Misrepresenting Martin (1992)

Martin (1992: 574):

The first point that needs to be made is that the interpretation of language and context here is indeed multi-structural and polysystemic.  System/structure theory has been re-involved in the description on a number of different levels — rank, stratum and plane — most of which involve metafunctional diversity and so can be analysed simultaneously as particle, wave and prosody; in addition, synoptic and dynamic perspectives on text as system and text as process have been introduced.

Blogger Comments:

[1] As previously explained, context is interpreted here as types of language, rather than as context, with the result that, being context, these types of language are not considered language.

[2] This is misleading, in that it overstates what has actually been done.  Almost none of the systems that Martin has provided specifies structural realisations.  This is partially disguised by the fact that some networks do include realisation statements; however, these merely provide textual instances of the feature.

• Of Martin's 49 discourse semantic system networks, across four metafunctions, only 4 specify structural realisations, and all are confined to the interpersonal metafunction.
• Of Martin's 11 register system networks, not one specifies any structural realisations; that is, no register structures are specified by register systems.
• Martin provides 0 genre systems — only taxonomies of types (factual and story genres); that is, no genre structures are specified by genre systems.

[3] This is misleading, in that it overstates what has actually been done.  Martin has not provided a rank scale for his planes of register and genre, and in the case of the stratum of discourse semantics, only one of the four metafunctional systems, the interpersonal, includes a rank scale: exchange and move.  The ranks discussed in the experiential dimension of discourse semantics — the clause and group — are ranks of a different stratum: lexicogrammar. 

[4] The level that does not involve metafunctional diversity is genre.  Metafunction is thus another dimension in which the model is inconsistent with the architecture of SFL theory, which follows from the misinterpretation of genre as context.

[5] To be clear, these are the favoured modes of structural realisation only, varying according to metafunction.  Significantly, these were introduced in the section on genre, the plane without metafunctional diversity. 

[6] This confuses text with language.  Text is only the instance pole of the cline of instantiation.  Language is the entire cline, from systemic potential to actual instance, with every point on the cline providing a different perspective.

'Text as system', therefore, is the instantial system; i.e. the system of an actual text, not the system of the language as a whole.  It is the instance viewed from the system pole.  Halliday & Matthiessen (1999: 384): 

If we look at logogenesis from the point of view of the system (rather than from the point of view of each instance), we can see that logogenesis builds up a version of the system that is particular to the text being generated: the speaker/writer uses this changing system as a resource in creating the text; and the listener/reader has to reconstruct something like that system in the process of interpreting the text — with the changing system as a resource for the process of interpretation. We call this an instantial system.

'Text as process', on the other hand, is the process of instantiation that occurs at the instance pole of the cline of instantiation during logogenesis.  As previously explained, Martin misconstrues 'process' as structure, the syntagmatic axis, "viewed dynamically".  That is, he confuses the instantiation of the system as instance with the axial realisation of the system as structure.

Why Martin Prefers His Own Model To Hasan's

Martin (1992: 572):

English Text's preferred position is to treat mode differences as simply difference in mode and to derive all text types from genre networks elaborated along the lines illustrated above.  These networks are in a sense systemic formulations of what Hasan's [sic] refers to as a culture's "array of existing conventions".  The model suggests however that these arrays are relevant for all genres, not just those constitutive in mode.  This avoids the problems inherent in Hasan's apparently materialist reading of context, which leads her to derive some texts from their context of situation and others with respect to their cultural heritage.  The ancillary/constitutive opposition is in any case a cline, which creates considerable uncertainty about how to model context for texts in 'middling' modes.

Blogger Comments:

[1] A preferred position is not a reasoned argument; it is merely a stance, an attitude, a pose.

[2] This is doubly misleading.  On the one hand, it falsely implies that Hasan does not "treat mode differences as simply difference in mode", and on the other hand, it falsely implies that Martin does.  As previously explained here, Martin allocates the mode system of medium to his register and the system of rhetorical mode to his genre.  In SFL theory, mode is a system of context, whereas register and genre are complementary perspectives on functional (diatypic) varieties of language.

