Saturday 30 April 2016

Confusing Text Type (Genre) With Text Structure (Semantics)

Martin (1992: 503):
Approaching genre from a teleological perspective is also useful in accounting for the way in which texts typically move through stages to a point of closure and are explicitly treated by speaker/listener as incomplete when closure is not attained (having mentioned closure it is important to stress that genre, like all semiotic systems, is a dynamic open system (see Lemke 1984) and so in [sic] constantly evolving;

Blogger Comments:

[1] This is more true of clauses than it is of genres (text types), and yet, clause structure can be usefully accounted for without approaching the clause from a teleological perspective.  All repetitions of a previously established process have a previously established endpoint.  This does not necessitate the adoption of a teleological perspective.

A functional theory of language is concerned with functions, not purposes — the functions of clauses, the functions of text types, and so on.

[2]  Note the confusion of text type (genre) with text structure (semantics).

[3] This confuses the termination of a process (the closure of a text) with the boundary conditions (open or closed) of dynamic systems.  A closed system is isolated from its environment, whereas an open system interacts with its environment and derives energy from it.

[4] To be clear, genres are types of semiotic systems.

Friday 29 April 2016

Misrepresenting Purpose And Intention

Martin (1992: 503):
It should be stressed here that bringing telos into contextual theory at this point in no way implies that text is being interpreted as the realisation of speakers' intentions; genres are social processes, and their purpose is being interpreted here in social, not psychological terms. Nor does the model imply that the cultures as a whole are goal-directed, with some over-riding purpose governing the interaction of social processes. Social processes negotiate with each other and evolve, as noted above in the motivation for a level of ideology superordinate to genre and register. The metaphor of intentionality, in other words, is just as inappropriate for explaining why a culture has the social processes it does as for explaining why an individual speaker produces certain kinds of text. With these qualifications in mind, the notion of telos is a useful one for glossing systemic relations between combinations of field, mode and tenor choices at the level of genre.


Blogger Comments:

[1] It is useful here to consider the congruent construals of 'purpose' and 'intention' in the lexicogrammar of English.  Purpose is an enhancing expanding relation between processes (because intention Q so action P), in which a desiderative mental process of intention is the reason for another process.  That is, intention and purpose congruently involve mental processes.  Speakers undergo mental processes, genres do not.  The purposes of genres are the purposes of their speakers, whether viewed linguistically, socially or psychologically.

The functions of genres, on the other hand, are modelled in SFL theory by the contextual features (field, tenor and mode) of the situation type that the particular text type realises.

[2] This confuses orders of experience.  Genres, as processes, are types of semogenic processes; types of logogenesis.  Social processes are of the material order of experience, semogenic processes are of the semiotic order of experience.

[3] Social processes do not "negotiate" with each other.  Speakers negotiate through the interpersonal metafunction of language.  The metaphor is a misleading one.

[4] This continues the misunderstanding of stratal relations as hyponymic.

[5] This is manifestly untrue.  Intentionality — metaphorical or otherwise — can indeed explain 'why an individual produces certain kinds of text'.  For example, the intention of speaker A to learn something from speaker B about grammar explains why speaker A produces texts involving questions about grammar.

[6] The utility of the notion of telos has not been demonstrated. It has merely been asserted without supporting argument.

Thursday 28 April 2016

Assigning Purpose To Theoretical Dimensions

Martin (1992: 502-3):
The register variables field, tenor and mode can then be interpreted as working together to achieve a text's goals, where goals are defined in terms of systems of social processes at the level of genre.

Blogger Comments:

[1] In SFL theory, field, tenor and mode are the metafunctional dimensions of context, not register.  A register realises the field, tenor and mode of a situation type.

[2] In SFL theory, field, tenor and mode do not "work" and texts do not have goals:
  • at the instance pole of the cline of instantiation, field, tenor and mode are dimensions of the situation that is realised by a text;
  • it is speakers and writers who have goals; see the next post.

the register variables field, tenor and mode
can
then
be interpreted
as
[[[working together || to achieve a text's goals]]]
Token/Identified
Process:

relational

Value/Identifier

Value realised by clause nexus of purpose:

working
together
to achieve
a text's goals
a
x b
Process: material
Accompaniment
Process: material
Goal



cf.
We
can interpret
the register variables field, tenor and mode 
as
[[[working together || to achieve a text's goals]]]
Assigner
Process: relational
Token/Identified

Value/Identifier

Wednesday 27 April 2016

Problems With The Non-Argument For Genre As Context

Martin (1992: 502):
As noted above, the fact that notions of purpose and effect do not correlate with any one metafunctional component of language and have been associated with different variables in the development of register theory suggests that a teleological perspective on text function might be better set up as superordinate to — rather than alongside or incorporated in — field, mode and tenor.

