Fawcett's 'Item'

Fawcett (2010: 226-7):

The third of the three major categories in the present theory of syntax (with 'unit' and 'element') is the item. This term includes both 'word' (in its traditional sense) and 'morpheme'. Strictly speaking, the concept of 'item' lies outside syntax, since items are a different manifestation of meanings at the level of form from syntax. The four manifestations are: (1) items (words or morphemes, their relationship being described below); (2) syntax (i.e., the concepts that define relations between items), and (3) either intonation or punctuation (depending on whether the text is spoken or written). However, we need to bring the concept of 'items' into the picture to complete the account of syntax, because syntax only ends when elements are expounded by items.

In the present theory of syntax, the lowest syntactic category on each branch of the tree in a tree diagram representation of a sentence is an element (e.g., the head of a nominal group). And each such lowest element is expounded by an item — or as we shall see shortly, by items (in the plural).

Notice, then, that an element such as the head of a nominal group is not 'filled' by the unit of the 'word', as it would be in "Categories" and, in principle, in IFG. (In practice, however, elements of groups are almost always shown in IFG as being expounded directly by words, roughly as advocated here — the reason being that the description in IFG does not go below the 'rank' of the 'group' (which of course includes its elements).

Items are quite different from the other categories that we have discussed so far, because they do not have an internal structure that it [sic] relevant to a generative grammar and so have, in spoken language, what Firth calls "phonetic and phonological 'shape' " (1957/68:183). (He also introduces the term "graphic exponents" for what he terms the "companion study" of written language.) It is precisely the fact that they have a phonological or graphological shape that differentiates them from the relatively more abstract categories of syntax that we have been considering in the earlier sections of this chapter.

As we shall now see, languages vary quite considerably in how soon, as one moves down the layers of structure in the representation of a text-sentence, one escapes the abstract categories of syntax and reaches the first category that has the phonological or graphological "shape" of an item.

 

Blogger Comments:

[1] To be clear, on the one hand, it is true that the concept of 'item' lies outside syntax, since it is concerned with modelling morphology rather than syntax. However, on the other hand, morphology and syntax can be seen as two manifestations of the same phenomenon, as Halliday (2002 [1961]: 51) argues in the source of Fawcett's model, Scale & Category Grammar:

Traditionally these terms have usually referred to “grammar above the word” (syntax) and “grammar below the word” (morphology); but this distinction has no theoretical status. … But it seems worthwhile making use of “syntax” and “morphology” in the theory, to refer to direction on the rank scale. “Syntax” is then the downward relation, “morphology” the upward one; and both go all the way.

[2] To be clear, this incongruously models form at different levels of symbolic abstraction — content ("items and syntax") and expression ("intonation and punctuation") — as being at the same level.

[3] As previously observed, although Fawcett claims that the Cardiff Grammar does not include a rank scale in its architecture, its syntagmatic categories are nevertheless ranked on a scale from highest to lowest.

[4] This is potentially misleading. To be clear, the reason why elements of groups are "almost always" shown as realised by words in IFG is because that is how the theory models the relation. Elements of structure at a higher rank (e.g. group) are realised by units of the rank below (e.g. word). The only exceptions are instances of rank-shift, as when the Qualifier of a nominal group is realised by an embedded clause or prepositional phrase.

[5] To be clear, in SFL Theory, both syntax and morphology ('items') have "phonological and graphological shape" to the extent that lexicogrammar is realised in the expression plane systems of phonology and graphology.

Problems With The Notion Of Places Without Elements

 Fawcett (2010: 225-6):

It is important to emphasise that in such cases the "raised" element does not actually become an element of that higher unit; it simply occupies a place in that unit. As will now be clear, it is the fact that the theory contains the concept of 'place' as well as 'element' and 'class of unit' that makes it possible to handle this otherwise problematical phenomenon in a principled manner. It is not clear from the published works of Halliday, Matthiessen etc. how such problems would be handled in the Sydney Grammar, either descriptively or in a generative version of the grammar.

Blogger Comments:

[1] To be clear, the claim here is that the who in who were you seen by? does not serve as structural element of the clause. That is, the claim here is that who does not serve as the Theme, Senser or Complement of this clause, and, consequently, that these elements are absent from this clause. In other words, the incongruous claim here is that the clause who were you seen by? has no Theme, no Senser and no Complement.

[2] As can be seen from the previous post, and from [1] above, the concept of 'place' is unwarranted theoretically, and leads to unintended absurdities such as the clause who were you seen by? having no Theme element.

[3] To be clear, Halliday & Matthiessen (2004: 137) are quite explicit on the matter:

In the selection of the WH- element, the category of Complement can extend to include the minor Complement of a prepositional phrase. Here the WH- element is conflated with the minor Complement of a prepositional phrase serving as a circumstantial Adjunct in the clause. Since the WH-element is thematic, the minor Complement of the prepositional phrase is given the status of Theme, while the minor Predicator appears within the Rheme, in the position the Adjunct has when it is not thematic …

 See the previous post for exemplifying clause structures.

The Key Rôle Of 'Place' In Handling 'Raising' Phenomena

Fawcett (2010: 224-5):

However, there is one aspect of syntax where the concept of 'place' assumes major theoretical importance. It is when a place in one unit — in practice, always a clause — is occupied by an element from a lower unit. This phenomenon occurs in cases such as Who were you seen by?, Who did you give it to?, Where did you say you put it? and so on.

Consider the first example, the analysis of which is shown in Figure 13. Here, the initial item who fills Place 28 in the clause. This is the place at which 'sought' elements (realised as wh-items) typically occur. However, the element who is not an element of the clause, but of the prepositional group that fills the Complement. In other words, in this example it has been made to occupy a place in the unit above the unit of which it is an element.

… Essentially the same principles apply in the analysis of examples such as Who did you say you saw there? (See Section 11.7 of Chapter 11 for the concept of discontinuity, and see Fawcett (in press) for the analysis of these and other types of discontinuity.

 

Blogger Comments:

[1] Reminder: Fawcett's claim is that the Cardiff Grammar does not use a rank scale ('lower', 'above'), and does not feature units, though it does feature classes of units.

