. As Halliday says in his introduction to IFG, "although each strand of meaning in the clause will be described independently in its own terms, [...] a clause is still one clause — it is not three" (1994:36). It is clear, then, that the next stage for the representations in IFG is that they should be 'conflated'.

Blogger Comments:

[1] This is misleading, because it is untrue. There is no temporal sequencing of this sort in the architecture of SFL Theory, since it is relational in its organisation, not modular.  That is, the theoretical architecture maps the relations between all the options, across systems, ranks and strata, and the process of instantiation is the selection of a configuration of such relations.

Fawcett's model (Figure 4), on the other hand, is modular: it comprises a sequence of modules where one module produces an output as the input for the next module. Here again Fawcett is misrepresenting his own model as Halliday's.

[2] This is misleading, because it is untrue. To be clear, this has been described, it is merely the case that Fawcett does not understand that the three metafunctional clause structures are integrated by the syntagm of group/phrase units that realises them.

[3] This is misleading, because it is untrue. As previously noted, 'structure' conflation' is Fawcett's misunderstanding of the integration of metafunctional clause structures in their realisation as a syntagm of units of the lower rank.

Fawcett's Claim That There Are Two Versions Of Halliday's Grammar

Fawcett (2010: 124):

The first part of the answer is that there appear to be two versions of the Sydney Grammar that differ from each other on this matter. Roughly speaking, the two seem to correspond to each of the two 'strands' of the theory that I characterised in Section 5.2 of Chapter 5 as the 'text-descriptive' and the 'theoretical-generative'. When Halliday or Matthiessen are writing from a 'text-descriptive' viewpoint, they typically write in terms of what we shall call the structure conflation model, but when they are writing from the 'theoretical-generative' perspective they tend to write in terms of what we shall call the element conflation model.

Blogger Comments:

[1] This is misleading, because it is untrue. There is only one version in this regard. What Fawcett mistakes for two versions are two aspects of the one theory.

On the one hand, what Fawcett calls 'the structure conflation model' is actually the integration of metafunctional clause structure in the syntagm of group/phrase rank units that realise it; for example, in the clause Ian Curtis was Joy Division's singer, the experiential structure Token ^ Process ^ Value is realised by the syntagm nominal group ^ verbal group ^ nominal group.  Structures themselves cannot be conflated because, as previously noted, a structure is actually the relation between elements.

On the other hand, what Fawcett calls 'the element conflation model' is actually the case of two or more functions being realised by the same unit of a syntagm; for example, in the clause Ian Curtis was Joy Division's lyricist, the nominal group Ian Curtis realises the conflation of Theme, Subject and Token.

[2] For the confusions involved in Fawcett's distinction between 'text-descriptive' and 'theoretical-generative', see the original post here.

On The 'Supposed' Version Of Halliday's Grammar

Fawcett (2010: 123-4):

We shall begin by asking how the representations of structure found in an IFG-style analysis fit into the theoretical-generative model of language that is used by Halliday and others. Specifically, we shall make explicit the implications of taking the position that the several strands of functional structure found in such representations are part of the generative grammar. In other words, we shall make a leap of faith (but only for a couple of pages) and assume that somewhere there is (or will be in the future) a systemic functional grammar that is capable of generating structures such as those in Figure 7 of Section 7.27 

Adopting such an approach requires, as we shall see, the addition of a new component to the overall model of language and a new 'level' — or perhaps we should say 'stage' — in the representation of instances. The three stages of representation in such a model would be as follows: 

1. the selection expressions for each unit (such as a clause), i.e., the features that display the unit's meaning potential (which could if we wished be displayed according to the clause's strands of meaning, as in the lower half of Figure 10 in Section 7.8), 2a. a set of several different structures, each corresponding to a major strand of meaning, and each consisting of a number of 'functions' (e.g., as in Figure 7 of Section 7.2, and as currently in IFG), 2b. the single structure into which these must finally be integrated (e.g., as in the last line of Figure 6 in Section 4.9 of Chapter 4, or in the upper half of Figure 10 of Section 7.8). 

Thus the first type of representation corresponds to the 'selection expression' of features chosen in the system network, i.e., to the instance at the level of meaning in Figure 4 in Section 3.2 of Chapter 3. However, each of the second and third types can be seen as corresponding, in their way, to the instance at the level of form in terms of Figure 4 in Chapter 3 (which is why they are labelled "2a" and "2b", rather than "2" and "3"). 

But what is the relationship of the last two to each other? Does the Sydney Grammar first generate a set of several different structures for each clause, one for each line of analysis in Figure 7 of Section 7.2, and does it then bring into operation a set of 'mapping rules' which match up all of these structures and then integrate them into a single structure? This is the clear implication of Halliday's statement (cited earlier in Section 4.9 of Chapter 4) that 

it is the function of the lexicogrammar to map the structures onto one another so as to form a single integrated structure that represents all components [= 'components of the grammar, in the sense of 'metafunctions'] simultaneously. (1977/78:128) 

If this was indeed how the Sydney Grammar worked, it would require another major component in the grammar itself to perform this complex task of 'mapping'. But is it?

Blogger Comments:

[1] This is misleading.  What Fawcett actually does is ask how (his misunderstanding of) Halliday's model of clause structure fits his own model (Figure 4), which, as previously demonstrated, is invalidated by its own internal inconsistencies.

[2] This is misleading. Grammatical systems that specify clause structures appear in Matthiessen (1995) and Halliday & Matthiessen (2004, 2014).

[3] This is misleading. Here Fawcett misrepresents his own model (Figure 4) as Halliday's model, and adjusts it so as to generate metafunctional clause structures.

[4] To be clear, on Halliday's model, the three metafunctional structures of the clause are integrated by the syntagm of units at the rank below — such as nominal group ^ verbal group ^ prepositional phrase — that realises (represents) the elements of clause structure.

[5] This is misleading, since neither Figure 6 nor Figure 10 is consistent with the model of structure in Halliday (1994). Figure 6 is from Halliday (1970) and represents a superseded stage of the model (with the word 'combined' strategically added by Fawcett):

and Figure 10 is Fawcett's model, not Halliday's:

On The 'Supposed' Vs The 'Implemented' Versions Of The Sydney Grammar

Fawcett (2010: 122-3):

We turn now to Question 1c. This was: 

1c. In the representation at the level of form, is the conflation that occurs between the realisations of the various strands of meaning a conflation of whole structures or a conflation of individual elements? 

