Not Recognising The Stratification Of Content: Dictionary Definitions

Martin (1992: 273):

The folk-conception of a dictionary entry as a "form" and the definition as its "meaning" is of course a misleading one.  In fact the relationship between words and definitions is not stratal (i.e. words are not being associated with the semantic primes they realise).  All that is going on is that one wording is being reformulated as another.

Blogger Comment:

Halliday & Matthiessen (2004: 237) see this rather differently, because they apply SFL theory to definitions:

Naming and defining are linguistic exercises, in which the word is Token and its meaning is its Value. … 

a ‘gazebo’	is	a pavilion on an eminence
Token		Value

(= ‘the word gazebo means [names, is defined as] a pavilion on an eminence')

Misconstruing The Context As Register

Martin (1992: 271-2):

Something of the role played by lexical relations is illustrated in [5:1] below.  A mother is talking with her 10 year old son about his experiences at school earlier in the day.  Towards the end the mother realises that she may have misinterpreted the field, and interrupts to clarify the situation.  It is the lexical cohesion that has been misread, not the NEGOTIATION, IDENTIFICATION or CONJUNCTION. …

Meanwhile it suffices to comment that it is the relations between lexical items, not lexical items per se that have been misunderstood.  And the misreading of these relations is the misreading of the text's field. … This means that although considerations of register have been set aside as far as possible in Chapters 2, 3 and 4, with lexical cohesion the register variable field needs to be brought more clearly into the picture.

Blogger Comments:

[1] In the 33-line text, the child begins talking about choosing teams in kickball then switches to playing baseball, upon which the mother realises she had mistaken kickball for soccer.  The first field is identified in the second clause:

In s…kickball we have two captains, right?

and second field is identified in the 25th clause:

I got a home run today.

It is not the lexical cohesion that has been misunderstood — not the lexical relations of repetition, synonymy, hyponymy, meronymy and collocation.  The mother misread the first field because she misheard 's…kickball' as 'soccer'.  It was the experiential meaning that was misunderstood, not the textual cohesion.

[2] Given what is to come, it is worth mentioning that this use of situation and field is consistent with SFL theory.  The situation is the instance of the context that is realised by the text, and the field is the ideational dimension of context.  Note that this is second order field — what the language is about —which Martin later (op cit: 292) dismisses as less 'appropriate' than his conception of first order field.

[3] This misconstrues context as register.  To use Martin's words (p260): 'this short-coming is a very serious one indeed'.  In SFL theory, field is not a register variable, field is a dimension of context, and context is more symbolically abstract than language.  Register, on the other hand, is not more abstract than language; register is a text type: a functional variety of language (not of context) that varies with the situation type (context) being realised.

The relation between language and context is realisation: an identifying relational process.  The relation between language as system (potential) and language as register (subpotential) is instantiation: an ascriptive relational process.

The relation between context (situation type) and register (text type) is realisation.  The lower order of symbolic abstraction, text type (language), realises the higher order, situation type (context).

Misrepresenting The Chapter 4 Model Of Conjunction

Martin (1992: 271):

To this point, the point of departure for exploring discourse systems has been a grammatical one: general MOOD classes (realising NEGOTIATION), closed system nominals (tracking participants) and clause complex structures (used extensively to code conjunctive relations in spontaneous spoken monologue).  In this chapter on the other hand the point of departure is lexis.  It is the contribution made by open system items to discourse structure that is under consideration.  This is an ambitious undertaking, in at least two respects.  First, lexis has received less attention in functional linguistics than has grammar, and so there is less to build on… .  And second, the scope of experiential meaning coded through lexis in any language is vast, which fact alone makes it harder to bring under analytic control.  Nevertheless, lexical relations have an important role to play in discourse structure.

Blogger Comments:

[1] This misrepresents what was done in chapter 4.  Non-trivially, the chapter applied Halliday's logical semantic relation of expansion — not clause complex structures — without distinguishing whether it was deployed structurally in logical complexes, or non-structurally to cohesively mark textual transitions.  Trivially, the model was applied to written as well as spoken mode, and to dialogue as well as monologue.

[2] It is true that the systems of sufficient delicacy to account for lexis have not been developed in SFL theory.  As Halliday and Matthiessen (2004: 46) have estimated:

It would take at least 100 volumes of the present size to extend the description of the grammar up to that point [of maximum delicacy] for any substantial portion of the vocabulary.