[2] To be clear, in SFL Theory, the first two examples are clauses, whereas the latter two are clause complexes. In the clause examples, the minor Complement of a prepositional phrase is conflated with the thematic WH- element:

Halliday & Matthiessen (2014: 163-4) explain:

In the selection of the WH- element, the category of Complement can extend to include the minor Complement of a prepositional phrase. Here the WH- element is conflated with the minor Complement of a prepositional phrase serving as a circumstantial Adjunct in the clause. Since the WH-element is thematic, the minor Complement of the prepositional phrase is given the status of Theme, while the minor Predicator appears within the Rheme, in the position the Adjunct has when it is not thematic…

However, this is not the case for the two clause complex examples.  Here the Adjunct and Complement of projected clauses — not the Complement of a prepositional phrase — are each thematised beyond the (now included) projecting clause:

Here the first clause is a thematic agnate of Did you say where you put it? and the second clause is a thematic agnate of Did you say who you saw there?.

But see Matthiessen (1995: 417), who (bizarrely) interprets the thematised element of the projected clause as functioning as an element of the projecting clause.

[3] To be clear, in SFL Theory, the who in Who were you seen by? does serve as an element of clause structure:

• textually as Theme,
• interpersonally as Complement/WH-,
• experientially as Senser (along with by).

Importantly, the Range/Complement of a prepositional phrase functions as an indirect participant in the Process of the clause — indirect, because the participation is mediated by the minor Process/Predicator of the prepositional phrase. Halliday & Matthiessen (2014: 312, 329):

We can make a contrast, then, between direct and indirect participants, using ‘indirect participant’ to refer to the status of a nominal group that is inside a prepositional phrase. We have already seen that the participant roles of (1) Client, Recipient and Receiver and (2) Scope, Behaviour and Verbiage are sometimes expressed ‘indirectly’ in this sense, as in gave money to the cashier, plays beautifully on the piano. … The preposition, it was suggested, acts as a kind of intermediary whereby a nominal element can be introduced as an ‘indirect’ participant in the main process.

[4] Reminder: Fawcett (in press) is still unpublished, 21 years after the first edition of this book.

Confusing Nominal Group Head With Thing

Fawcett (2010: 222n):

²² … Rather similarly, while there is a very strong probability that a nominal group will have a head, some nominal groups do not. … But if we were to decide that we wished to say that two in Give me two was a head, we would need to have a realisation statement of the form: 'If Feature A is selected the realisation is X, but if Feature B is selected the realisation is Y', in order to cover the cases where of them was and was not made overt. The realisation rules for cardinals are already quite complicated (to accommodate cases such as around fifty, five thousand two hundred, over five thousand two hundred etc.), and it is therefore preferable to avoid the unnecessary additional complication of having two (and so every other cardinal) expound two different elements. And the same general principle applies to the very rich, the poor, the old and infirm, etc. …On these grounds (and others) it is greatly preferable to build into the grammar the possibility of having occasional nominal groups that have no head and clauses that have no Main Verb. Allowing for this avoids losing many useful generalisations such as those identified above. Note, however, that this is only possible when the grammar works by locating elements at places, rather than relying on the presence of some "anchor" element in relation to which all other elements are ordered.

Blogger Comments:

[1] This is misleading, because it is untrue. In SFL Theory, all nominal groups have a logical Head, though not all nominal groups have an experiential Thing; see further below.

[2] To be clear, in SFL Theory, two in Give me two is the Head of a nominal group, and the realisation statement is simply conflate Head/Numerative. In SFL Theory, the structures of these nominal groups are analysed as follows:

[3] To be clear, in SFL Theory, the realisation statement for these instances is conflate Head/Epithet. In SFL Theory, the structures of these nominal groups are analysed as follows:

[4] To be clear, Fawcett's Main Verb corresponds to the Predicator in SFL Theory, and SFL Theory accommodates the absence of a Predicator, whether through ellipsis or in some relational clauses. For example, Halliday & Matthiessen (2014: 476):

With the additive, the Process of a relational dependent clause may be implicit; the marker is the conjunctive preposition with (positive) or without (negative):

||| I told the whole story of the six-minute Louvre at The Kennedy Centre || with President Carter there … |||

||| Without chlorine in the antarctic stratosphere, || there would be no ozone hole. |||

[5] This is misleading, because it is untrue. As demonstrated above, and previously, Fawcett makes this claim on the basis of his misunderstandings of SFL Theory.

Fawcett's 'Crucial Test Case' Against The SFL Method Of Ordering Elements

Fawcett (2010: 221-2):

The crucial test case in English for this approach is the clause. The fact is that there is an enormous variety of sequences in which the elements of the English clause may come — both absolutely and in relation to each other. But the first question is whether there is any element that it always present and always in one position. The only candidate for this 'anchor' role is the Main Verb — but the places at which it occurs do in fact vary. The most frequent reason for this variation is that, when it is expounded by a form of the verb be, it is often conflated with the Operator. Consider the two positions of the Main Verb is in Ivy is here and Is Ivy here? Clearly, the Main Verb in the second example precedes the Subject Ivy . There is in fact no element in the English clause that is (1) always located at the same place and (2) always present. And the situation with respect to the groups in English is similar, e.g., the head of a nominal group and the apex of a quality group are not obligatory, as the relevant sections of Fawcett (in press) show.²²

Blogger Comments;

[1] To be clear, this is true, but irrelevant. The sequencing of elements is specified by realisation statements such as Subject^Finite, conflate Theme/Subject, conflate Subject/Actor etc., and the consequences of such realisation statements.

[2] As previously explained, this is irrelevant, because ordering realisation statements (e.g. Finite^Subject) only apply to features whose realisation statements also specify the insertion of both elements (e.g. +Subject, +Finite), and because they specify the relative ordering of elements, an atypical location of one element does not affect the relative ordering of the elements.

[3] To be clear, according to SFL Theory, the motivation for sequence variation is textual: e.g. to highlight an element as Theme, or background it as Rheme. This can be seen in polar interrogative clauses, where the polarity of the Finite element is thematised, and in the choice of clause voice, operative vs receptive, which selects which participant is thematised. 