Blogger Comments:

[1] This is manifestly untrue.  Notions of purpose (Fawcett 1980) and effects (Firth 1950) were not 'associated with different variables in the development of register theory'.  Each was featured in  early models of context.

[2] This misunderstands stratification.  The term 'superordinate' misconstrues relations between strata as a relationship of hyponymy (delicacy), instead of symbolic identity (realisation); see Halliday & Matthiessen (1999: 145).

[3] The argument here for establishing genre as a contextual stratum is as follows:
  • Premiss 1: Because notions of purpose and effect do not correlate with one metafunction only, and
  • Premiss 2: because purpose and effect have been associated with different variables in the development of register theory,
  • Conclusion: it follows that a teleological perspective on text function might be modelled as "superordinate" to field, mode and tenor.
As noted in [1] and [2] above, Premiss 2 features an untruth, and the Conclusion mistakes stratification for hyponymy.  Leaving these aside, the conclusion does not follow from the premisses. This is because the number of metafunctions and the different theorisings of purpose and effect are entirely irrelevant to the question of levels of symbolic abstraction in the stratificational hierarchy.

[4] Teleology has its Western philosophical origins in Aristotle, who 'believed in purpose as the fundamental concept in science' (Russell 1961: 90).  However, it ceased to be part of scientific explanation with the rise of science in the 17th Century.  As Bertrand Russell (1961: 523) makes clear:
Another thing that resulted from science was a profound change in the conception of man's place in the universe. … Moreover purpose, which had since Aristotle formed an intimate part of the conception of science, was now thrust out of scientific procedure. … The world might have a purpose, but purposes could no longer enter into scientific explanations.

Tuesday 26 April 2016

Problems With The Non-Argument For Register As Context

Martin (1992: 501-2):
Before beginning however, it is important to note that English Text extends the use of the term register as defined by Halliday.  Halliday uses the term simply to refer to language as context's expression plane — the linguistic meanings (entailing their expressions) at risk in a given situation type.  English Text extends the notion to cover in addition part of context's content plane; register is used in other words to refer to the semiotic system constituted by the contextual variables field, tenor and mode.  As outlined above, in the model of context developed here, register is the name of the metafunctionally organised connotative semiotic between language and genre.  This means that instead of characterising context of situation as potential and register as (context's) actual, English Text treats register as a semiotic system in its own right, involving notions of both system and process.


potential (system)
actual (process)
Halliday (1978)
context of situation :
register ::
English Text
register :
language

Blogger Comments:

[1] No argument is offered that identifies any problems with Halliday's model of register, or as to why Martin's model is to be preferred.  The new model is merely announced, as if a new consumer good.

[2] Martin previously used the term 'meanings at risk' in conjunction with modal responsibility, which in SFL theory, is the meaning of Subject, an interpersonal function of the clause.  Here Martin uses the term to misrepresent Halliday's notion of register, the subpotential of language (all metafunctions) that realises a situation type.  That is, the use of the term is both inconsistent within Martin's own model, and a misconstrual of Halliday's model.

[3] This claim is both inconsistent with what follows, and a misunderstanding of the principle of stratification.  It is inconsistent with what follows because register is simply equated with Halliday's context.  It is inconsistent with stratification because it construes register as both language and context.

[4] This confuses context with language and stratification with instantiation.  To equate register with context is to construe it as more abstract than language (stratification) instead of as a subpotential of language (instantiation) — a major category error that makes the model untenable.

[5] This confuses stratification with instantiation.  Situation type is related to register by realisation, whereas the relation of potential to actual is the relation between system and instance (text).

[6] The implication here is that Halliday's model does not treat register as a semiotic system in its own right.  To be clear, in SFL theory, each register is a semiotic system that realises a situation type.