On the other hand, SFL theory models the discourse function of lexis as lexical cohesion, in which the types of lexical relations include those of repetition, synonymy, hyponymy, meronymy and collocation.

[3] In SFL theory, the experiential meaning "encoded through lexis" is brought "under analytic control" by

• construing meaning as a higher level of abstraction than the lexicogrammar, and 
• modelling lexis as most delicate grammar, such that grammatical systems represent the most general of the systems whose ultimate specification is lexical items.

Confusing Cohesive Relations With Logogenesis

Martin (1992: 267):

A further parallel has to do with the distinction between reference to participants as opposed to text reference and external as opposed to internal conjunction:

TEXT REFERENCE : INTERNAL CONJUNCTION ::

PARTICIPANT REFERENCE : EXTERNAL CONJUNCTION

Both text reference and internal conjunction construct text as part of rather than as an interpretation of the world. 

Blogger Comment:

In SFL theory, 'constructing text' is logogenesis, the unfolding of the text at the instance pole of the cline of instantiation as systemic potential is actualised though the selection of features and the activation of realisation statements.  The ideational selections are a construal of experience of the world as meaning.  The text is part of the world both materially (e.g. soundwaves, pages, pixels) and semiotically (the content plane that realises the cultural situation).

Reference and conjunction, on the other hand, in SFL theory, are systems of the textual metafunction on the content stratum of lexicogrammar.  They are resources for relating elements in the text through cohesive relations rather than by structural means.

Misrepresenting The Discourse Systems Of Conjunction & Identification

Martin (1992: 265-6):

In this chapter, an account of a third major discourse system, CONJUNCTION, has been presented; in addition the relatively minor system of CONTINUITY was briefly reviewed.  CONJUNCTION represents the discourse semantics of clause complex relations, much as IDENTIFICATION underlies nominal group DEIXIS and NEGOTIATION underlies MOOD.  The association of these discourse systems with metafunctions and their unmarked realisation in lexicogrammar is summarised in Table 4.16.

Blogger Comments:

[1] The discourse system of continuity is presented as system of the logical metafunction.  It was demonstrated in earlier posts that most markers of continuity were not conjunctions marking logical relations, but adverbs serving as adjuncts that realise interpersonal functions.  (It was also shown that elements serving experiential and textual functions were also misconstrued as continuity items.)

In SFL theory, continuity refers to a cohesive system, a non-structural resource of the textual metafunction, that marks relations by means of continuatives.  (In the discourse semantic model of conjunction, this SFL textual system was subsumed under the logical relation of addition.)

[2] Contrary to this claim, the proposed system of conjunction does not represent the discourse semantics of clause complex relations.  It was demonstrated in previous posts that the model does not discriminate between structural logical relations between clauses in a clause complex and nonstructural textual relations of conjunctive cohesion between (groups of) messages.  In the former case, it was shown that the model represents a logically invalid reorganisation of clause complex relations; in the latter case, since it is not clause complexes that are being modelled, it is not a model of clause complex relations.

[3] Contrary to this claim, as demonstrated in previous posts, the system of identification is not a discourse model of the use of deixis in the service of cohesive system of reference.  Whereas grammatical reference is a system for marking the identifiability of elements in a text, discourse semantic identification is instead concerned with tracking the persistence of specific instantial participants through a text.

[4] Contrary to the theoretical architecture of SFL theory and the principle of symbolic abstraction, this use of the term 'underlies' places the higher level of symbolic abstraction (the discourse semantic systems of identification and negotiation) below the lower level of symbolic abstraction (the grammatical systems of deixis and mood).

[5] In SFL theory, cohesive conjunction is a system of the textual metafunction, not the logical metafunction.

[6] This again confuses markedness — as in: Subject is the unmarked Theme in declarative clauses — with congruence: non-metaphorical realisations of the semantics in the lexicogrammar.

Confusing Textual Conjunction With Logical Relations

Martin (1992: 264-5):

The realisation of conjunctive relations within clauses causes problems for the definition of messages provided in 4.6 above where ranking, non-projected clauses were proposed as constituting the event-line in reticula.  The following messages from [4:199] can be used to illustrate the problem (nominalisations in bold-face).