[4] To be clear, in SFL Theory, such instances are not actually conflations of Finite ('Operator') and Predicator ('Main Verb'). Halliday & Matthiessen (2014: 153):

There are two lexical verbs in English, be and have, where, strictly speaking, the simple past and simple present forms consist of Finite element only, rather than of a fusion of Finite with Predicator. This is shown by the negatives: the negative of is, was is isn’t, wasn’t – not doesn’t be, didn’t be. Similarly with have (in the sense of ‘possess’, not have in the sense of ‘take’): the negative forms are hasn’t, hadn’t…

In Fawcett's examples, the ordering of the elements Subject and Finite is specified by the realisation statement Subject^Finite for the feature 'declarative', and the realisation statement Finite^Subject for the interrogative feature 'yes/no' (Halliday & Matthiessen (2014: 162).

[5] This is misleading, because it is untrue. In SFL Theory, every nominal group has a Head element. Fawcett's false claim derives from his repeated confusion of (logical) Head with (experiential) Thing. (Fawcett's quality group does not feature in SFL Theory; see previous posts for the misunderstandings on which it is formulated.)

Fawcett's Argument Against The SFL Method Of Ordering Elements

Fawcett (2010: 221):

The problem with the first method is that if, in a given instance, the element that is used as the 'anchor' point for placing another element were to be either missing from the unit or located in an untypical position in it, then the statement for placing any 'dependent' element in the structure would become much longer. This is because it would have to include a set of conditional rules, whose role would be to specify what should be done under various possible scenarios if the 'anchor' element were not to be both (1) present and (2) in its typical place. And these conditional rules would become exponentially more complex as the grammar was extended to handle the great range of possible variations in the sequence of elements in the English clause that occurs in natural language texts — especially in the varying positions of the various types of Adjunct. Indeed, it is only practicable to use this first approach in a highly limited sub-set of cases — i.e., (1) where the grammar is small (e.g., one that has been developed for illustrative purposes such as that in Halliday 1969/81) and (2) where every unit recognised in the grammar has at least one element that is both (a) obligatorily present and (b) always occurs in the same position.

Blogger Comments:

[1] This is misleading, because ordering realisation statements (e.g. Finite^Subject) only apply to features whose realisation statements also specify the insertion of both elements (e.g. +Subject, +Finite).

[2] This is misleading, because ordering realisation statements specify the relative ordering of elements, and an atypical location of one element does not affect the relative ordering of the elements.

[3] To be clear, these claims are irrelevant, since they derive from the misunderstandings identified above ([1] and [2]).

The SFL Method Of Sequencing Elements In A Unit

Fawcett (2010: 221):

Systemic functional linguists have explored two ways of locating elements in an appropriate sequential relationship to each other in a unit. The first — which may at first sight appear to be the simplest — is to locate each element in its 'place' by relating it to some previously located element. This approach depends crucially on the existence of what we might term an 'anchor' element in each unit, i.e., the existence of a 'pivotal' element that is always present. It was this method that Halliday used in his seminal first description of a generative SF grammar (Halliday 1969/81). Surprisingly, references to this as a method of sequencing elements in a SF grammar are still found in current descriptions by Halliday (1993: 5405) and Matthiessen (1995: 23-4).

I say "surprisingly" because, when Mann and Matthiessen were working on the large-scale computer implementation of Halliday's SF grammar in the Penman Project in the late 1970s, this approach caused problems. When faced with the additional complexities of building an [sic] large, generative grammar, they found that in practice they had to turn to the second method — to which we shall come in a moment (Matthiessen, personal communication).

Blogger Comments:

[1] To be clear, the first method of sequencing elements is Halliday's theoretical approach, whereas the second is Fawcett's method for adapting SFL Theory to the limitations of computers for the purpose of text generation.

[2] This is potentially misleading. In terms of the realisation statements that specify the sequencing of elements, what is required is the insertion of the elements that are to be sequenced. The notions of 'pivotal' and 'anchor' are irrelevant, as demonstrated by realisation rules like Finite^Subject, where neither element is 'pivotal' or an 'anchor', even in Fawcett's terms, since the 'anchor' or 'pivotal' element of the clause in the Cardiff Grammar is the 'Main Verb' (p201).)

[3] To be clear, the cited instances are as follows. Halliday (1995 [1993]: 272):

(c) 'Order' an element with respect to another, or to some defined location (e.g., order finite auxiliary before subject);

Matthiessen (1995: 23-4):

Structuring statements determine the organisation within one layer of a function structure:

(1) the presence of a grammatical function (insertion; e.g., "insert Subject," symbolised +Subject), 

(2) the relative sequence of two grammatical functions (ordering; e.g., "order Subject Finite," symbolised Subject^Finite), and  

(3) the constituency relationship between two functions (expansion; e.g., "expand Mood Subject", symbolised Mood(Subject)).

[4] The word "surprisingly" here is misleading, because it is the opposite of what is true. The need to adapt a theory of human language to the limitations of computers for text generation would not be surprising to anyone engaged in such a project, unless they were unable to distinguish between theory and its context-specific application.

Fawcett's Concept Of Places In A Unit

Fawcett (2010: 220):

Elements, as we have seen, are the 'components' of units. However, there is another category that intervenes between a unit and its elements. This is the concept of the places in a unit. In other words, elements occur at places in units.

In the analysis of texts the places are usually omitted (as in Figure 10 in Chapter 7). However, in the full, generative version of the grammar they play an essential role, as in Figure 13 in Section 10.4.4 below, and as illustrated in the generation of a text-sentence in Fawcett, Tucker & Lin (1993).

Like a number of other concepts that are central in a modern SF grammar (such as the concept of 'system' itself and 'lexis as most delicate grammar') the concept of 'place' can be found in "Categories" (as I pointed out in Section 2.3 of Chapter 2). The word is used rather than defined as a central concept, but its use there seems to be consistent with the sense in which it is used here.

Blogger Comments:

[1] To be clear, this is inconsistent with the architecture of SFL Theory, in which the constituents ('components') of a unit are units of the rank below. Elements, on the other hand, are the functions that those lower rank units serve in the structure of the higher rank, as exemplified by a nominal group serving as the Phenomenon of a mental clause.

[2] To be clear, in SFL Theory there are no pre-existing places in structures at which elements appear. Instead, the sequencing of elements in a structure is specified by realisation rules in system networks. For example, in the type of example that Fawcett comes to discuss (p224-5), the different ordering of elements according to MOOD selection:

• the feature 'indicative' activates 'realised by insert Mood, expand: insert Finite, insert Subject',
• the more delicate feature 'declarative' activates 'realised by the order Subject ^ Finite, whereas
• the interrogative feature 'yes/no' activates 'realised by the order Finite ^ Subject', and
• the disjunct interrogative feature 'WH' activates 'realised by insert Wh, the order Wh ^ Finite'.