Since this question concerns an issue that is internal to the representation of structure at the level of form, it might appear less important than the ones that we have just been considering. But it is in fact the key question for our present purposes, because the answer to it shows why it is necessary to question the status of the representations of functional structure in IFG. 

The reason for asking Question 1c is that we need a clear answer on the issue of precisely what the syntactic phenomenon is that actually gets conflated — i.e., is it whole, clause-length structures, or is it single elements of the clause? As we shall see, the theoretical status of the 'multi-strand' representations in IFG depends upon the answer.

To discover this answer we must look not in IFG but in the accounts by Halliday and others of work in the theoretical-generative strand of the theory. And we shall discover that, within the Sydney Grammar version of a SF grammar, there are two conflicting pictures of how the grammar works — and that only one of these has a sound theoretical base, so making it implementable in the computer. In what follows, therefore, I shall refer to the 'supposed' and the 'implemented' versions of the Sydney Grammar. 

Blogger Comments:

[1] This is misleading, because 'the level of form' is a component of Fawcett's model only. In SFL theory, 'levels' can refer to strata or rank, and Fawcett's 'level of form' is theorised as the stratum of lexicogrammar, within which, lexicogrammatical form is theorised as a rank scale of units — clause, group/phrase, word, morpheme — each of which is the entry condition for a system of functions.

In SFL theory, a syntagm of units at a lower rank  — e.g group/phrase — is the means of integrating the three function structures of the rank above — for the clause: those of theme, mood and transitivity.  When one of those units, e.g. a nominal group, realises Theme, Subject and Actor, those three functions are said to be 'conflated'.  It is not the structures that are conflated, because a structure is actually the relations between the elements. Halliday & Matthiessen (2014: 83-4):

The significance of any functional label lies in its relationship to the other functions with which it is structurally associated. It is the structure as a whole, the total configuration of functions, that construes, or realises, the meaning. The function Actor, for example, is interpretable only in its relation to other functions of the same kind – other representational functions such as Process and Goal. So, if we interpret the nominal group I as Actor in I caught the first ball, this is meaningful only because at the same time we interpret the verbal group caught as Process and the nominal group the first ball as Goal. It is the relation among all these that constitutes the structure.

[2] This is misleading, because it is untrue.  In Halliday's theory, there is only one model of "how the grammar works", and Fawcett's reasons for believing otherwise are based on his own misunderstandings of the theory — such as those aspects of the theory described above in [1] — as will be seen in the unfolding of the posts that follow.

On How Best To Represent "The Multifunctional Nature Of Language"

Fawcett (2010: 122):

So far in Section 7.3 we have so far been concerned with the question of whether a representation of a text by a single (though multifunctional) level of analysis is adequate. And the answer has been that we should, in principle, represent a text in terms of its systemic features, because, in Halliday's words, "the system has priority". And I have suggested that it is preferable to show the multifunctional nature of language at the level of meaning, since it is in fact the "strands of meaning" (IFG p.34) that we wish to model. 

Blogger Comments:

[1] To be clear, by "representation of a text", Fawcett means a representation of an analysis of clause structure in terms of the three metafunctions.

In terms of SFL Theory, "a single level of analysis" is an analysis of language in terms of the lexicogrammatical stratum only, without regard to semantics (or phonology/graphology). Such an analysis of the content plane is adequate in the absence of grammatical metaphor, because, in such circumstances, the semantics and lexicogrammar are in agreement (congruently related).

[2] To be clear, even though, in SFL Theory, "the system takes priority", Fawcett's perspective is firmly focused on structure, even to extent of representing systemic features as elements of structure (Figure 10), as previously demonstrated.

[3] As previously noted, this contradicts Fawcett's earlier claim that two elements at his level of form, Complement and Adjunct, are classed as experiential.

[4] To be clear, in SFL Theory, 'the three distinct kinds of meaning that are embodied in the structure of a clause' (Halliday 1994: 34) constitute an analysis at the level of wording (lexicogrammar), not meaning (semantics). This is because grammatical forms are viewed 'from above', and so are identified in terms of the meanings they realise, such as nominal group serving as Subject, etc. Halliday & Matthiessen (2014: 49):

Being a ‘functional grammar’ means that priority is given to the view ‘from above’; that is, grammar is seen as a resource for making meaning – it is a semanticky kind of grammar. But the focus of attention is still on the grammar itself.

On Fawcett's "Fact" That The Multifunctionality Of Language Lies At The Level Of Meaning Rather Than Form

Fawcett (2010: 121-2):

We shall now turn briefly to Question lb, i.e., "Should the representation show the multifunctional nature of language?" The answer to this question goes a long way to explaining why the representations in IFG are as they are — and it is as follows. Given (1) that the multifunctional nature of language is a central principle of the systemic functional model of language, and (2) that the representation of functional structure is the only type of representation that has so far been developed to the point where it can be made generally available through publication (for use in describing texts, etc), it is useful that it should show, if it is possible to do so without misrepresenting the model, the multifunctional nature of text. 

However, as we shall see in Section 7.8, the picture of what is both possible and desirable changes when we have the option of showing the multifunctional nature of language in the representation of the meaning potential, i.e., at the level of meaning. The question then becomes: "If the multifunctional nature of language is shown in the representation at the level of meaning, is it also necessary — or even appropriate — to show it at the level of form too?" My answer is that the multifunctionality of language lies at the level of meaning rather than form, and that a representation such as that in Figure 10 in Section 7.8 reflects this fact more accurately than the IFG-style diagram in Figure 7.

Blogger Comments:

[1] This is misleading, because it misrepresents a question about levels in Fawcett's model, meaning and form, as a question about levels in Halliday's model (semantics, lexicogrammar and phonology).

[2] To be clear, on the one hand this contradicts Fawcett's previously stated view that, in his own model, the forms Complement and Adjunct (see Figure 10 below) are experiential elements, in terms of metafunction:

In contrast, the concepts of 'Complement' and 'Adjunct' — but not 'Predicator' — have a central role in the Cardiff Grammar. Here, a Complement is an 'experiential' element of the clause that is 'predicted' by the Process (i.e., a Participant Role), while an Adjunct is one that is not (i.e., a Circumstantial Role) — or an element expressing a different type of meaning.