[3] This is also doubly misleading.  

Firstly, the comparison here is with Hasan's model.  The discussion of Hasan's work focused on deriving text structures.  Here Martin offers his alternative "preferred position" but in doing so switches to deriving text types.  That is, it does not address the issue he raised about Hasan's model, and yet purports to be offering a better alternative.

Secondly, the reason Martin has switched from deriving text structures to deriving text types is to divert attention away from the fact that he has not devised any genre systems for deriving text structures — merely given excuses for not doing so; see here.  He has however provided small taxonomies of factual genres (Fig. 7.26) and story genres (Fig. 7.27), neither of which, of course, generates structures.

[4] This is a false claim.  Martin's taxonomy merely classifies types of story in terms of a few semantic features.  It provides no information about the conventions of text (semantic) structure associated with specific types.

[5] This is also doubly misleading.  On the one hand, it falsely implies that Hasan claimed these were relevant to constitutive mode, and on the other hand, it falsely implies that Hasan claimed these were relevant only to constitutive mode.  As Hasan (1984: 78) pointed out with regard to this "array of existing conventions":

But to say that the structure of a nursery tale is controlled by artistic conventions is to explain nothing …

[6] The claims here are that:

• Hasan's model of context is materialist;
• it is this materialist perspective that 'leads her to derive some texts from their context of situation and others with respect to their cultural heritage; and
• Martin's model avoids such problems inherent in Hasan's model.

All three claims are, of course, false.  This can be demonstrated as follows.

[a] Firstly, it is not true that Hasan's model of context is materialist.  Hasan uses Halliday's model of context, which conceives as context as the culture as a semiotic system.  In contrast, as previously demonstrated, Martin has used the term 'context' to refer to:

• material setting,
• co-text,
• register (diatypic variety of language viewed from the system pole),
• genre (diatypic variety of language viewed from the instance pole).

Materialism is clearly anathema to Martin, despite the fact that he doesn't understand what it is.  Here's one definition:

Materialism is a form of philosophical monism which holds that matter is the fundamental substance in nature, and that all phenomena, including mental phenomena and consciousness, are results of material interactions.

By way of contrast with Martin, Halliday and Matthiessen (1999: 609) clarify the philosophical underpinning of SFL theory as materialist in this sense:

… we are prepared to acknowledge a broadly materialist position …

[b] Secondly, it is untrue that a materialist perspective of context leads Hasan 'to derive some texts from their context of situation and others with respect to their cultural heritage'.  There are three untruths here, in addition to the false claim about Hasan's perspective being materialist:

• a materialist perspective — since this is Martin's misrepresentation — is thus not the reason for Hasan's two methods of analysis;
• Hasan does not "derive" texts, but text structures — Martin's omission of 'structure' misleadingly invites the interpretation as 'text type', on the basis of the preceding co-text; and
• Hasan does not "derive" text structures "with respect to their cultural heritage" — but from the semantic features of such texts, as explained in the previous post.

[c] Thirdly, it is untrue that Martin's model avoids the problems inherent in Hasan's model.  This also doubly misleading.  On the one hand, the "problems" in Hasan's model are not problems, but misrepresentations and misunderstandings on Martin's part, as demonstrated above and in the previous post.  On the other hand, Martin's model lacks what Hasan's model provides, since, unlike Hasan's, it does not account for text structures; it is merely a taxonomy of text types (genres).

[7] The claim here is that because ancillary vs constitutive opposition is a matter of degree — modelled as a cline — it "creates considerable uncertainty about how to model context for texts in 'middling' modes" when using Hasan's model.  Martin's confusions here are threefold:

• the mode system cline is a way of modelling context;
• Hasan is concerned with modelling the structural potential (semantics) of text types — i.e. language — not context;
• uncertainty does not arise, because both of Hasan's methods can be used for determining the structural potential of text types that realise intermediate values on the mode cline.