See Halliday & Matthiessen (2014: 162).

[3] Figure 13 (p225):

[4] This is misleading, because 'Categories' (Halliday 1961) set out the architecture of a long superseded theory, Scale & Category Grammar, not a "modern" Systemic Functional Grammar.

Misrepresenting The SFL Model Of Structure

 Fawcett (2010: 220):

Moreover, the representation of any layers of structure other than the one that reflects directly the choices in the system networks is redundant. (It is not even useful for pedagogical purposes, since it is misleading to suggest that a string of elements of structure that function as the realisations of a set of individually chosen meanings are in some sense "a single element".)

An[d] elements of structure should be defined in terms of the function it serves, i.e., by the aspect of the meaning of the unit that it expresses, rather than by its position in the unit. Since each element in each unit realises a different type of meaning, it should have a different label. The position of an element in its unit is a matter that is handled separately, through the concept of 'place' — the concept that is to be discussed in the next section.

Blogger Comments:

[1] This is misleading. To be clear, the Mood element, for example, is specified by the realisation statement 'insert Mood' which is activated by the selection of the feature 'indicative'; see the Mood system in the previous post.

[2] This is misleading. To be clear, Mood and Theme are both single elements, each of which can consist of constituent elements.

[3] On the one hand, this confuses formal constituents with functional elements. In SFL Theory, it is formal constituents (e.g. nominal and verbal groups of the clause) that are interpreted in terms of the function they serve (e.g. Senser and Process elements of clause structure). 

On the other hand, it is misleading because it falsely claims that, in SFL Theory, elements are defined in terms of position instead of function; see the previous posts on Fawcett's misunderstandings of nominal group structure.

[4] To be clear, as previously noted, Fawcett's principle is violated by Fawcett's own model, where there are multiple elements with the same label, such as the Modifier element of the nominal group (p306):

[5] To be clear, in SFL Theory, the position of an element in a structure is specified systemically.

Fawcett's Argument For 'Secondary Structure'

Fawcett (2010: 219-20):

In developing a SF grammar, both for very large computer implementations and for text analysis, priority must be given to the "most delicate" possible of structural descriptions, because a full account of the meaning potential of a unit (such as a clause) requires statements about each element in its own right. It is the individual elements of a unit that carry the different meanings that are the focus of interest for a functional grammarian. 

For example, the fact that the main mood meanings of a clause are realised by the configuration of the Subject and Operator is best shown by stating the semantic feature that generates this configuration in an analysis of the meaning potential of the clause, as in Figure 10 in Chapter 7 (rather than introducing an additional layer of structure as part of its syntax, as in the case of "Mood + Residue" in IFG).

Blogger Comments:

[1] To be clear, Systemic Functional Theory is a model of language as a phenomenon. Text analysis is an application of the theory, and computer implementations require the modification of the theory to accommodate the limitations of computers.

[2] To be clear, on the one hand, this priority flatly contradicts Fawcett's previous argument against what he regards as the "most delicate" structures; see the previous post. On the other hand, this priority is inconsistent with SFL Theory, because it argues for the view 'from below' (structure and form) rather than the view 'from above' (system and function).

[3] To be clear, the focus of interest for a Systemic Functional grammarian is the system that specifies structural realisations.

[4] To be clear, as previously noted, in Systemic Functional Linguistic Theory, structural realisations are specified systemically as realisation statements. Halliday & Matthiessen (2014: 162):

[5] To be clear, as previously noted, Fawcett's Figure 10 confuses systemic features of the clause with elements of structure:

Moreover, the only semantic feature that corresponds to the configuration of Subject and Operator is the unexplained 'information giver'. Significantly, Fawcett does not provide the semantic networks from which these features are derived.

Fawcett's Argument Against 'Secondary Structure'

Fawcett (2010: 219):

A more general reason why we should expect the use of the concept of 'secondary structure' to wither away in modern SF grammars is that we now have much fuller functional descriptions of languages than those that were available when Halliday was writing "Categories". The result is that it no longer feels adequate simply to label all the elements that precede the head in a nominal group as a single 'modifier', or to label all the thematised elements in a case of multiple Theme as a single Theme. In IFG, for example, we would not find a nominal group such as that lovely porcelain vase from China that you broke last week me [sic] being analysed as if it had just three elements of structure at the 'primary' degree of delicacy (i.e., Modifier, Head and Qualifier, as these terms are used in "Categories"). Instead it would be analysed immediately into its supposedly 'secondary' structure, i.e., as having (in IFG terms) a Deictic, an Epithet, a Classifier, a Head and two Qualifiers.

Blogger Comments:

[1] To be clear, as previously explained, the concept of a more delicate 'secondary structure' was a feature of Halliday's superseded theory, Scale & Category Grammar (1961), and does not feature in SFL Theory.

[2] To be clear, having "much fuller functional descriptions of languages" is irrelevant to the theoretical principle of delicacy of structure in Halliday (1961). Halliday originally proposed the notion of delicacy of structure (elaboration) to distinguish it from the compositional rank scale (extension). However, with the development of SFL Theory, it became clear that the organising principle was not delicacy (elaboration), but composition (extension), though applied to functional elements, not formal constituents (the rank scale).

[3] To be clear, on the one hand, here Fawcett has become confused, switching his argument against 'secondary structure' to an argument against 'primary structure' ("a single Modifier", "a single Theme"). On the other hand, the "argument" — "it no longer feels adequate" — is merely a personal attitudinal evaluation, not an instance of reasoning on the basis of evidence; see [4] and [5].

[4] To be clear, on the one hand, the reason why a 'Categories' analysis would not appear in IFG is because IFG is an exposition of the current theory: Systemic Functional Grammar (1994), whereas Categories is an exposition of the superseded theory: Scale & Category Grammar (1961). On the other hand, Fawcett is again arguing against what he believes is a primary structure, while purporting to be arguing against what he takes to be secondary structure.