[4:199] j. Let's be clear: I personally favour the initiative

            k. and ardently support disarmament negotiations to reduce the risk of war.

… It may be argued that the conjunctive relations in [4:1999j-k] are too deeply buried to worry about.  But this objection will not hold for [4:204] where they are explicitly realised through circumstantial relations (through and resulting from) in clause structure. …

[4:204] a.  Government and police sources agreed that the force's problem is lack of morale through a lack of discipline resulting from the absence of a strong figurehead.

The projection in this text is just a clause long in its written form, but can be unpacked into a three message clause complex as follows:

[4:205] (= SPOKEN PARAPHRASE of the projection in [4:204])

a.  the police force has problems

b.  because they aren't being disciplined

c.  because the force doesn't have a strong commissioner.

As can be seen from these examples, unpacking the grammatical metaphors characteristic of written text will have dramatic consequences for any text's conjunctive structure.  But not unpacking them means that a good deal of the logical structure of many texts will be missed. …

As suggested above, Rhetorical Structure Theory does not appear to have addressed this issue.  In practice, relational propositions which are realised within clauses through TRANSITIVITY relations are simply ignored. 

Blogger Comments:

[1] This confuses conjunction with logical semantic relations.  In SFL theory, conjunctive relations are the non-structural deployment by the textual metafunction of the logical semantic system of expansion.  It is merely one of the many manifestations the transphenomenal fractal types, expansion and projection, in the system.  Conjunctive relations don't obtain within a clause, but logical semantic relations do, as in the relations between the clause nucleus and other participants and circumstances.

[2] In terms of SFL theory, there is no conjunctive relation in [4:199] — there is neither a conjunction nor a conjunctive Adjunct — linking the clause complex to the opening simplex.

[3] Unpacking a metaphorical clause into a more congruent clause complex is irrelevant to the conjunctive relations in a text, because, in SFL theory, there are no conjunctive relations obtaining either in a clause or in a clause complex.

[4] The issue is ignored because it only arises because of a misunderstanding of SFL theory.

[5] Logical relations are not relational propositions.  A proposition is a speech function (interpersonal semantics) that is realised in clause mood.

Confusing Semantic Relations With Logogenesis

Martin (1992: 263-4):

Summarising these observations as discourse maxims one might come up with the following:

i.  Sort out what you've said (worry about what's next as it comes).

ii. Keep surfacing (don't get too embedded or too deeply dependent).

Whatever the moral drawn or how framed, there is abundant evidence that texts are produced progressively by speakers who constantly monitor what they say as they say it and for the most part make retrospective connections between new messages and what has already been said.  The deep regressive structures proposed by Rhetorical Structure Theory are thus inappropriate as a model of text production and need to be complemented by a more process oriented account.  The flatter, covariate structures proposed by English Text on the other hand seem more promising as a first step in the development of dynamic representations.

Blogger Comments:

This is the final section of a critique of Rhetorical Structure Theory.  

[1] This stage of the argument provides some advice on how to organise spoken texts, in terms of a moral to be drawn from observing two texts.

[2] This stage of the argument presents a commonsense observation as an investigative finding.

[3] Retrospective connections are deemed appropriate, but regressive structures inappropriate, for a model of text production, on the basis of the discourse maxims and commonsense observation.

[4] This confuses the semantic relations in a text with the unfolding of the text (logogenesis).

[5] The relative "flatness" of the English Text model reflects the fact that it is less able to account for the depth of dependencies in a text.

As argued previously, covariate "structures" are not structures.  They are units (messages) conjoined by logical relations, but the conjoined units do not form an integrated whole. (A reticulum is a representational device, not a linguistic structure.)

Maintaining That Clause Simplexes "Cause Problems For The Clause Complex Analysis"

Martin (1992: 262):

A number of passages from [4:2] will be reviewed; some of these were commented on in 4.1 above because they cause problems for the clause complex analysis, in spite of the fact that Halliday's system was specifically designed to handle the dynamics of clause combining.

Blogger Comment:

The passages that were commented on (1992: 162-3) were not clause complexes, but clause simplexes involving cohesion, and so do not "cause problems for the clause complex analysis".   Of the four, only one involved a conjunctive relation; the other three deployed demonstrative reference.  See previous critiques here and here.