On the other hand, it denies the metafunctional distinctions of the tone group in phonology: the interpersonal system of TONE versus the textual systems of TONALITY and TONICITY.

[3] To be clear, a representation such as that in Figure 10 in Section 7.8 (p148):

misrepresents what Fawcett takes to be (paradigmatic) features of semantic systems (e.g. repeated past, periodic frequency, positive, unassessed) as elements of (syntagmatic) function structure. Moreover, some of what Fawcett takes to be (paradigmatic) features of semantic systems (e.g. agent, subject theme, unmarked new) are, instead, elements of (syntagmatic) function structure.

Misconstruing Paradigmatic Selection Expressions And Syntagmatic Structures As Two Levels Of Instances

Fawcett (2010: 121):

Indeed, when Halliday is writing in theoretical-generative terms, he always writes with the assumption that there will also be a representation in terms of the systemic features. He cannot but do so, because if there is no selection expression of features there can be no input to the realisation rules (as we saw in Chapters 3 and 5). However, for text-descriptive purposes such as those for which IFG is intended, Halliday seems to take a different position. This is that, while it may in principle be desirable to represent the systemic features as well as the functional structures, it is either not possible or not necessary to provide this level of description. Section 7.8 will discuss the question of when it will become possible to provide a systemic description of English.

The implications of the argument so far for the representations given in IFG is that their status is somewhat diminished in importance. This is because they represent the grammatical structure of language — even though, to cite Halliday's words once again, "in systemic theory the system takes priority".

Blogger Comments:

[1] For the confusions involved in Fawcett's distinction between 'theoretical-generative' and 'text-descriptive', see the previous post here.

[2] To be clear, a representation of a clause in terms of systemic features is an identification of the paradigmatic features that classify that unit as potential or instance. In terms of the architecture of SFL Theory, this is distinct from a representation of a clause as a syntagmatic structure (that realises paradigmatic selections). As previously noted, once or twice, Fawcett's model (Figure 4) misconstrues structures as instances.

[3] This is misleading, because it misrepresents Fawcett's model as Halliday's model. It is only in Fawcett's model (Figure 4) that selection expressions (misunderstood as meaning instances) are the input for realisation rules (misconstrued as form potential).

[4] To be clear, Matthiessen's Lexicogrammatical Cartography: English Systems (1995) already provides a (necessarily partial) systemic description of English.

[5] This is both misleading and a non-sequitur. Even ignoring all the theoretical misunderstandings in Fawcett's argument, it does not follow that the structural representations in Halliday (1994) are "somewhat diminished in importance" merely because system is given priority over structure in theorising language. The "importance" of the structural representations in IFG is that they are most directly related to text analysis, the purpose for which the book was written (Halliday 1994: x, xxvii).

Misrepresenting Fawcett's Views As Halliday's

Fawcett (2010: 120-1):

To summarise so far: Halliday recognises that texts should in principle be represented in terms of their features in the 'meaning potential' as well as their functional structures. Indeed, he states that "the system takes priority". However, the fact that IFG is written "specifically for those who are studying grammar for text analysis purposes" ((1994:x) has led him to focus on representations of texts that display the functional structure rather than the more abstract systemic features, one reason being that it is easier to understand such analyses. The answer to Question la is therefore that Halliday, like me, believes that in a SF model of language a text should, in principle, be analysed in terms of both its meaning potential and its structure — this being the model summarised in Figure 4 of Section 3.2 of Chapter 3 (or perhaps for Halliday the topological variant in Figure 5 in Section 4.7 of Chapter 4, with its unsatisfactory representation of 'instantiation' and 'realisation').

Blogger Comments:

[1] To be clear, for Halliday, 'meaning potential' is language as system, whereas as text is language as instance.

[2] As previously pointed out, Halliday presented function structures in IFG, not because they are easier to understand, but because they are more directly related to text analysis (Halliday 1994: x, xxvii).

[3] This is misleading, because it misrepresents Fawcett's understanding as Halliday's. For Halliday, 'meaning potential' is language as system, not language as text.

[4] This is misleading, because it misrepresents the architecture of Fawcett's model (Figure 4) as the architecture of Halliday's theory. For the original critique of Figure 4, see here.

[5] This is misleading, because it misrepresents Fawcett's reworking of his own model (Figure 5) as a possible architecture of Halliday's theory. Moreover, as previously demonstrated, Figure 5 is not a topological variant of Figure 4, and the architecture of both models, Figures 4 and 5, is invalidated by Fawcett's own misunderstandings of instantiation and realisation. For the original critique of Figure 5, see here.

On The Diagrammatic Representation Of Systemic Analyses

 Fawcett (2010: 120):

There are in fact three further practical reasons why IFG does not provide the reader with systemic analyses. The first is that if Halliday had included the systemic analyses too — and so also the system networks and realisation statements that would make that possible — the book would have had to be two to three times as long (Halliday 1994:xv). The second reason is that Halliday has never published full sets of his current system networks, and it would have been odd if their first public appearance had been in what was intended as a text book. As we saw in Section 4.5 of Chapter 4, it was only through the publication of Matthiessen (1995) that the system networks which accompany IFG were finally published. And the third reason is that it is not obvious how best to represent, in a readily interpretable diagram, the analysis of a clause in terms of its features. A simple list of all the features that have been chosen is clearly inadequate, because it may include the names of thirty or forty features, not all of which are immediately interpretable. There have in fact been very few published attempts — in any version of the theory — to develop any such method of displaying the systemic features of a text. The diagrams in Fawcett (1980: 195 and 231) are early forerunners of the method of presentation currently used in the Cardiff framework, an example of which is presented in Section 7.8. So far as I know, Fawcett (forthcoming a) will be the first work to introduce a method of analysing texts in terms of their systemic features. 

Blogger Comments:

[1] This is misleading. To be clear, the first two editions of IFG were not intended as a textbook.  Halliday (1994: xxvi):

This book is not a textbook of English; it is an interpretation of the English code. No attempt is made to 'teach' the categories. But an attempt is made to interpret some of them, especially the difficult and important ones like Subject.