[5] This is misleading, because it misapplies SFL Theory as expounded in IFG. On the one hand, Fawcett misrepresents an experiential structure as a secondary structure of logical structure — while placing a logical element (Head) in an experiential structure. On the other hand, Fawcett neglects to mention how the logical structure is expanded, or how the two Qualifiers differ in terms of the constituency of the nominal group:

Fawcett's Argument Against 'Primary' And 'Secondary' Structures

Fawcett (2010: 218-9):

In my view, then, it is unhelpful to have 'primary' and 'secondary' structures in the representation of a clause. Firstly, it implies that elements form a 'multiple element', when the evidence from the syntactic distribution of the elements themselves shows that they do not. Secondly, they risk introducing misunderstandings about the number of 'strands of meaning' in a clause.

However, while the concept of 'more delicate structure' has no place in a modern theory of SF syntax, I have to admit that the Cardiff Grammar's description of English still contains a few historical remnants of the concept. One is found in the convention that the term "modifier" is used as the second part of the names of many different elements in the nominal group, e.g., the "affective modifier", the "epithet modifier", and so on. And the same goes for Adjuncts and, to a lesser extent, determiners in the nominal group and Auxiliaries in the clause. (See Appendix B for examples.) Moreover, at the introductory level of text description we simply use "m" for all the different types of 'modifier', "X" for the different types of 'Auxiliary Verb', and so on.

Blogger Comments:

[1] As previously explained, the architecture of SFL Theory does not include the notion of primary and secondary structures distinguished on the basis of delicacy. This was a feature of Halliday's superseded theory, Scale & Category Grammar (1961) — the theory on which Fawcett's "modern theory of SF Theory" is based.

[2] This is misleading and confused. It is not that elements "form a multiple element", but that one element such as Theme, Mood or Residue can itself be composed of elements.

[3] This is misleading. On the one hand, Fawcett does not produce the evidence that he claims exists, while on the other hand, in a functional grammar, the criterion for structural elements is their function (the view 'from above'), not their syntactic distribution (the view 'from below').

[4] This is misleading, because it is untrue. The number of strands of meaning in the clause is unambiguously three, because they are metafunctionally defined. The average academic can distinguish between the theoretical principle and the number of rows in a representation of structure.

[5] This is misleading. To be clear, presenting a single line of structural analysis with multiple elements with the same label, as illustrated in Appendix B (pp305-6), does not exemplify Halliday's (1961) double-layered structures of primary and secondary delicacy:

Misrepresenting Halliday On Structural Layers In SFL Theory

Fawcett (2010: 218):

Thus Halliday continues to use the "Categories" concept of having a 'primary', 'secondary' and if necessary a 'tertiary' structure within each strand of text analysis. However, it is not clear how far this concept still has its former central role in Halliday's current theory. While it is not mentioned in the summary of "Systemic theory" given in Halliday (1993), it is there by implication in his realisation operation "Split" (for a discussion of which see Section 9.2.3 of Chapter 9). And the concept that there is a 'primary' and a 'secondary' structure plays a fairly prominent role in the examples of clause analysis in IFG, often appearing twice in the analysis of a single clause (once for MOOD and once for THEME, whenever there is a case of 'multiple Theme'). Its main justification seems to be that it is thought to add insight to the description. (The theory does not require it, as I have shown in Section 9.2.3 of Chapter 9).

Blogger Comments:

[1] This is misleading, because it is not true. The notion of primary and secondary structures, conceived of in terms of delicacy (elaboration) of structure, was a feature of Halliday's superseded theory, Scale & Category Grammar (1961) only. In SFL Theory, there are no primary and secondary structures, and the principle behind the layered structures of Mood and Theme is composition (extension), not delicacy (elaboration).

[2] To be clear, adding insight to a description is a positive quality of a theory. See previous posts on Mood and Theme for the theoretical motivations for layered structuring that are unknown to Fawcett.

[3] This is misleading, because it is not true. See, for example, two of the previous clarifying critiques of Section 9.2.3, 'A comparison of the realisation operations in the two frameworks':

• Misrepresenting Halliday (1993) On The Realisation Statement 'Expand'
• Misrepresenting Halliday (1994) On Maximally Extended Theme

Fawcett's Argument Against Layered Representations Of Theme

 Fawcett (2010: 218, 218n):

A similar argument applies to the IFG analysis of Theme, as in Halliday (1994:55). Here the analyses frequently show three layers of structure within one strand of meaning for the 'theme' meanings in a clause.²¹ So, while Halliday may claim that the concept of "secondary structure" does not add another layer of structure, in constituency terms, it does add one or more additional lines of analysis to virtually every representation of a clause.

²¹ The IFG example is Well but then Ann surely wouldn't the best idea be to join the group? It is hard to see why any analysis other than the third and most delicate is required in such cases of multiple Theme. It is unhelpful to imply that all of the thematised elements constitute "the theme" in any semantically unified sense. Similarly, there is little point in giving a line of analysis to showing where two or more happen to share a metafunction. The fact is that several different elements of the clause all happen to have been "thematised" at the same time, each for its own reason. (However, I should add that I do not consider all of the 'early' elements in the above example to be thematised, in that the items well, but, then and wouldn't are not early in the clause because of a systemic choice. But since there are also several genuinely 'thematised' elements, my point remains relevant to that example.)

Blogger Comments:

[1] To be clear, Fawcett's "argument" was that the inclusion of Mood and Residue in structural analyses makes them harder for him to read; see previous post.

[2] This is not misleading, because it is true — apart, perhaps, from the inclusion of the word 'frequently'. Halliday (1994: 55-6) provides five examples of multiple Themes.

[3] This is misleading, because 'secondary structure' is a feature of Halliday's superseded theory, Scale & Category Grammar (1961), not Systemic Functional Grammar (e.g. Halliday 1994). In Scale & Category Grammar, secondary structures were modelled in terms of delicacy, not constituency (rank). Halliday (2002 [1961]: 48):

Subsequent more delicate differentiations are then stated as secondary structures. These are still structures of the same unit, not of the unit next below; they take account of finer distinctions recognisable at the same rank. Rank and delicacy are different scales of abstraction: primary group structures differ in rank from primary clause structures, but are at the same degree of delicacy; while primary and secondary clause structures differ in delicacy but not in rank.

[4] This is true, but it is unproblematic from a theoretical standpoint. Fawcett's "argument" is merely that it makes the analysis harder for him to read.