Confusing Textual Relations With Logogenesis

Martin (1992: 261-2):

The text is repeated below to highlight the significance of regressive dependency and depth in RST analysis.  Note for example that the second message, Tempting as it may be, is a satellite to [4:199c], which is in turn part of a satellite to [4:199g-i], which is in turn part of a satellite to [4:199n].  Before this clause could be generated in other words the text would have to be processed backwards from its end message to one of its deepest points. …

It would perhaps be unfair to suggest that RST analysis treats [4:199] as a text that unfolds backwards from its end instead of working forwards from is beginning.  The picture is in fact more complex than this, involving both progressive and regressive dependencies.  However, it is clear that Rhetorical Structure Theory is not at all concerned with the way in which text unfolds as a process, with what has been realised taken as a point of departure for what comes next.  Somewhat ironically, given the similarity between the RST interpretation and clause complex analysis, rhetorical structures are very synoptic in orientation.  They account for text as a finished, pre-conceived, edited product.  The model has been developed mainly in the context of providing an account of written texts, so this development is hardly surprising.

Blogger Comments:

[1] This confuses textual relations within a text (the concern of RST) with logogenesis: the unfolding of text at the instance pole of the cline of instantiation.

Moreover, if 'regressive dependency' were a problem for RST, it would also be a problem for Martin's 'logical' model, as the following quote (1992: 240) makes clear:

Conjunctive structures have been treated in other words as typically anaphoric, as reference chains.  Conjunctions do sometimes point forward however.

See the (20/8/15) critique of this here.

[2] The suggestion is, more importantly, inaccurate, based, as it is, on the confusion of textual semantic relations with logogenesis.  It would perhaps be unfair to suggest that the assessment is more hypocritical than critical.

Criticising Others For Not Making The Same Mistakes

Martin (1992: 260):

The critical point is that because it so closely modelled on clause complex interdependency, Rhetorical Structure Theory cannot at present handle simultaneous conjunctive structures in text.  Since internal and external conjunction very commonly assign simultaneous layers of structure to a text, much as experiential and interpersonal metafunctions in the grammar do for the clause, this short-coming is a very serious one indeed.

Blogger Comments:

[1] As demonstrated in previous posts, the "simultaneous" conjunctive structures only arise from theoretical misunderstandings of Martin, such as 

• mistaking parataxis for addition,
• misapplying the logical semantic category concessive,
• mistaking interpersonal functions for conjunctive functions.

[2] As demonstrated in previous posts, Martin has misunderstood Halliday & Hasan's (1976) distinction between internal and external conjunction throughout, misconstruing it as the distinction between 'constructing text' and 'constructing field'.

[3] An important difference here is that the experiential and interpersonal structures of the clause are different metafunctional angles on the same unit, whereas any simultaneous conjunctive structures would be from just one angle: Martin's logical.

[4] Given the above, no shortcoming in RST has been identified, serious or otherwise.

Misconstruing Circumstantial Comparison As Conjunctive Comparison

Martin (1992: 258-9):

Equally serious is Rhetorical Structure Theory's claim that conjunctive relations in text can be reduced to a single chain.  The short-comings of this position can be demonstrated with respect to [4:3] above.  Consider first the relationship between clause complexes 2 and 3:

[4:3] 2.c. Modern capitalism thrives on expansion and credit,
           d. and without them it shrivels.
        3.e. Equally however, it requires the right context,
           f. which is an expanding world economy:
          g. a national economy is distinct and severable from other national economies in some senses but not all. 

These are simultaneously related through internal comparison and concession.  The third clause complex is assigned the same status in the argument as the second through the internal comparative conjunction equally, at the same time as it is related to it by however: 'although I've just argued that modern capitalism thrives on expansion and credit and without them shrivels, it is equally important to say that it requires the right context, which is…'.

Blogger Comments:

[1] The conjunctive relation here is neither comparison nor concession, and there is only one relation, not two.  The conjunctive relation, which is marked by however, is extension: addition: adversative ('on the other hand'), not concessive ('despite this').

[2] The adverb — not conjunction — equally functions experientially as a circumstance of Manner: comparison.  Like all circumstances of Manner, it 'construes the way in which the process is actualised' (Halliday & Matthiessen 2004: 267), as demonstrated by its unmarked textual agnate:

However, it equally requires the right context.