[2] To be clear, the obvious way to present a systemic — not structural — analysis of a clause is to identify the features that were selected in its instantiation.  A list of features is not inadequate because all of the features are immediately interpretable, because the meaning (valeur) of each feature is its relation to other features in the system. Halliday (1994: xxvii):

A systemic grammar is not syntagmatic but paradigmatic; hence there is no difference in describing something and relating it to everything else, because the description of any feature is its relationship to all the others.

[3] To be clear, Fawcett's method of presentation in Section 7.8 — Figure 10 (p148) — confuses systemic features (paradigmatic axis), such as 'repeated past', 'positive' and 'unassessed', with structural elements (syntagmatic axis), such as 'overt agent', 'overt affected', 'subject theme' and unmarked new':

This confusion of the paradigmatic and syntagmatic axes is inherent in Fawcett's model — Figure 4 (p36 ) — where selected features (paradigmatic axis) are misconstrued as the 'meaning level' counterpart of 'form level' structure (syntagmatic axis):

 

For the internal inconsistencies in Fawcett' model, see:

On 'The Main Components Of A Systemic Functional Grammar'.

Misrepresenting Halliday (1994) On Structural Representations

Fawcett (2010: 119-20):

The recognition of this situation leads us to ask a new question. This is: "Since the grammar is is systemic as well as functional, why doesn't Halliday show representations of the features in IFG?' Halliday gives a partial answer to this question when he writes in the "Foreword" to IFG that, because he wrote this book "specially for those who are studying grammar for text analysis purposes, I did not include the systemic part" (1994:x). Later he explains that "structures are less abstract; they are so to speak 'nearer the text'" (1994:xxvii). In other words, the clear implication is that Halliday decided that IFG should present the structural representations of texts rather than the systemic representations, on the grounds that the structural representations are likely to be easier for the reader to understand. Notice, however, that here Halliday is describing a strategy for overcoming the difficulties of communicating the theory's concepts for a particular class of readers, not making a statement about the status in the theory of the structural representations shown in IFG.

Blogger Comments:

[1] To be clear, the presentation of structural representations in Halliday (1994) was not motivated by any supposed deficiency in the intellectual ability on the part of those who analyse texts ("easier for a particular class of readers to understand").  As Halliday (1994: x, xxvii) points out, structures are more directly related to the analysis of text:

The reason the present work is not called an introduction to systemic grammar is that that is not what it is. Since it was being written specifically for those who are studying grammar for purposes of text analysis, I did not include the systemic part: that is, the system networks and realisation statements, which constitute the main theoretical component (and would be central if the book was an introduction to systemic grammar). What is presented here is the functional part: that is, the interpretation of the grammatical patterns in terms of configurations of functions. These are more directly related to the analysis of text. …

The reason for using structural rather than systemic representations for discourse analysis is that structures are less abstract; they are so to speak 'nearer' the text. The most direct move in the analysis of a text is to give it a structural representation, and this is what is done here. All the structural analyses could be interpreted in terms of the features selected.

[2] To be clear, here Fawcett misleads through selective omission. On the second of the two cited pages, Halliday does indeed make "a statement about the status in the theory of the structural representations shown in IFG — one that is inconsistent with the Fawcett model (Figure 4) that gives equal weight to system ("meaning potential") and structure ("instances of form"). Halliday (1994: xxvii):

This is not an account of systemic theory, nor does it present the system networks for English grammar (there is a brief foray into network representations at the end of Chapter 10). It presents the structures which are the 'output' of the networks — which collectively realise the sets of features that can be chosen. But it is not a 'structural' grammar (still less a 'structuralist' grammar in the American sense). Such grammars are syntagmatic, having structure as their main organising concept, and bringing in special devices to relate one structure to another. A systemic grammar is not syntagmatic but paradigmatic; hence there is no difference in describing something and relating it to everything else, because the description of any feature is its relationship to all the others. Obviously we have to describe one part of the grammar at a time; but it is important to think of every section as being part of the network as a whole.

Misrepresenting Halliday On "Representations At The Levels Of Meaning And Form"

Fawcett (2010: 119):

It therefore seems that Halliday's answer to Question la above is, in principle, that there should be representations at the levels of both meaning and form. Moreover, this principle should clearly be applied to the analysis of texts as well as to the generation of texts. Indeed, the concept of 'priority to the system' is precisely what is implied by the fact that the name of the theory is "systemic functional grammar" — and not "structural functional grammar". So what, one might reasonably ask, does Halliday consider that the systemic representation of a clause would be like? The answer is that we do not know. 

Perhaps the most surprising fact of all about Systemic Functional Grammar as a theory of language is that that there is no generally agreed way of showing, in the representation of the analysis of a text, the features in the system network that have been chosen in generating it. What makes this fact even stranger is that all of the SF linguists who engage in the 'theoretical-generative' task of building such grammars recognise that it is not merely desirable to have a representation of the features; it is necessary. This is because the 'selection expression' is an integral part of the operation of the grammar, as we saw in Chapter 3 and as is demonstrated in Appendix A.

Blogger Comments:

Reminder of Fawcett's Question 1a (p117):

1a. Is it desirable — or indeed necessary — to have representations of a text at the levels of both form and meaning? 

[1] To be clear, the levels of meaning and form feature only in Fawcett's model; Halliday's model postulates the strata of semantics, lexicogrammar and phonology. So any suggestion of Halliday endorsing Fawcett's model of meaning and form is misleading.

[2] To be clear, one way to provide a systemic representation of a clause is to identify the systemic features that specify its structural realisation.  Halliday and Matthiessen (2004) provide examples for analyses of:

• theme (pp101-3), 
• mood (pp158-9), and
• transitivity (pp304-5).

[3] To be clear, 'selection expression' is "an integral part of the operation of the grammar" in Fawcett's model (p36):

However, as can be seen above, in Fawcett's flowchart, a selection expression is misunderstood as an instance of a system and restricted to only one level: meaning.  In SFL Theory, a selection expression is constituted by all the features that identify a given unit, of a given stratum, whether as potential or instance. This is most clearly demonstrated by the phonological selection expression [voiced, bilabial, stop] which identifies the phoneme /b/ as both potential and instance.