[7] This misunderstands the meaning of system and process in SFL theory.  The term 'system' is itself shorthand for system–&–process.  Halliday & Matthiessen (1999: 507):
As we conceive of it, the term “semiotic” is framed within a linear taxonomy of “physical – biological – social – semiotic”; and the term “system” is a shortened form of “system–&–process”, there being no single word that encapsulates both the synoptic and dynamic perspectives (we have referred to the term “climatic system” with the same observation on how it is to be understood).
[8] Here the false dichotymy of system/process (see [7]) is misconstrued as two levels of symbolic abstraction.  Note that, previously, Martin has interpreted process as (dynamic) structure

[9] This misconstrues the relation between language–as–system and language–as–register (instantiation) as:
  • register is realised by language, and
  • register as potential, and language as instance.
These are inconsistent with each other, and the second reverses the relation between language and register on the cline of instantiation, relocating register from the midway point to the system pole, and language from the system pole to the instance pole.


Conclusion:
  1. No argument is offered to support the model; 
  2. the model is inconsistent with itself across several dimensions; 
  3. the model is inconsistent with a sound knowledge of the architecture of SFL theory.

Monday 25 April 2016

Self-Contradiction

Martin (1992: 501):
In broad terms the development of work on register and context reviewed above can be interpreted in terms of a gradual de-materialising and concomitant semioticising of frameworks for relating language to situation (see in particular Hasan's 1977, 1979, 1985/1989 use of contextual features to generate text structure).

Blogger Comment:

This is inconsistent with Martin's own claim (p497) that the early work of Firth (1950, 1957b, 1957c) developed context 'more abstractly as a level of language'.  The confusion of semiotic context with material setting is Martin's, as demonstrated here, here, here and here.  The material and semiotic are distinct orders of experience.

Sunday 24 April 2016

Confusing Context With Text Type

Martin (1992: 501):
Overall it would appear that "rhetorical purpose" is the wild card in contextual description, being variously categorised under field (Halliday 1965), tenor (Gregory 1967), mode (Halliday 1978, 1985/1989) and as a separate contextual variable in its own right (Firth 1950effects, Ure & Ellis 1977 — rôle, Fawcett 1980 — pragmatic purpose). The main reason for this is that purpose is difficult to associate with any one metafunctional component of the lexicogrammar or discourse semantics.  The effect of a text is the result of all components of its meaning.  This makes associating the notion of rhetorical purpose with Bakhtin's more global notion of speech genres an attractive one (cf. Gregory 1982).

Blogger Comments:

[1] The introduction of the word 'purpose' here is misleading.  It is falsely presented as a common feature of the various models of context, despite it being restricted to just one (Fawcett 1980).  This unwarranted fudge invalidates Martin's argument.

[2] Here purpose is not distinguished from effect.  This confuses two types of cause: purpose and result.  The meaning of purpose is 'because intention Q, so action P', whereas the meaning of result is 'because P so result Q'.  This confusion also invalidates Martin's argument.

[3] Martin's main confusion here is between mode, the rôle played by language, and genre a type of text.  This is a confusion of stratification (mode is a dimension of the context stratum) with instantiation (a genre is a subpotential of language).  The confusion arises from not distinguishing text types from the contextual features — such as mode: narrative — by which they are identified.

Saturday 23 April 2016

Confusing Context With Semantics

Martin (1992: 500):
At the same time however it is important to note the uncertainty in Halliday's writing as to whether mode is meant to cover genre.  At times this connection is explicitly made (e.g. "mode covers roughly Hymes' channel, key and genre" 1978: 62).  Elsewhere Halliday would appear to disassociate genre from any one contextual variable (e.g. In the most general terms there are two other components of texture.  One is the textual structure that is internal to the sentence: the organisation of the sentence and its parts in a way which relates it to its environment.  The other is the 'macrostructure' of the text, that establishes it as a text of a particular kind — conversation, narrative, lyric, commercial correspondence and so on" Halliday & Hasan 1976: 324).

Blogger Comments:

There is no uncertainty on this matter in the quotes presented above.  The first quote is about the textual metafunction at the level of context, whereas the second quote is about the textual metafunction at the level of semantics.  That is, each quote is concerned with a different level of symbolic abstraction.

In the first quote, rhetorical mode is identified with Hymes' notion of genre, which, in SFL theory, is the part language plays in the situation type.  

The second quote, on the other hand, is concerned with the textual structure (semantics) of a text type ("genre") that realises a particular rhetorical mode (situation type).

The dimensions in play here are
  • stratification: context versus semantics, and
  • instantiation: situation type as subpotential of context, and text type as subpotential of semantics.

Friday 22 April 2016

Purpose, Genre And Register

Martin (1992: 499):
Instead, taking Halliday's (1978) model [of context] as a baseline, the problem of purpose will be explored, since it bears critically on the relationship between genre and register to be further developed below.

