David Rose On The Relations Between Grammar, Semantics and Text

David Rose wrote to sys-func and Sysfling on 26 March 2015 at 09:20:

That said, there is an interesting ambivalence in recent IFG editions about the relations between grammar, semantics and text, both realisational and instantial. … 

In fact IFG explicitly models ‘text’ both as the unit of analysis of the semantic stratum, consistent with discourse semantic theory, and as instantiating grammatical systems. It also discusses domains of instantiation of semantic systems as stretches of discourse, also consistent with discourse semantic theory. What it doesn’t admit is the discourse semantic systems described in English Text, Working with Discourse, and a considerable body of other scholarly work. 

My own view is that this omission has as much to do with personal institutional politics within the SFL leadership, as it does with the relative merits of the theory.

Blogger Comments:

 [1] In SFL, the word 'text' is used to refer to the highest semantic unit and to the instance pole of the cline of instantiation.  Its valeur depends on what it is contrasted with.  It is misleading to call this 'ambivalence', which means:

• uncertainty or fluctuation, especially when caused by inability to make a choice or by a simultaneous desire to say or do two opposite or conflicting things.

[2] The notion of text as a semantic unit was imported, unaltered, by Martin into 'discourse semantic theory' from Halliday's Systemic Functional linguistics; hence the "consistency".

[3] Each text is an instance of systemic potential — whatever the linguistic stratum; just as, above the linguistic strata, at the level of context, situation is an instance of culture.

[4] IFG (Halliday and Matthiessen 2004), as the name suggests, is an introduction to functional grammar, not 'discourse semantic systems'.  Halliday and Matthiessen (2004: 550-1):

for a systemic description of this [cohesive reference] as a semantic system, see Martin (1992).

Martin's English Text and Martin and Rose's Working With Discourse and 'a considerable body of other scholarly work' are also listed in the References.  

Considering how inconsistent 'discourse semantic theory' is with the core architecture of Systemic Functional Linguistics (e.g stratification and metafunction), its citing in IFG is an act of great generosity, especially in face of bullying insults, such as those of Rose quoted above (appraisal analysis here).

For a rigorous assessment of the "relative merits" of 'discourse semantic theory', including the degree to which it is consistent with (a sound knowledge of) Systemic Functional linguistic theory, see here.

David Rose On 'Semantics Instantiating Grammar'

David Rose wrote to Annabelle Lukin on sys-func and Sysfling on 26 March 2015 at 09:20:

The main problem with your argument is that it hinges on this partial quoting of me

in (b), semantics is 'merely instantiating grammar systems in text’

This section of my analysis actually discussed ‘discourse semantic systems’ not ‘semantics'

The models differ in their view of discourse semantic systems, such as conjunction, ideation, appraisal 

In a) discourse semantic systems are not possible, as semantics is consumed with grammatical functions, in b) they merely instantiate grammar systems in texts, In c) grammatical structures realise higher rank functions, but structure and function are properties of both grammatical and semantic strata

Blogger Comments:

[1] Semantics does not instantiate grammar, regardless of whether Rose is discussing semantics or 'discourse semantic systems'.  Instantiation is the relation between system and instance, not the relation between strata, such as semantics and grammar.

Using the verb 'instantiate' is problematic, because it does not serve as an attributive Process, and instantiation is an attributive relation.  Rose's use of the verb results in a construal of discourse semantic systems as less abstract, a lower stratum, than grammar systems:

discourse semantic systems	merely	instantiate	grammar systems	in texts
Token		Process: identifying	Value	Location

[2] The 'discourse semantic systems' conjunction and ideation are Martin's (1992) rebranding of two of Halliday's lexicogrammatical resources of cohesion: conjunction and lexical cohesion (see Halliday & Matthiessen 2004: 524ff).  Martin took Halliday's textual lexicogrammatical systems and reconstrued them as ideational systems — conjunction is logical, ideation is experiential — on a higher stratum of content.  Doing so violated the meaning of both stratification and metafunction.

[3] Appraisal is a genuine interpersonal semantic system realised in, and construed by, the lexicogrammar.  Viewed from below, it lies within evaluation, between grammar and lexis.  Halliday (2008: 49):

With options in the way something is evaluated (“I approve / I disapprove”) or contended (“I agree / I disagree”), the borderline between grammar and lexis is shaded over; systems of appraisal, as described by Martin & White (2005), represent more delicate (more highly differentiated options within the general region of evaluation.