Misrepresenting The Cardiff Flowchart As Consistent With The Architecture Of SFL

Fawcett (2010: 118-9):

However, Halliday's words in "Systemic theory" go even further, clearly implying that it is not only possible to provide this second level of representation, but that it is also desirable. He states that 

in systemic theory the system takes priority [my emphasis]; the most abstract representation [...] is in paradigmatic terms. [...] Syntagmatic organisation is interpreted as the 'realisation' of paradigmatic features, the 'meaning potential'. (Halliday 1993:4505) 

And a few paragraphs later he foregrounds the importance of the representation of the systemic 'meaning potential' more strongly still, writing that 

the selection expression constitutes the grammar's description of the item [e.g., a clause]. (Halliday 1993:4505) 

This is heady stuff. However, the ambitious view of the essential nature of a semiotic system expressed in the above quotations is shared by all of those who work in the framework of the Cardiff Grammar, and by many other systemic functional linguists. There are two important consequences that follow directly from accepting this position for our present concerns. Firstly, Halliday's words suggest that the diagram of language shown in Figure 4 in Section 3.2 of Chapter 3 can indeed stand as a summary of the framework that is common to both the Sydney and the Cardiff Grammars. Secondly, they show that Halliday believes that, if we are to make a full analysis of a clause, we must give equality — and indeed "priority" — to the features that have been selected in generating it — i.e., to the features that specify its 'meaning potential'. All this is, we should note, is what should happen "in principle"; we shall come shortly to what is currently done in practice. 

Blogger Comments:

[1] To be clear, the notion that the system takes priority is not a "view of the essential nature of a semiotic system".  It is a statement of an underlying principle on which Systemic Functional Linguistics is theorised.

[2] This is very misleading.  On the one hand, it is simply a non-sequitur, since giving priority to the system does not logically entail that Figure 4 "can indeed stand as a summary of the framework that is common to both the Sydney and the Cardiff Grammars".  On the other hand, Figure 4 cannot "stand as a summary of the framework that is common to both the Sydney and the Cardiff Grammars" because it is inconsistent with the architecture of SFL Theory and invalidated by its own internal inconsistencies, as previously explained (e.g. here).  Reminder:

[3] As previously explained, Fawcett misunderstands Halliday's 'meaning potential' (language as system) to be networks at the level of semantics.

Confusing Systemic Features With Structural Elements

Fawcett (2010: 118):

Let us begin with Halliday's own description of the status of the analyses offered in IFG. He writes: 

This book [...] presents the structures which are the 'output' of the networks — which collectively realise the sets of features that can be chosen. (Halliday 1994:xxvii) 

So far so good; this is indeed what the analyses of texts in IFG does, in Halliday's terms. But a few lines below that he adds that 

all the structural analyses [given in IFG] could be reinterpreted in terms of the features selected [my emphasis]. (Halliday 1994:xxvii) 

The significant thing about these passages is that they show that Halliday recognises that there are, in principle, two types of representation for each clause. The first type corresponds to those in IFG, and these are presented to us as instances at the level of form. However, since these "realise the sets of features than can be chosen" (in the 'meaning potential'), it is clear that there is also, in principle, a second type of representation — one that is expressed in terms of the systemic features from which the structure is generated. It is a pity that Halliday does not give an example of what he thinks such representations might be like, but I shall give an example of one possible way of such showing such representations in Section 7.8.

Blogger Comments:

[1] To be clear, the two types of representation of a clause differ in terms of axis: paradigmatic system vs syntagmatic structure.   That is, one type represents the clause as features of systems, whereas the other represents the clause as structures.

[2] This is misleading, because it misrepresents Fawcett's model ("instances at the level of form") as Halliday's model.  Halliday (1994) presents structures at the stratum of lexicogrammar.

[3] As previously noted, Fawcett misunderstands Halliday's 'meaning potential' — language as system — as just the semantic system, or, in terms of his own model (Figure 4): potential at the level of meaning.  In Halliday's model, SFL Theory, the systems that specify grammatical structures are grammatical systems.

[4] Examples of what "such representations might be like" were given in Halliday & Matthiessen (2004) for clause rank textual feature selections (pp101-3), interpersonal feature selections (p169) and experiential feature selections (pp304-5).

[5] To be clear, the example representation that Fawcett provides in Section 7.8, Figure 10, presents his clause features as if they were (features of) structural elements of the clause. Fawcett (2010: 148):

Fawcett's Three Sub-Questions On The Status Of Representations In Halliday (1994)

Fawcett (2010: 117):

Now that we have examined the nature of representations such as that in Figure 7, we are ready to consider their theoretical status. We shall do this by breaking down the first question that we asked in Section 7.1 into three sub-questions: 

1a. Is it desirable — or indeed necessary — to have representations of a text at the levels of both form and meaning? 

1b. In a model with representations at the level of both form and meaning, is it desirable — or indeed necessary — to show explicitly the multifunctional nature of language at both levels? 

1c. In the representation at the level of form, is the conflation that occurs between the realisations of the various strands of meaning a conflation of whole structures or a conflation of individual elements? 

In answering these questions, we shall make two discoveries about the status of the representations of texts in IFG which are likely to surprise most readers. 

Blogger Comments:

[1] Reminder: Fawcett (2010: 117):

Let me begin by expressing our first question in relatively general terms (though we shall later divide it into three sub-questions): 

1. What is the status in systemic functional theory of the representations of the functional structure of text-sentences given throughout both IFG and the various associated texts?

[2] This is misleading, because it misrepresents levels of Fawcett's model, form and meaning, as levels in Halliday's SFL Theory, on which IFG (Halliday 1994) is based.  The linguistic levels (strata) in SFL Theory are semantics, lexicogrammar and phonology/graphology. Grammatical form is modelled as a rank scale of units, each of which provides the entry condition to systems of functional features.

[3] This is misleading.  What Fawcett terms the "conflation of whole structures" is the integration of the three metafunctional structures of the clause in a syntagm of clause constituents, such as nominal group ^ verbal group ^ prepositional phrase.

[4] In examining Fawcett's answers to these questions, we shall discover that Fawcett misleads by misrepresenting Halliday's theory, which is unlikely to surprise readers of this blog.