[5] Halliday (1994: 55):

[6] Fawcett's wording here, 'It's hard to see why', is an example of the logical fallacy known as the argument from incredulity. To be clear, the reason why the "most delicate" analysis — the first, not third — is insufficient is precisely because it would fail to represent the fact — not implication — that 'all the thematised elements constitute the Theme in a semantically unified sense'.

[7] To be clear, the point was more pedagogical than theoretical. In subsequent editions, Halliday & Matthiessen (2004: 79-80; 2014: 107), the metafunctional classifications of the multiple Themes are provided separately from the clause analysis:

[8] To be clear, this is an example of a logical non-sequitur, since it is entirely irrelevant to the claim made in the preceding sentence; see [7].

[9] To be clear, in English, being put first is what realises Theme selection. Halliday & Matthiessen (2014: 387):

The textual meaning of the clause is expressed by what is put first (the Theme); … and by conjunctions and relatives which if present must occur in initial position.

Continuatives and conjunctions are inherently thematic because their textual functions are specific sub-types of the function of Theme. Halliday & Matthiessen (2014: 109):

Why do these items favour thematic position in the clause – or, to put the question more meaningfully, why are they associated with thematic function, either characteristically or, in some cases, inherently? In the most general sense, they are all natural Themes: if the speaker, or writer, is making explicit the way the clause relates to the surrounding discourse (textual), … it is natural to set up such expressions as the point of departure. The message begins with ‘let me tell you how this fits in’ … 

Those that are inherently thematic are the (textual) continuatives and conjunctions. As the language evolved, they have, as it were, migrated to the front of the clause and stayed there. Essentially they constitute a setting for the clause (continuative), or else they locate it in a specific logical-semantic relationship to another clause in the neighbourhood (conjunction). In either case, their thematic status comes as part of a package, along with their particular discursive force.

It might be added that, in Fawcett's model, as shown in Figure 10, Subject is always conflated with Theme, meaning that its thematisation is not a matter of systemic choice.

 

Fawcett's Argument Against Mood And Residue

Fawcett (2010: 217-8):

Let us now address the question of how far we need, in a modern SF grammar, the "more delicate differentiations" in "structure" that Halliday introduces in "Categories" (1961/76:63). There the 'primary' structure of the nominal group was said to be "M H Q", i.e., "modifier + head + qualifier", and the distinctions within the "modifier' between "deictic", "numerative" and "epithet", etc., were said to be a matter of 'secondary' structure.

The first point to make is that these "secondary structures" do not — in principle at least — constitute another layer of structure in the representation. Halliday emphasises that they are "still structures of the same unit, not of the unit next below" — the key point being that "they take account of finer distinctions recognisable at the same rank" (by which he means "in the same unit"). However, the fact is that when he introduces such "secondary structures" to the representation of a text-sentence alongside the "primary structures", he adds another line to the analysis — as the presence of two lines of analysis for MOOD in Figure 7 in Chapter 7 clearly illustrates. Indeed, every finite clause in IFG is analysed terms of (1) its "Mood" and "Residue", and then within the "Mood" (2) its "Subject" and "Finite".

If the line showing the "Mood" and "Residue" is removed from such diagrams, the display of the analysis becomes simpler to read, while still preserving the essential insight that it is the relationship of the Subject and the Finite that is the primary expression of the meaning of MOOD. In other words, I am suggesting here that the additional layer of analysis into "Mood" and "Residue" detracts from the insightfulness of the diagram rather than adding to it. Figure 10 illustrates the way in which the semantic information can be extracted directly from the relationship of the Subject and Operator (or Finite). 

Blogger Comments:

[1] To be clear, here Fawcett has stopped outlining his own model, and has returned to one of his previous critiques of Halliday's model. However, despite claiming to be concerned with "a modern SF grammar", he focusses here on Halliday's superseded theory, Scale and Category Grammar (1961), which featured neither metafunctions nor system networks.

In SFL Theory, 'delicacy' does not apply to structure, there are no secondary structures, and the structures of the nominal group are differentiated in terms of metafunction: logical (Modifier ^ Head) vs experiential (Deictic ^ Numerative ^ Epithet etc.).

[2] This is misleading. Without acknowledging the fact, here Fawcett switches attention from Scale and Category Grammar to SFL Theory. In Scale & Category Grammar, the elements of clause structure are simply Subject, Predicator, Complement, Adjunct (Halliday 2002 [1961]: 47). The Mood and Residue blocks were not theorised in Scale & Category Grammar. 

[3] This is not misleading, because it is true for indicative clauses.

[4] To be clear, Fawcett's argument against the theoretical categories of Mood and Residue is that their representation in structural analyses makes them harder for him to read.

[5] To be clear, the structural element Mood is functionally motivated. Halliday & Matthiessen (2014: 142, 143):

The Mood is the element that realises the selection of mood in the clause; and it is also the domain of agreement between Subject and Finite. …

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

(2) Within the indicative, what is significant is the order of Subject and Finite:

(a) The order Subject before Finite realises ‘declarative’;

(b) The order Finite before Subject realises ‘yes-no interrogative’;

(c) In a ‘WH- interrogative’ the order is: 

(i) Subject before Finite if the WH-element is the Subject;

(ii) Finite before Subject otherwise.

[6] This misleading, because Figure 10 (below) does not show how "the semantic information can be extracted directly from the relationship of the Subject and Operator (or Finite)". It merely juxtaposes Fawcett's semantic and syntactic categories for one instance. Moreover, the semantic networks from which these features are derived are not provided anywhere in this publication.

Note also that Fawcett incongruously relabels Halliday's Mood element we would as 'information-giver'.

Confusing Structure With (Transcendent) Ideational Denotation

Fawcett (2010: 216-7):

What is the relationship between one element and its 'sister' elements in a unit? This has been a major focus of interest for some grammarians, leading to arguments about whether such relationships are those of 'daughter dependency' or 'sister dependency' (e.g., Hudson 1976). (For a brief comparison of the two, see Section 11.2 of Chapter 11.) Here I offer a new answer to the question asked above. It is one that follows directly from the adoption of the framework outlined in Chapter 2 and exemplified in Appendix A.