Tom Bartlett On Function And Structure

Tom Bartlett wrote to David Rose on sys-func and Sysfling on 23 March 2015 at 20:13:

The comment you pick up on (and run with, if I may say so...) simply made the point that the two because stretches in the speech  [the one a phrase, the other a clause] perform a similar function with regard to the utterance as a whole in providing alternative causes for the event described and this is why they can coordinate despite their different structures (and the different kinds of meanings these structures construe as structures). As far as models are concerned, there is nothing in what I said that posits a stratal rather than realisational relationship between function and structure;

Blogger Comments:

There are several interwoven confusions here.

There is no realisational relation between function and structure.

There is no stratal relation between function and structure.

The stratal relation is one of realisation.

There is a realisational relation between strata; e.g. the lower stratum of content realises the higher stratum of content.

There is a realisational relation between system and structure; this is the relation between the paradigmatic and syntagmatic axes.

There is a realisational relation between function structure at a higher rank and the syntagms (forms) of the lower rank, as when a Process (clause rank) is realised by a verbal group complex (group rank).

What Bartlett said construes two levels of symbolic abstraction on the content plane, two strata related by realisation, even though the technical terms he uses — "stretches of speech" and "the utterance as a whole" — construes them as if in a part-whole relation.

It is this stratal relation that allows us to speak of the same meaning (semantics) being realised by different wordings (lexicogrammar), as when the meaning 'cause' is variously realised as a dependent clause or as a prepositional phrase.

David Rose On Grammatical Metaphor As Stratal Tension

David Rose wrote to sys-func and Sysfling on 23 March 2015 at 19:29:

Another thing that’s interesting in both the ‘re-mapping’ and ‘stratal tension’ models, is the role of lexis. For example, in IFG 10.5.2 there are many examples of ideational metaphors ‘unpacked’ to congruent clause complexes. Like Obama's example…

America has carried on | Circ:cause not simply because of the skill or vision of those in high office

-> America has carried on || not simply because those in high office are very able || or see clearly

Recognising lexical relations such as skill-able and vision-see is necessary to interpret the metaphors, but lexical relations tend to be backgrounded in discussions of grammatical metaphor. To me they indicate that discourse semantic systems like ideation are in tension with grammatical systems like transitivity. It is the lexical item ‘vision’ that construes the process of seeing, and ‘skill’ that construes the quality able, while the grammar construes them as possessions of those in high office.

Blogger Comments:

[1] Lexical relations are a resource of the textual metafunction (the text-forming resources), so it's hardly surprising they are "backgrounded" in discussions of ideational metaphor (incongruent construals of experience).

[2] The relation between lexical relations (cohesion: non-structural textual grammar) and transitivity (structural experiential grammar) is not one of 'stratal tension', since both are construals of the same stratum: lexicogrammar.

Regarding the rebranding of Halliday's cohesion as 'discourse semantics', I have just started a new blog called Discourse Semantic Theory (here) which demonstrates how and why 'discourse semantics' is inconsistent with the theoretical architecture of SFL.  This will be done by paying close attention to its major exposition, page by page, in Martin's English Text.

David Rose On Major Theoretical Differences Between 3 SFL Models: Grammatical Metaphor

David Rose wrote to sys-func and Sysfling on 23 March 2015 at 08:35:

The models also differ in their treatment of grammatical metaphor

In c) as ‘stratal tension’ between discourse semantic and grammatical systems (both contributing to meaning of metaphors), so conjunction is realised congruently as clause complexity, and incongruently as transitivity or circumstantiation 

In b) as ‘re-mapping’ between semantic meanings and their grammatical realisations, prompting the treatment of an ideational metaphor as clause rank complexing (‘not simply because the vision…') 

In a) Tom’s comment suggests that the transferred meaning is treated as function, and the grammatical realisation as structure.

Blogger Comments:

Rose presents the difference between how Martin's derived model (c) and Halliday's original model (b) treat grammatical metaphor as a distinction between a 'stratal tension' and a 'remapping' between content strata, but since both involve the distinction between congruent and incongruent grammatical realisations of semantic features, 'stratal tension' is largely Martin's rebranding of Halliday's idea in less precise terms.  In this regard, there is the question of what a model of semantics needs to provide in order to be able to distinguish congruent from incongruent realisations in the grammar and whether the 'discourse semantics' model satisfies those requirements.

However, Rose implies, by inclusion vs omission, that it is only in Martin's derived model (c) that both strata contribute to the meaning of metaphors.  Of course, given that grammatical metaphor, like all semogenesis, involves a realisation relation between strata, it is nonsensical to say that one stratum does not "contribute".

More importantly, as Halliday & Matthiessen (1999: 283) point out, the metaphorical form is junctional: it also embodies semantic features deriving from its own incongruent lexicogrammatical properties.  That is, grammatical metaphor is a means of simultaneously construing the meanings of both the congruent and incongruent grammatical realisations.