Misrepresenting The Integration Of Meanings Into A Single Syntagm As The Conflation Of Structures

Fawcett (2010: 116):

For our present purposes, however, the main point to notice is that a diagram such as that in Figure 7 embodies a very significant extension of the concept that a single element such as we in Figure 7 is the 'conflation' of three 'functions'. In such diagrams the whole clause is presented as the 'conflation' of five or more functional structures (as well as also involving 'more delicate' layers of analysis). In Sections 7.4 and 7.5, I shall challenge Halliday's concept that it is whole structures that are conflated with each other, rather than individual elements. Moreover, this central proposal of Halliday's is also challenged, as I shall show, by the actual practice in the theoretical-generative strand of work in the Sydney Grammar, i.e., in the computer implementations of Halliday's model by Mann, Matthiessen and Bateman — and even, surprisingly, by Halliday's own writings on generation.

Blogger Comments:

Reminder:

[1] This is misleading. To be clear, we is not an element of function structure.  As a clause constituent, the nominal group we realises the conflation of three clause functions: Theme, Subject and Actor.

[2] This is misleading. On the one hand, a clause involves only three types of function structure — theme, mood and transitivity — and the relation between the Mood and Residue blocks and their elements in one of composition, not delicacy; e.g. Subject and Finite are parts of the Mood element, not subtypes of it.

On the other hand, it confuses (functional) structure with (formal) syntagm. To be clear, the 'simultaneous structures' of the clause are integrated into a single syntagm — a syntagm being a string of clause constituents (different classes of group ± prepositional phrase). Halliday & Matthiessen (2014: 74):

The clause, as we said, is the mainspring of grammatical energy; it is the unit where meanings of different kinds, experiential, interpersonal and textual, are integrated into a single syntagm.

[3] This is misleading, because it misrepresents Fawcett's concept as Halliday's. That is, the concept that whole structures are conflated is Fawcett's, not Halliday's, which means that Fawcett's challenge will be to his own misunderstanding of Halliday's model, rather than to Halliday's model itself.  In terms of logical fallacies, this can be seen as a variant of the Straw Man argument.

Deploying Two Logical Fallacies In Misrepresenting Halliday (1994) On Thematic Structure

Fawcett (2010: 116): 

The two examples on pages 375-6 of IFG provide a final example of how complex Halliday's representations can become: each has ten lines of analysis. The reason is that Halliday shows three lines of analysis within the THEME line, on the grounds that there are three degrees of structural delicacy in them. However, it should be said that Halliday's analysis of 'multiple theme' in this way is one with which many other systemic functional grammarians disagree — as also do other grammarians (e.g., Huddleston 1988).

Blogger Comments:

[1] On the one hand, this is misleading, because the lines of analysis are not restricted to clause rank analysis. On the other hand, it deploys the red herring logical fallacy, since the number of represented analyses is irrelevant to the value of the analyses themselves.

[2] This is misleading, because it misrepresents Halliday's analysis. The theme analyses represent one line of meaning realised in clause structure, textual meaning, and present the Theme element as consisting of textual, interpersonal and topical (experiential) elements.

Halliday (1994: 375):

Halliday (1994: 376):

[3] This is misleading. The mere fact that people disagree with an analysis does not, in itself, constitute an argument against the analysis.  This constitutes a version of the logical fallacy known as argumentum ad populum.  Huddleston's arguments will be examined for validity when Fawcett presents them later in his publication.

The Reason Why Fawcett Has Been Repeatedly Misrepresenting Halliday (1994) On Clause Structure

Fawcett (2010: 115-6, 116n):

It may come as a surprise to some readers that an IFG-style analysis in fact requires seven separate lines of representation (six of these being structural). This is because the writings of the Sydney grammarians regularly present the view that there are either three or at most four strands of meaning, each corresponding to one of the metafunctions. In Chapter 2 of IFG, for example, Halliday introduces the concept that a clause has multiple structures under the beguilingly simple section heading of 'Three lines of meaning in the clause" (p. 33). Yet, the fact is — as I have just demonstrated — that a analysis of a sentence in IFG terms regularly requires seven lines of analysis (and sometimes even more, as we shall see shortly).  

In IFG, therefore, the task of analysing a text — and so the model of language that underlies it — is rather more complex than we are at first led to expect. Indeed, in the analyses of 'The 'silver' text" on pages 368-85 of IFG there are often eight or more lines of analysis, because Halliday adds an extra line for the analysis of the 'unpacked' interpretation of examples that contain cases of what he terms "grammatical metaphor".…⁶

⁶ I should point out that the seven or eight lines of structure found in an IFG-style analysis do not correspond to the eight major types of meaning that I consider to be present (if only by their non-realisation) in a typical clause. See Section 7.8 for a brief introduction to the eight strands of meaning that are recognised in the Cardiff Grammar.

Blogger Comments:

[1] This is misleading.  As previously demonstrated, in arguing about clause structure, Fawcett switches between clause, 'text-sentence' and text in order to deceive the reader.  In erroneously arriving at seven lines of representation for the clause, Fawcett counts interpersonal structure twice, includes both information and logical structure, despite the fact they are not clause structures, and includes cohesion despite the fact that it is not a structure of any grammatical unit.

[2] This is correct.  There are three lines of meaning realised in clause structure: textual, interpersonal and experiential.

[3] This is misleading.  A statement about the number of lines of meaning realised in the structure of one one rank unit, the clause, is not a statement about the complexity of carrying out a complete text analysis using all the dimensions of the theory.

[4] This is correct.  In the analysis of the 'silver' text, Halliday demonstrates how to deploy the theoretical resources outlined in the body of the book, including the three metafunctional structures of the clause, logical relations between clauses in complexes, information unit structure, and the unpacking of grammatical metaphor.

[5] To be clear, it is Fawcett's model of the clause — not Halliday's — that proposes eight lines of meaning.  This suggests that Fawcett's motivation for misrepresenting Halliday's model is to disguise the fact that his own model is less elegant.

Fawcett's Claim That Halliday's Analysis Of A Clause Typically Involves At Least Seven Lines Of Analysis

Fawcett (2010: 115):

We come finally to the seventh line of analysis shown in the standard analyses of texts in IFG . This shows textual 'cohesion', but as Halliday does not propose it as a type of structure we shall pay no further attention to it here. 