The circumstantial Adjunct equally also contributes to the texture structurally, as marked Theme, while the conjunctive Adjunct however contributes non-structurally, as a marker of cohesive conjunction.  (The punctuation also suggests both are packaged as contrastive New information.)

equally	however	it	requires		the right context
circumstantial Adjunct	conjunctive Adjunct	Subject	Finite	Predicator	Complement
Manner: comparison		Carrier	Process: relational		Attribute
Theme: marked	Rheme

[3] If this were the meaning, then the function of equally would be interpersonal, not logical (or textual), since it enacts an evaluation of the proposition.

Misapplying Textual Reference To Logical Dependency

Martin (1992: 240):

In all of the examples considered so far, messages have been shown to depend retrospectively on preceding ones.  Conjunctive structures have been treated in other words as typically anaphoric, as reference chains.  Conjunctions do sometimes point forward however.  This happens in correlative paratactic structures like that in [4:189] where either predicts an ensuing alternative just as or predicts a preceding one. … A small number of internal conjunctions operate in this way; for example first and on the one hand.  In [4:192] first predicts a succeeding argument, just as second presumes a preceding one.

[4:192] a.  I'd say it's a material process.
            b.  First, the tense is [present-in-present].
            c.  Second, the Agent is not conscious.
            d.  Finally, this process can't project.

Blogger Comments:

[1] In SFL theory, 'the different kinds of pointing', phora, are features of reference, a nonstructural resource of the textual metafunction.  Here they are applied to conjunction, which is construed as a structural resource of the logical metafunction.

[2] Phoric relations are here construed as obtaining between a conjunction in one message and another message within the same conjunctive structure, and the 'pointing' is interpreted as predicting prospectively (cataphora) or presuming retrospectively (anaphora).

However, whereas anaphoric and cataphoric reference point forward or backward in the text to where the identity presumed by the reference item can be recovered, in this use of the terms for conjunctive relations, the pointing forward of backward in the text is merely to the location of the conjoined message.

Misapplying Logical Nesting To Textual Relations

Martin (1992: 238):

The point is that "nesting" is a feature of conjunctive relations whether they are expressed through a clause complex structure or not.

Blogger Comment:

In SFL theory, nesting, or internal bracketing, occurs in logical structures whenever there is a change in taxis or in logico-semantic relation (Halliday & Matthiessen 2004: 382).  Cohesive conjunction, on the other hand, is a non-structural resource of the textual metafunction.

Misconstruing An Implicit Conjunctive Relation As Internal

Martin (1992: 237):

A text with a simultaneous internal and external structure, including an implicit internal consequential is illustrated below.

[4:187] a.  Ben was unlucky.

            b.  He had to take steroids for his injured hamstring

            c.  and then they introduced more sophisticated tests.

Blogger Comment:

By the SFL definition of the term — see previous posts — there are no internal relations in this example.  

In SFL theory, the logical relation in the clause complex (b. c.) is enhancement: temporal: later (rebranded as 'successive'), and the textual (cohesive conjunction) relation between the message (a.) and the message group (b. c.) is implicit enhancement: cause: reason (rebranded as 'consequential').

With regard to implicit conjunction, in general, Halliday & Matthiessen (2004: 549) recommend:

It is perhaps as well, therefore, to be cautious in assigning implicit conjunction in the interpretation of a text.  It is likely that there will always be other forms of cohesion present, and that these are the main source of our intuition that there is a pattern of conjunctive relationships as well.

Misconstruing Parataxis As Addition (Extension)

Martin (1992: 236):

Additives are treated in this way to simplify the representation since they often combine with other external relations (e.g. and so, and then, and thus etc.) and are extremely frequent in certain modes.

Blogger Comment:

In such cases, the expansion relation is determined by the second conjunction: so, then, thus.  When functioning structurally (logically), rather than cohesively (textually), the and only marks the paratactic interdependency of clauses in a nexus.  Both examples given (237, 239) are of paratactic enhancement — temporal and cause— with no additive (extension) relation:

He had to take steroids for his injured hamstring	and then they introduced more sophisticated tests
1	x 2 (temporal)

On top of all that he won his race	and so had to be tested
1	x 2 (cause: result)

The Omission Of Projection And Hypotactic Elaboration From The Logic Of Discourse Semantics

Martin (1992: 235):

The Hartford school modelled conjunctive relations between actions, more or less equivalent to the TRANSITIVITY function Process in systemic theory.  Since TRANSITIVITY relations are treated as part of lexicogrammar in the model developed here, an appropriate discourse semantics unit needs to be found.  The unit message will be adopted here, where this is realised as a ranking clause that is neither a projection, nor a hypotactically dependent elaborating clause.  This means that locutions and ideas, elaborating beta clauses and all embedded clauses will be treated as part of messages rather than as conjunctively related units in their own right.