Let us take as an example the relationship between a modifier and the head in the English nominal group. In the framework of a systemic functional grammar the relationship is not, I suggest, the direct one that form-centred grammarians consider it to be. In formal and traditional grammars, it is simply assumed that what the modifier modifies is the head. Here, however, the general function of the modifiers in a nominal group is regarded as being to describe the referent. (See Fawcett (in press) for the sub-types of 'description' that the various sub-types of modifier express, e.g., 'colour modifiers', 'affective modifiers', 'general epithet modifiers' etc.) Similarly, the function of the head of the nominal group is (assuming that it is a noun) to state the 'cultural classification' of the referent. The referent is thus the object to which the nominal group refers, and it is the function of the noun at the head of the nominal group to express what Lyons (1977:206-7) terms the "denotation" (of some class of 'thing'). In other words, the head realises one type of meaning that relates to the referent, while the modifier realises another. So both the modifier and the head relate, via the meanings they express, to the referent — but they are related only indirectly to each other. Thus a modifier does not in fact 'modify' (or 'describe') the head; it modifies (or describes) the referent which the head denotes.

This general principle applies to all 'sister' relationships between elements, and it applies to all units. From this viewpoint, the question of whether an element is dependent on a 'sister' element such as the 'head' or on a 'mother' unit is beside the point; the 'dependency' is not in fact 'syntactic' at all, and what we observe in syntax is the realisation of dependence in the system networks.

Blogger Comments:

[1] To be clear, this confuses relations between formal units with relations between functional elements.

[2] To be clear, this confuses structure with ideational denotation ("describing a referent").  In SFL Theory, a structure is the relation between the elements of a unit. Halliday & Matthiessen (2014: 451):

Note that, although it is the functions that are labelled, the structure actually consists of the relationships among them.

In SFL Theory, which takes an 'immanent' view of meaning, the ideational denotation of a wording (lexicogrammar) is its ideational meaning (semantics). However, in other theories, which take a 'transcendent' view of meaning, the ideational denotation is to a domain outside language.

[3] To be clear, as Fawcett explicitly makes clear by the wording "meaning that relates to the referent", his model takes a 'transcendent' view of meaning, a view that is seriously inconsistent with the epistemological assumptions from which SFL Theory is constructed.

[4] To be clear, this is inconsistent with SFL Theory, in which Modifier–Head is a univariate structural relation. Halliday & Matthiessen (2014: 390):

We refer to this kind of structure as a univariate structure, one which is generated as an iteration of the same functional relationship (cf. Halliday, 1965, 1979): α is modified by β, which is modified by γ, which is ... .

[5] To be clear, if 'dependency' is a syntagmatic relation between formal units, then it is 'syntactic'.

[6] To be clear, this is inadvertently consistent with the architecture of SFL Theory — or, at least with the general principle that syntagmatic structures realise paradigmatic selections.

Fawcett's Linkers, Deictics & Quantifiers

Fawcett (2010: 216, 216n):

There is a good reason why, in all of these cases, the same element is permitted in more than one unit. It is that the same meaning is carried in relation to the entity being expressed, e.g., 'co-ordination' of one of several types in the case of the Linker. So, with the exception of these well-motivated cases, the principle that a unit and its elements of structure are mutually defining holds good.²⁰

²⁰ To complete the picture, it should be said that there are two elements that occur in a quality group when it is being used to express a 'superlative' or 'ordinative' meaning, which have considerable similarities to the deictic determiner and the quantifying modifier in the nominal group, e.g., as in the use of the and two in the underlined portion of the most interesting two of Shakespeare's plays. See Appendix B for the names of the elements.

Blogger Comments:

[1] To be clear, the paratactic structural relation ('co-ordination') marked by a Linker is to other units, not to other elements within the same unit. In SFL Theory, a structure is the relation between the elements of a unit. Halliday & Matthiessen (2014: 451):

Note that, although it is the functions that are labelled, the structure actually consists of the relationships among them.

[2] To be clear, the coyly withheld names of the elements that resemble the deictic determiner and quantifying modifier of Fawcett's nominal group appear to be the quality group deictic and the quality group quantifier (p304). In SFL Theory, the quality group is a portion of the nominal group, as demonstrated in previous posts.

Fawcett's 'Starter' And 'Ender' Elements

 Fawcett (2010: 215-6, 216n):

… the present model also includes punctuation (and intonation, which we shall not discuss further here). Two elements that are expounded by punctuation marks in written text are the starter and the ender. These can occur in any unit. The starter only occurs in an embedded unit, and it is expounded by a comma, a dash or a bracket, while the ender occurs with both embedded and unembedded units, and it is expounded by a wide range of punctuation marks — but most frequently by a full stop at the end of a clause (as shown in Appendix B). And there are equivalent exponents of the starter and ender in spoken text.¹⁹

¹⁹ The meanings of intonation are realised, like the meanings of punctuation, in items (in a broad sense of the term) including those that expound the starter and the ender. For a general account of the model of intonation in the Cardiff Grammar, see Tench (1996), and for a description of the first stage of its implementation in COMMUNAL, see Fawcett (1970).

Blogger Comments:

[1] To be clear, in SFL Theory, punctuation marks are a component of graphology, the systems of the expression plane for written mode. In Fawcett's model, punctuation is incongruously located at the level of syntax, which is a model of the content plane of language.

[2] To be clear, here Fawcett introduces two elements named for their structural position, starter and ender, despite having previously (pp213-4n) criticised the labels 'pre-deictic', and 'pre-numerative' as "purely positional labels". Moreover, neither of these functional syntactic elements is realised by syntactic forms.

[3] To be clear, this is a bare assertion, unsupported by evidence or argument. Moreover, if tone groups were structured with demarcative elements, such elements would only consistently demarcate boundaries of the information unit, a unit which does not feature in Fawcett's model of syntax. 

Further, the distribution of tone groups (realising information units), the system of TONALITY, is only one of three systems of intonation, the others being TONICITY (tonic placement, realising the focus of New information), and TONE (pitch movement realising interpersonal distinctions).

Fawcett's 'Linker' And 'Inferer' Elements

Fawcett (2010: 215):

However, there are a few exceptions to the basic principle, i.e., cases where an element occurs in several different classes of unit. But in all such cases this is because a higher principle is at work, as we shall see. The exception that is met most frequently is the element Linker (&) — or linker (with a lower case "l") in groups and clusters. This is expounded by and, or, etc, and such items occur at (or near) the start of any unit. (However, some of the items that expound them, such as but and so, only occur with some units.)

The second item of this type is the inferer element. This may occur in any class of group, and is expounded by even, only and just. (When these items occur in a clause with a similar meaning they expound the Inferential Adjunct.)