Halliday's analysis of a clause therefore typically involves at least seven lines of analysis. It is as if the variation in the structures of the various lines of analysis is seen as a phenomenon of language that is to be celebrated. But we must ask: "Is it a linguistic phenomenon rather than a metaphenomenon, i.e., a product of Halliday's version of the theory?". I shall argue in Section 7.4 (1) that it is indeed a metaphenomenon; (2) that this approach to representing the multi-functional nature of language brings with it enormous problems; and (3) that there is a preferable approach that achieves the same goals. 

Blogger Comments:

[1] This raises the question as to how, or if, the Cardiff Grammar integrates cohesion into its model.

[2] This is misleading.  Here Fawcett has switched back from his own terminology, 'text-sentence', on which his argument has been constructed, to the SFL notion of 'clause' and repeated his false claim that a clause "typically involves at least seven lines of analysis", instead of three.

To recap, in these seven lines of analysis, Fawcett includes:

• information structure, which is not a system of the clause, but of the informations unit,
• logical structure, which is not a system of the clause, but of the clause complex,
• cohesion, which is not a system of the clause, or any other grammatical unit,

and misrepresents the interpersonal structure of the clause as two lines of analysis, instead of one.

[3]  The fatuousness of this remark can be made self-evident by transposing its type to physics:

it is as if the number of different subatomic particles is a physical phenomenon that is to be celebrated.

[4] To be clear, any proposed linguistic structure can be nothing other than a construal ("product") of one theory of language or another.  Such construals are assessed in terms of their validity.

[5] It will be seen in the examination of Section 7.4 that the enormous problems that Fawcett identifies only arise, once again, from his own misunderstandings of Halliday's theory.

[6] To be clear, the approach that Fawcett finds preferable is his own, and whether or not it achieves the same goals will be examined when his approach is finally revealed.

On Predicator, Complement and Adjunct As Labels For Otherwise Empty "Boxes"

Fawcett (2010: 115):

It is interesting to note that the inclusion of the clause elements "Predicator", "Complement" and "Adjunct" as 'secondary' elements in the "Residue" is the last remaining trace in the descriptive framework found in IFG of a set of concepts that were central in the Scale and Category description of English. Indeed they, together with "Subject" were the primary elements of clause structure in "Categories". Interestingly, these elements are not referred to in the explication of "The 'silver' text" (IFG pp. 368-85). So what is their role?

In IFG, their role seems to be little more than a way of labelling another row of boxes that would otherwise remain empty, like the "Residue" (of which they are said to be the corresponding more delicate analysis) and the Rheme. But in Halliday (1970/76b) and (1977/78) they played a far more important role, as we saw in Section 4.9 of Chapter 4. 

Blogger Comments:

[1] To be clear, Predicator, Complement and Adjunct feature in the Mood structures in the analysis of the 'silver' text, and a Predicator is explicitly mentioned in the discussion of clause 7 (p377).  However, the discussion of the text generally takes a systemic perspective — e.g. MOOD selection — rather than a structural one.

[2] On the one hand, this misunderstands the notion of structure in SFL Theory. Halliday & Matthiessen (2014: 451):

A multivariate structure is a configuration of different functional relationships … . Note that, although it is the functions that are labelled, the structure actually consists of the relationships among them.

That is, the textual structure of the clause is the relation between Theme and Rheme; the interpersonal structure of the clause is the relation between Mood (elements) and Residue (elements).

On the other hand, the "labelling of boxes" is the assignment of clearly defined theoretical functional categories to the data. For example, the label 'Predicator', inter alia, identifies the element that specifies the Process that is predicated of the Subject (Halliday & Matthiessen 2014: 152-5). The  label 'Complement' specifies any element that was passed over for the choice of Subject (ibid.), and the label 'Adjunct' specifies any element that could not have been selected as Subject (ibid.). (Note that neither conjunctive nor modal Adjuncts are part of the Residue.)

[3] This is misleading, because it is untrue.  See the earlier post: Seriously Misrepresenting Halliday's Early Theorising.

On Complement And Adjunct As Experiential Elements In The Cardiff Grammar

Fawcett (2010: 115n):

5. In contrast, the concepts of 'Complement' and "Adjunct' — but not 'Predicator' — have a central role in the Cardiff Grammar. Here, a Complement is an 'experiential' element of the clause that is 'predicted' by the Process (i.e., a Participant Role), while an Adjunct is one that is not (i.e., a Circumstantial Role) — or an element expressing a different type of meaning. 

Blogger Comments:

To be clear, the notion of 'Complement' and 'Adjunct' as experiential meanings is inconsistent with the notion of experiential meaning, since neither is a construal of experience as meaning.  Halliday & Matthiessen (1999: 511):

Ideationally, the grammar is a theory of human experience; it is our interpretation of all that goes on around us, and also inside ourselves. There are two parts to this: one the representation of the processes themselves, which we refer to as the "experiential"; the other the representation of the relations between one process and another, and it is this that we refer to as the "logical". The two together constitute the "ideational" metafunction, whereby language construes our experiential world.

Moreover, as such, neither is structurally related to a Process. Halliday & Matthiessen (2014: 451):

A multivariate structure is a configuration of different functional relationships … . Note that, although it is the functions that are labelled, the structure actually consists of the relationships among them.

On The Unjustified Inclusion Of Predicator, Complement And Adjunct In Mood Structure

Fawcett (2010: 114-5, 115n):

Halliday's reasons for including this line of analysis as part of the interpersonal component are dependent on the fact that the Subject and Finite are involved in the expression of MOOD meanings. While I agree that these two elements contribute to the clause's interpersonal meaning (Fawcett 1999), I see no reason to include the other elements here. Halliday's stated reason for including Complements and Adjuncts in this line of analysis is that a Complement is an element that (typically) can become a Subject (though not in examples such as clever in Ivy is clever, as Halliday admits) and that an Adjunct is one that cannot. While this is largely true it does not seem a sufficient justification for including them in the analysis of every clause, and nor does it explain why the 'Predicator' is included there. In fact, this line of analysis does not explain anything about the particular text that is currently being analysed, since these elements patently do not carry the meaning associated with the Subject in the clause currently being analysed. (The Complement might, if the concept of the syntactic transformation were to be introduced to the theory, but Halliday would clearly not wish to do that.)