Blogger Comments:

[1] In SFL theory, the "purpose" of a functional variety of language is modelled in terms of the rhetorical mode of the situation type that the language realises.  Mode is concerned with the rôle that language plays.

[2] In SFL theory, register and genre — in the sense of text type — are the same phenomenon viewed from different poles of the cline of instantiation.  Register is the view of functional varieties of language from the system pole; text type is the view from the instance pole.  Importantly, these are identified by the contextual features — field, tenor and mode — that they realise.  That is, situation type (context) and register/text type (language) are different levels of symbolic abstraction.

Thursday 21 April 2016

Misunderstanding Stratification And Context

Martin (1992: 498):
Register is defined as "the configuration of semantic resources that the member of a culture associates with a situation type.  It is the meaning potential that is accessible in a given social context" (Halliday 1978: 111). 
Defining register in these terms pushes considerations of context such as those addressed by Malinowski and Firth one level up, to what Halliday refers to as context of situation — presumably what is referred to as situation (non-linguistic phenomena) in Fig. 7.4.  Context of situation is then organised metafunctionally into field, tenor and mode as described above.

Blogger Comment:

This is a very strange misunderstanding indeed.  Halliday's definition of register — as the meaning that realises a situation type — has no bearing whatsoever on the organisation of strata.  Register and situation type are midway points on the cline of instantiation — for the strata of semantics and context, respectively.

Halliday's notion of context derives from Malinowski and Firth, and is structured semiotically in terms of field, tenor and mode, as Martin has already explained (p494) with a quote from Halliday (1978).

Wednesday 20 April 2016

Misrepresenting Halliday On Formal And Contextual Meaning

Martin (1992: 497-8):
Halliday (1961) similarly includes context as a level of language concerned with the relationship between form and extra-textual features of the situation.  This is glossed as semantics in Fig. 7.4 (from Halliday, McIntosh and Strevens 1964: 18), although it is important to keep in mind that Halliday (1961) recognises both formal and contextual meaning, with contextual meaning "an extension of the popular — an traditional linguistic — notion of meaning" (1961: 245).  As Halliday clarifies in Thibault (1987: 614), 
"the whole system is meaning creating.  Meaning is the product of the interrelations among the parts";
 it is a mistake in other words to make
"too close a tie-up between 'meaning' and the notion of a specifically 'semantic' level."

Blogger Comments:

[1] See the clarification in the preceding post.

[2] This is very misleading, especially when juxtaposed with this co-text. In this very early paper by Halliday, 'formal meaning' is the "information" of information theory, and 'contextual meaning' Halliday equates with 'semantics', as stated in Halliday et al. (1964). The following quote from Halliday (1961) clarifies the matter:
Language has formal meaning and contextual meaning. Formal meaning is the "information" of information theory, though (i) it can be stated without being quantified and was in fact formulated in linguistics independently of the development of information theory as a means of quantifying it, and (ii) formal meaning in lexis cannot be quantified until a method is found for measuring the information of non-finite ("open") sets. The formal meaning of an item is its operation in the network of formal relations. 
Contextual meaning, which is an extension of the popular — and traditional linguistic — notion of meaning, is quite distinct from formal meaning and has nothing whatsoever to do with 'information'. The contextual meaning of an item is its relation to extratextual features; but this is not a direct relation of the item as such, but of the item in its place in linguistic form: contextual meaning is therefore logically dependent on formal meaning.
[3] This continues the confusion between stratification and semogenesis. To be clear, this quote from Halliday refers to semogenesis — 'meaning creating' — not to stratification.

[4] The notion of meaning as a semantic level is meaning as a stratum in the stratification hierarchy, which is a means of parcelling out the complexity of the meaning making process into levels of symbolic abstraction.  Halliday (2008: 14) clarifies this narrower use of the word 'meaning':
Realisation is the relationship among strata… wordings realise patterns of meaning, which we refer to as the stratum of semantics. (Note that “meaning” is here being used in its narrower, more specific sense, to refer just to patterns in semantics.)

Monday 18 April 2016

Misrepresenting Firth On Context

Martin (1992: 497):
Systemic approaches to context derive from the work of Malinowski (1923, 1935) who argued that texts have to be understood in relation to their context of situation and context of culture.  Malinowski developed these ideas with respect to the problem of translating specific texts in particular contexts, leaving it to Firth (1950, 1957b, 1957c) to develop context more abstractly as a level of language.  For Firth, context was one of a number of levels of analysis (alongside grammar, morphology, lexis, phonology and phonetics) required for linguistics to make statements of meaning about text (Firth 1935/1957a: 33).