Blogger Comments:

[1] To be clear, in SFL Theory, Fawcett's 'Linker' and 'linker' serve the same function, differing only in rank. 'Linker' marks paratactic relations between clauses in clause complexes, while 'linker' marks paratactic relations between groups and phrases in group and phrase complexes. Fawcett does not acknowledge a rank scale in his model.

[2] To be clear, in SFL Theory, in this function, the adverbs even, only and just serve as the Head elements of adverbial groups which serve as modal Adjuncts of intensity: counter-expectancy in the interpersonal structure of the clause; see Halliday & Matthiessen (2014: 189). Fawcett's 'Inferential Adjunct' is not discussed elsewhere in this publication.

Giving Priority To Form Over Function

 Fawcett (2010: 215):

The principle that each element in a unit realises a different type of meaning brings with it the great advantage that it makes text analysis easier. This is because, once you have worked out what element an item expounds, you typically also know automatically what class of unit the element belongs to. This advantage applies both to human text analysts (perhaps consulting a summary of the syntax of English such as Appendix B) and to a computer parser of text-sentences, e.g., as built for the COMMUNAL Project (Weerasinghe & Fawcett 1993 and Weerasinghe 1994).

Blogger Comments:

[1] This is misleading, because it falsely implies that this principle holds for Fawcett's Cardiff Grammar, but not for SFL Theory. To be clear, in SFL Theory, each of the clause elements Senser, Process and Phenomenon, for example, realises a different type of meaning.

[2] This nicely demonstrates, once again, that the Cardiff Grammar gives priority to form over function. In this case, the functional element is merely the means of ascertaining the class of formal unit.

Confusing Stratal Relations With Structural Relations

 Fawcett (2010: 214-5):

The second important consequence is that, since every syntactic unit realises a different class of semantic entity, we should expect that, in principle, every element in every class of unit will be different from every element in every other class of unit. Thus we shall not use the terms "modifier" and "head" for more than one class of unit, as some form-centred grammars do (sometimes supplemented by other general, quasi-functional terms such as "complement", "adjunct" and "specifier").  

Here the terms "head" and "modifier" are used only for nominal groups. So elements in the structure of other groups that may at first appear to be partially equivalent to the "modifier' and "head" of a nominal group are given different names, e.g., "temperer" and "apex" in the quality group. The reason is simple: they express different meanings. 

Thus the head of a nominal group such as large houses tells us the 'cultural classification' of the object in terms of the culture associated with the language, and the modifier large tells us 'what sort' of thing it is (here, specifically 'what size' it is). But the apex in a quality group such as very stupid tells us that the 'quality' is 'stupid' and the temperer very tells us the 'quantity' of that 'quality'. 

The important generalisation is that each element of structure appears — in principle — in one (and only one) class of unit, and this holds good both within each unit and across the different classes of unit.

Blogger Comments:

[1] To be clear, this confuses a stratal relation (between syntactic unit and semantic entity) with the structural relations between elements in a unit. Elements of a unit function relative to other elements in the same structure, and such structures may be multivariate or univariate. As previously explained, in a univariate structure the relation between elements is iteratively one of modification, and this is why, in SFL Theory, it is applied across different classes of unit, such as nominal, verbal, adverbial, conjunction and preposition groups.

[2] This is misleading, because the implication here is that SFL Theory is a form-centred theory and the Cardiff Grammar is not. As this blog has demonstrated over and over, unlike SFL Theory, Fawcett's systemic-functional syntax gives priority to form over function (and structure over system).

[3] This is misleading, because the "quasi-functional" terms Complement and Adjunct name elements in both Fawcett's functional syntax and SFL Theory.

[4] To be clear, this creates a theoretical inconsistency since it posits different terms for the same structural relation, as a result of confusing stratal relations with structural relations; see [1].

[5] To be clear, this confuses the logical (univariate) structure of nominal groups, Modifier ^ Head, with the experiential (multivariate) structure, Epithet ^ Thing, and mistakes the meanings (semantics) they realise for culture (context).

[6] To be clear, this simply confuses degree (very) with quantity.

[7] To be clear, this clause is tautological, since it merely repeats the meaning of the previous clause, but as if it were new information.

Confusing Formal Unit With Functional Element

Fawcett (2010: 214, 214n):

The basic principle is, as stated above, that an element is defined in terms of its function in expressing meaning. This has two important consequences for describing languages.

The first is the principle that every element in a given class of unit serves a different function in that unit. Sometimes, as in the case of the different types of modifier in an English nominal group, the differences between the functions of elements appear to be very slight — but they are always there.¹⁸ Sometimes, when an element serves one of several functions (three at least in the case of the head of a nominal group), a more general — but still functional — label such as "head" must be used.  

¹⁸ If there is no differences at all between the function served by two units, as with scrumptious and delicious in (That was) a superbly scrumptious, dead delicious sweet, they should be treated as two co-ordinated units that fill the same element.

Blogger Comments:

[1] On the one hand, from the perspective of SFL Theory, this confuses unit with element. An element doesn't serve different functions, because an element is the function served by a unit.

On the other hand, from the perspective of SFL Theory, this principle fails to deal with units such as the following clause, in which the element 'Location' features three times, and the element 'Adjunct' appears four times:

[2]  To be clear, from the perspective of SFL Theory, Fawcett's "different types of modifier" confuses two distinct types of structural relations: univariate with multivariate. A univariate structure is the iteration of the same functional relationship, modification, whereas it is a multivariate structure where each element has a distinct function within the whole. Halliday & Matthiessen (2014: 390):

We refer to this kind of structure as a univariate structure, one which is generated as an iteration of the same functional relationship (cf. Halliday, 1965, 1979): α is modified by β, which is modified by γ, which is ... . By contrast, the type of structure exemplified by Deictic + Numerative + Epithet + Classifier + Thing we call a multivariate structure: a configuration of elements each having a distinct function with respect to the whole.

[3] Again, from the perspective of SFL Theory, this confuses unit with element. An element doesn't serve different functions, because an element is the function served by a unit.

[4] Trivially, unlike the Cardiff Grammar, SFL Theory does distinguish between the two wordings

• a superbly scrumptious, dead delicious sweet, and
• a superbly scrumptious and dead delicious sweet (co-ordination):