Blogger Comments:

[1] To be clear, the interpersonal structure of the clause, the clause as exchange, realises a proposition or a proposal.  In the case of propositions, the choice of indicative MOOD is realised by the presence of the Mood element, Subject and Finite, which realises the nub of the proposition, with the Residue elements — Predicator, Complement, circumstantial Adjunct — realising the remainder of the proposition.

[2] To be clear, these are not the reasons for including Complements and Adjuncts as elements of the interpersonal structure of the clause, but criteria for differentiating the two.

[3] To be clear, this is a bare assertion, unsupported by reasoned argument.

[4] To be clear, Halliday & Matthiessen (2014: 151-2) identify four functions of the Predicator, including:

It specifies the process (action, event, mental process, relation) that is predicated of the Subject.

Misrepresenting Halliday (1994) On Types Of Clause Structure

Fawcett (2010: 114):

However, there are in fact at least three other lines of representation in the analysis of a typical text-sentence in Halliday (1994:368-5), in addition to the four shown in Figure 7. The fifth is the analysis of the logico-semantic relationships between clauses that Halliday introduces to represent 'paratactic' or 'hypotactic' relations, as in I'll come when I'm ready (see Section 2.6.1 of Chapter 2.)

The sixth line of representation arises from the fact that Halliday always shows two lines of structure for the 'interpersonal' strand of meaning. At the primary degree of delicacy the structure is 'Mood + Residue'. Here I am using the term "primary" that was introduced to the theory for this purpose in Halliday 1961/76), in order to make explicit the nature of the relationship between the two lines. Then at the secondary degree of delicacy the structure within the 'Mood' element is divided into 'Subject + Finite'. This is not a matter of constituency relations, Halliday emphasises, but of a more delicate analysis at the same layer of structure, as we saw in Section 2.4 of Chapter 2.

Blogger Comments:

[1] This is misleading, because it misrepresents Halliday' model.  Fawcett's Figure 7 depicts a clause analysis; the notion of a 'text-sentence' is exclusive to Fawcett's model, and so does not feature in Halliday (1994).  As previously demonstrated, on Halliday's model, there are three lines of structure in the clause: textual, interpersonal and experiential.  Fawcett's Figure 4 shows four lines of structure because he misrepresents the structure of an information unit as a structure of the clause.

The "at least three other lines of representation" that Fawcett refers to here are:

• logical structure, which does not apply to clause structure (hence Fawcett's strategic use of "text-sentence" instead of clause);
• interpersonal structure misrepresented as two lines of structure instead of one (see below); and
• cohesive relations, which do not apply to clause structure (hence Fawcett's strategic use of "lines of representation" instead of lines of structure).

[2] This is misleading, because here Fawcett uses Halliday's first theory, Scale & Category Grammar (Halliday 1961) — which featured neither system nor metafunction — to misrepresent Halliday's final theory, Systemic Functional Grammar.

In SFL Theory, delicacy refers to the elaboration of systems, from the most general features to the most particular.  It does not refer to structural relations.

The relation of the Mood element to the Subject and Finite elements, on the other hand, is one one of composition — a subtype of extension, not elaboration. Halliday & Matthiessen (2014: 143):

The presence of the Mood element, consisting of Subject plus Finite, realises the feature ‘indicative’.

Importantly, from the perspective of SFL Theory, the sense in which this functional relation is "not a matter of constituency relations" is that it is not a matter of formal (rank) constituency; that is, it is not a matter of clauses consisting of groups consisting of words consisting of morpheme.

To be clear, in SFL Theory, the mood structure of the clause realises one type of meaning: that of the interpersonal metafunction.

Misrepresenting Halliday (1994) On Its Representation Of Clause Structure

Fawcett (2010: 113-4):

Halliday's introductory writings often imply that each line of structure corresponds to one of his three (or four) 'metafunctions', i.e., broad types of  'meaning'. There are four such metafunctions when he splits the 'ideational' metafunction into the 'experiential' and the 'logical', e.g., as described in Halliday (1977/78:128), but in his introductory writings he frequently omits the 'logical' and only illustrates three (as in Figure 7). The interesting point here, then, is that the four 'strands of meaning' shown in Figure 7 — and in all equivalent representations in IFG — do not correspond to Halliday's four 'metafunctions'. Thus the THEME and INFORMATION lines of analysis both belong in his 'textual' metafunction; the MOOD analysis (which itself contains two lines) is the 'interpersonal' metafunction; and the TRANSITIVITY is 'experiential'. The fourth 'metafunction' (i.e., the 'logical') is omitted in Figure 7, because Halliday only introduces it when there is a 'logico-semantic relationship' between two or more clauses.

Blogger Comments:

[1] This is misleading.  Halliday explicitly states — over and over, across many publications — that each line of structure corresponds to the meanings of a separate metafunction.  For example, Halliday (1994: 35):

[2] To be clear, the logical metafunction is concerned with relations between units in complexes all along the rank scale, whether clauses in clause complexes, groups or phrases in group or phrase complexes, words in word complexes, or morphemes in morpheme complexes.

[3] This is misleading. To be clear, Fawcett's Figure 7 illustrates only three metafunctional strands of meaning: textual, interpersonal and experiential, and misrepresents information as a structure of the clause (instead of the information unit). Reminder:

[4] This is misleading. To be clear, there are no representations in IFG that are equivalent to Fawcett's Figure 7; that is, there are no representations showing all three lines of clause structure in one diagram, though there are several that show both MOOD and TRANSITIVITY. The nearest equivalents appear in Appendix 1, an analysis of the 'Silver' text, but these are not confined to clause structure, and additionally demonstrate both cohesion and information analyses, neither of which constitutes clause structure, as well as the analysis of grammatical metaphor.  For example, Halliday (1994: 371):

Clearly, Fawcett confuses this series of composite 'text analysis' diagrams in IFG with diagrams merely illustrating clause structure.

[5] This is misleading.  Mood structure represents only one line of meaning, interpersonal meaning, but the structure itself involves functional constituency.  For example, the Mood block of finite clauses consists of the Subject and Finite elements (± modal Adjuncts), and the Residue can consist of the elements Predicator, Complement(s) and circumstantial Adjunct(s).