These two meanings are themselves in an elaborating token-value relation within the semantic stratum, with the metaphorical Token realising the congruent Value (Halliday & Matthiessen 1999: 288).

David Rose On Major Theoretical Differences Between 3 SFL Models: Discourse Semantic Systems

David Rose wrote to sys-func and Sysfling on 23 March 2015 at 08:35:

The models differ in their view of discourse semantic systems, such as conjunction, ideation, appraisal

In a) discourse semantic systems are not possible, as semantics is consumed with grammatical functions, 

in b) they merely instantiate grammar systems in texts, and 

in c) they are distinct systems, where grammar and discourse both contribute to meaning

Blogger Comments:

[1] These 'discourse semantic' systems could easily be incorporated in Fawcett's Cardiff Grammar, as systems (functions) on the higher stratum of context, and as structures (forms) on the lower stratum.  Given that some of them derive from Halliday's grammatics (textual cohesion), this may already be the case. 

[2] This reflects Rose's ongoing confusion of instantiation with realisation (see elsewhere on this blog).  Semantic systems do not instantiate grammar systems ('in texts' or anywhere else).  Semantic systems are realised by lexicogrammatical systems.

Regarding the systems Rose mentions, while 'appraisal' is a genuine interpersonal semantic system, construed by the interpersonal grammar — the word 'discourse' is redundant — the others are a mixed bag. For example, Martin's logical 'discourse semantic' system of conjunction derives, at least in part, from Halliday's textual grammatical system of conjunction, and Martin's experiential 'discourse semantic' system of ideation derives both from Halliday's (non-structural) textual grammatical system of lexical cohesion and his (structural) experiential grammatical system of transitivity.

[3] As [2] indicates, integrating these diverse systems as Martin's 'discourse semantic' systems creates theoretical inconsistencies in terms of both metafunction and level of symbolic abstraction.

[4] This reflects Rose's ongoing confusion of semogenesis with stratification (see elsewhere on this blog).  All strata make meaning, in the sense of semogenesis, but stratification is a means of parcelling out the complexity of language as a hierarchy of symbolic abstraction, in which meaning (semantics) is realised by wording (lexicogrammar).  Does discourse "contribute" to meaning, or is it a semogenic process?

David Rose On Major Theoretical Differences Between 3 SFL Models: Structure And Function

David Rose wrote to sys-func and Sysfling on 23 March 2015 at 08:35:

This pointed to some major theoretical differences between 3 SFL models (that "make different and often opposing assumptions”)

In model a) structure and function are treated stratally, structure as grammar and function as semantics 

In b) structure and function are treated by rank - grammatical structures realise functions at higher ranks, and semantics is the meanings realised by grammar 

In c) grammatical structures realise higher rank functions, but structure and function are properties of both grammatical and semantic strata

Blogger Comments:

[1] The 3 SFL models that Rose has in mind are Halliday's original model (b), and two variations derived from it: Fawcett's Cardiff Grammar (a), and Martin's Discourse Semantics (c).

[2] 'Structure' and 'function' are each one half of two separate dichotomies, each involving the relation of 'realisation'.  On the one hand, 'structure' contrasts with 'system' — this is the relation between the syntagmatic and the paradigmatic axes.

On the other hand, 'function' contrasts with 'form' — this is the relation of symbolic abstraction that obtains between strata, and, in the rank scale, between elements of function structure, such as Senser^Process, and syntagm that realises them, such as nominal group^verbal group.

[3] In Fawcett's derived model, content system and structure are distinguished stratally.  Halliday & Matthiessen (1999: 429):

In Fawcett’s model, there is only one system–structure cycle within the content plane: systems are interpreted as the semantics, linked through a “realisational component” to form, which includes items and syntax, the latter being modelled structurally but not systemically…

[4] In Halliday's original model, system and structure are modelled at both strata of the content plane.  Halliday & Matthiessen (1999: 429):

…in our model there are two system-structure cycles, one in the semantics and one in the lexicogrammar. Terms in semantic systems are realised in semantic structures; and semantic systems and structures are in turn realised in lexicogrammatical ones.

[5] Grammatical 'structures' do not "realise higher rank structures".  Syntagms (forms) of a lower rank realise function structures of a higher rank.

[6] Rose's claim that only Martin's derived model posits "structure and function" on both content strata is demonstrably false.  For structure on both strata in the original model, see [4]; for function on both strata, consider, inter alia, the metafunctional organisation of both semantics and grammatics.