Blogger Comment:

This is misleading.  Firth did regard the text as part of his formulation of context, and he did regard context as integral to the study of language, but the following quote from Firth's A Synopsis Of Linguistic Theory, 1930-1955 makes clear that he distinguished language and context.  Firth (1962: 6):
Each function will be defined as the use of some language form or element in relation to some context.
This distinction is borne out by Firth's (op. cit.: 9) description of the constituents of the context of situation:
A. The participants: persons, personalities and relevant features of these.
(i) The verbal action of the participants.
(ii) The non-verbal action of the participants.
B. The relevant objects and non-verbal and non-personal events.
C. The effect of the verbal action.
Note that 'participants' refers to the interlocutors and any others relevant to the speech event, not to participants in processes in the clauses of the spoken texts that the interlocutors produce.

Sunday 17 April 2016

Inconsistent Claims About Discourse Semantics, Register, Genre And Ideology

Martin (1992: 496):
In this projection [Fig. 7.3 Language and its semiotic environment] metaredundancy (Lemke 1984) is reflected through the metaphor of concentric [sic] circles, with larger circles recontextualising smaller ones; the size of the circles also reflects the fact that the analysis tends to focus on larger units as one moves from phonology to ideology.  Thus the tendency in phonology to focus on syllables and phonemes, at the level of lexicogrammar to focus on the clause, at the level of discourse semantics to focus on the exchange or "paragraph", at the level of register to focus on a stage in a transaction, at the level of genre to focus on whole texts and at the level of ideology to focus on discourses manifested across a range of texts.  More in the spirit of Firth than Hjelmslev, this projection lends itself to a reading whereby meaning is constructed on all levels

Blogger Comments:

[1] In SFL, the tendency in phonology is to focus on the tone group, not on syllables and phonemes, and a tone group can be larger than a clause.

[2] To be clear, the lexicogrammar also affords the study of clause complex relations, and cohesive relations, some of which may obtain throughout an entire text.

[3] The units in Martin's model of discourse semantics (p325) are:
  • exchange and move (interpersonal),
  • participant (textual),
  • message (logical), and
  • message part (experiential)
The message is said (ibid.) to be realised in the lexicogrammar as 'Process (& transitivity rôles)', and the message part is said to be realised in the lexicogrammar as
  • Event (element at group rank),
  • Thing (element at group rank),
  • circumstance (element at clause rank),
  • Epithet (element at group rank) and
  • Manner adverb (element at group rank in Martin, but at clause rank in SFL theory).
That is to say, the focus in discourse semantics is on units that are, for the most part, smaller than the clause, not larger.  On the other hand, Martin's macro-Theme (introductory paragraph) and macro-New (text summary) are construed as functions at the level of text — which is said to be the focus at the level of genre, rather than the focus at the level of discourse semantics.

[4] Leaving aside the facts that, in SFL theory, register is language and context is not, there is the inconvenient truth that not all registers involve staged transactions.

[5] On Martin's own model, this confuses context (genre) with language (text).  In terms of SFL theory, it confuses text type (genre) with the meanings of a text (semantics).

[6] This confuses stratification with instantiation.  In SFL theory, the study of "discourses manifested across a range of texts" is the concern of register (text type, genre).

[7] This continues the confusion between semogenesis (making meaning) and stratification.  The whole point of stratification is to parcel out the complexity of meaning making into different levels of symbolic abstraction, thereby relating meaning to both lower levels of abstraction (wording and sounding) and to higher levels of abstraction (context).

Saturday 16 April 2016

Why The Argument For A Stratum Of Ideology Is Invalid

Martin (1992: 495-6):
Any configuration of this kind then needs to be qualified with respect to cultural diversity (cf. dialogism and heteroglossia in Bakhtin 1981).  Clearly, meaning potential is not evenly distributed across a culture (any more than material resources are).  Access to genre, register and language as semiotic resources is mediated through discourses of ethnicity, class, gender and generation, which discourses are in a continual process of negotiation with each other.  Not only is this process of negotiation manifest in all text, but it functions as well as the source of semogenesis, both contextual and linguistic.  It is for this reason that a fourth communicative plane, ideology, will be articulated here, with genre, and hence register and language as its expression form.