Blogger Comments:

[1] In SFL theory, transitivity is a system of the experiential metafunction.  A consistent discourse semantics unit with 'transitivity relations' would therefore be an experiential unit.  Here, however, it is a logical discourse semantics unit that is deemed 'appropriate'.  In the grammar, it is the rank units (forms) that are logically related in complexes.

[2] In SFL theory, the term 'message' is already used for a unit of the textual metafunction on the semantic stratum, and it is (groups of) messages that are conjoined by cohesive conjunction.  Here it is used for units related by conjunction, but, as conjunction is misconstrued as logical and structural, it is misconstrued as a unit of logical structure.  That is, although textual meaning has been misconstrued here as logical, the textual term has been retained, thereby creating a further theoretical inconsistency — cf 'clause as message', the thematic (textual) structure of the clause.

[3] In SFL theory, there are two major types of logical relations: expansion and projection.  The second of these major types, projection, is thus not included as a logical relation at the level of discourse semantics.

Similarly, within expansion, there are three major types: elaboration, extension and enhancement.  The first of these major types, elaboration, is thus not included as a logical relation at the level of discourse semantics — if the interdependency relation is hypotactic.

Excluding both major logical relations has serious ramifications for considerations of the distinction between congruent and incongruent stratal relations in determining grammatical metaphor.

The absence of projection follows from the rebranding of cohesive conjunction, which is the deployment of (only) expansion relations by the textual metafunction at the stratum of lexicogrammar, as the logical dimension of a higher stratum: discourse semantics.

On the other hand, in the discourse semantic model, types of elaboration are variously misconstrued as types of extension or enhancement, as demonstrated in earlier posts.

Misconstruing A Circumstantial Adjunct As A Continuity Item

Martin (1992: 233-4):

Again is used to mark the recurrence of a Process · Medium nucleus at a later date:

[4:180] Ben beat Carl in Rome
            and Carl lost again in Seoul. 

Blogger Comment:

In SFL theory, again may function as a conjunctive Adjunct or as a circumstantial Adjunct (Halliday & Matthiessen 2004: 133).  Here it functions experientially, identifying a temporal Location ('at a later time').

Ben	beat	Carl	in Rome
Agent	Process	Medium	Location: place
	Nucleus

and	Carl	lost	again	in Seoul
	Medium	Process	Location: time	Location: place
	Nucleus

Misconstruing Residue Substitution As Continuity

Martin (1992: 233):

The use of so and too to hold the Residue constant underlies their function in polarity focussed arguments such as the following where they resemble the clause substitutes not and so; so and too are always tonic in this function (tonic syllable in bold face):

[4:178] Ben won really.
            — // He did not. //
            — // He did so. //
            — // He didn't. //
            — // He did too. //

Blogger Comments:

[1] As demonstrated in previous posts, 'holding the Residue constant' is not the function of either too functioning as a conjunctive Adjunct or so functioning cohesively as substitute for Predicator and Complement.

[2] In SFL theory, not, so and too don't just resemble cohesive substitutes, that is their function in the example provided.  Substitution of the Residue occurs with positive polarity, whereas ellipsis of the Residue occurs with negative polarity.

Ben	won		really
Subject	Finite	Predicator	comment Adjunct: speech-functional: unqualified: factual
Mood		Residue

He	did not	Ø
Subject	Finite	Predicator	comment Adjunct: speech-functional: unqualified: factual
Mood		Residue

He	did	so
Subject	Finite	Predicator	comment Adjunct: speech-functional: unqualified: factual
Mood		Residue

He	didn’t	Ø
Subject	Finite	Predicator	comment Adjunct: speech-functional: unqualified: factual
Mood		Residue

He	did	too
Subject	Finite	Predicator	comment Adjunct: speech-functional: unqualified: factual
Mood		Residue