Blogger Comments:

[1] To be clear, genres/registers are types of language, not distinct semiotic modes.

[2] This is manifestly untrue.  The negotiation between "discourses of ethnicity, class, gender and generation" that "mediates access" to semiotic resources is a feature of only some registers of language.  Consider, for example, texts like the following:
There once was a young lady named Bright
Whose speed was much faster than light
She set out one day
In a relative way
And returned on the previous night.
[3] The claim here is that the negotiation between "discourses of ethnicity, class, gender and generation" that "mediates access" to semiotic resources is a source of semogenesis.  This misunderstands semogenesis.  In SFL theory, the logogenesis of all texts 'provides material for' ontogenesis, which 'provides material for' phylogenesis (Halliday & Matthiessen 1999: 18).

[4] The claim here is that, because the negotiation between "discourses of ethnicity, class, gender and generation" that "mediates access" to semiotic resources is a source of semogenesis, it motivates a more abstract level of context, above genre, termed ideology.  That is:
  • because logogenesis 'provides material for' ontogenesis (and ontogenesis for phylogenesis),
  • ideology can be modelled as a level of symbolic abstraction within context.
In logic, this is known as a non sequitur.  Semogenetic relations are distinct from — and so do not motivate — higher levels of symbolic abstraction.  This continues the confusion of semogenesis with stratification and the misconstrual of strata as modules.

In SFL theory, the ideologies of a culture are simply modelled within context, whereas the language that realises ideologies is, in the first instance, the concern of semantics.

Friday 15 April 2016

Theoretical Inconsistencies In Modelling Genre And Register As Context Strata

Martin (1992: 495):
The tension between these two perspectives will be resolved in this chapter by including in the interpretation of context two communication planes, genre (context of culture) and register (context of situation), with register functioning as the expression form of genre, at the same time as language functions as the expression form of register.  Register can then itself be organised with respect to field, tenor and mode, reflecting metafunctional diversity in its expression form, leaving genre to concentrate on the integration of meanings engendered by field, tenor and mode as systemically related social processes.

Blogger Comments:

[1] This misunderstands the theoretical concepts of 
  1. stratification,
  2. context, 
  3. genre and 
  4. register.
In SFL theory, strata are levels of symbolic abstraction, such that each lower stratum is a Token that realises an upper stratum, as Value.  Context is culture modelled as a semiotic system that is more symbolically abstract than language; language is a Token that realises context as Value.  Genre and register, on the other hand, are types of language; they are not more abstract than language, they do not realise language, they are language.  Because they are not more abstract than language, it is nonsensical to model them as strata above language.

[2] This misunderstands the theoretical concepts of
  1. stratification, and
  2. instantiation.
The relation between context of culture and context of situation is one of instantiation (ascription); a situation is an instance of a culture.  On the other hand, the relation between strata is realisation (identity); on this model, genre is realised by register.  Accordingly, it is theoretically inconsistent to equate the ascriptive relation between culture and situation as an identity relation between two strata, however conceived.

With regard to interpreting context of culture as genre and context of situation as register, note the inconsistencies with Martin's definitions of context of culture and context of situation here.

[3] This misunderstands the theoretical concepts of
  1. genre, and
  2. register.
In SFL theory, genre, in the sense of text type, is register viewed from the instance pole of the cline of instantiation; and contrariwise, register is genre (text type) viewed from the system pole of the cline of instantiation.  That is, genre and register are the same phenomenon viewed from different angles, instance and system.  Genre and register are not, therefore, different levels of symbolic abstraction, and, as such, it is theoretically inconsistent to model them as two strata related by realisation.

[4] This misunderstands the theoretical concepts of
  1. stratification, and
  2. register.
On the stratificational model, construing language as the 'expression form of register', is construing language as less symbolically abstract than register.  However, since register is language, register and language are of the same level of symbolic abstraction.  Register and language are not, therefore, different levels of symbolic abstraction, and, as such, it is theoretically inconsistent to model them as two strata related by realisation.

[5] In SFL theory, registers of language differ by the different contextual features — field, tenor and mode — that they realise.  Registers of language and contextual features are construed as different levels of symbolic abstraction.

[6] This confuses context with language.  In SFL theory, 'the integration of meanings engendered by field, tenor and mode' is the domain of the textual metafunction at the level of semantics.  So Martin's proposal is to model textual semantics as a higher stratum of context: genre.  It might be remembered that Martin models ideational semantics, activity sequences, as field, within his lower level of context: register.