Consequently, on this fourth tendered piece of evidence, Martin's claim that his model of register and genre avoids additional complications in Halliday's (and Hasan's) model is the exact opposite of what is true.

∞

The number of untruths here, so efficiently packed into such a small passage of text, amply justifies the type of assessment made by Peter Medawar (1961) of another author:

its author can be excused of dishonesty only on the grounds that before deceiving others he has taken great pains to deceive himself.

Strategically Misrepresenting Hasan

Martin (1992: 572):

Mode also impinges on Hasan's model of context and text structure, since for her only texts where the role of language is ancillary and whose environment is pragmatic can be derived from contextual configurations (in her discussion she is opposing text types such as service encounters to the nursery tale; Hasan 1984: 76).  Since for constitutive modes, context cannot predict text structure, Hasan suggests that for these texts what matters most is "the array of existing conventions" (1984: 78).  Hasan's model then is one which derives text structures in two fundamentally different ways, depending on mode (see Harris 1987: 36-7 for a related critique).

Blogger Comments:

[1] The reason mode "impinges" on Hasan's model of context and text structure is because Hasan understands that

• text structure is semantics, 
• semantic structure realises semantic systems, and
• semantics realises context (field, tenor and mode).

[2] This seriously misrepresents Hasan (1984: 78), who only raised this in order to point out that, by itself, it explains nothing:

The single most salient fact that appears most relevant is the overall adherence to an array of existing conventions.  But to say that the structure of a nursery tale is controlled by artistic conventions is to explain nothing, unless alongside this assertion we can also provide a convincing account of how artistic conventions themselves originate and how any change is successfully introduced into a body of pre-existing conventions.

In contrast, Hasan goes on to derive the structural potential of this text type from the semantics of the texts themselves.

[3] This is falsely presented as if it is a defect in the Hasan's theory, rather than a distinction that is motivated by the data.  In the passage immediately preceding the quote above, Hasan (1984: 78) explains :

I would suggest that the nature of the factors which motivate the elements of structure in such genres is relatively opaque.  This is because the environments in which such texts are either created or received bears only a tangential relationship to their inner unity.  It follows then that the elements of the structure of the nursery tale can neither be seen as fully governed by the author-audience interaction

To contextualise this and the previous misrepresentation, Hasan's (1984) paper was not easily accessible at the time that Martin was writing, being only published in a Nottingham Linguistic Circular.  However, the paper was eventually published in the 1996 collection Ways Of Saying: Ways Of Meaning.

[4] No indication is given as to how the critique in the following obscure paper relates to Martin's "critique" of Hasan.

Harris, S. 1987. "Court Discussion as Genre: some problems and issues".  Department of English, University of Nottingham.  Occasional Papers in Systemic Linguistics 2. 35-74.

∞

This misrepresentation of Hasan's work is strategic because its function is to demonstrate that Martin's model is less complex, and thus preferable.  See the next post for an assessment of the truth of this proposition.

Misconstruing Language Sub-Potentials (Genres) As Context Potential (Culture)

Martin (1992: 560):

Linguists' concern with constituency at the level of genre has meant that questions of field, in Pike's sense of the term, have not been actively pursued.  Hasan's notion of generic structure potential does generalise across a range of text structures, determining their generic identity:

The property of structure is what allows us to distinguish between complete and incomplete texts on the one hand, and between different generic forms on the other.  With some oversimplification, the assumptions here can be stated as follows: associated with each genre of text — i.e. type of discourse — is a generalised structural formula, which permits an array of actual structures.  Each complete text must be a realisation of a structure from such an array.  The generic membership of the text is determined by reference to the structural formula to which the actual structure can be shown to belong.  (Hasan 1977: 229)

But Hasan has not attempted to develop these structure potentials in the direction of system/structure theory; and as noted above, the question of systemic relations among structure potentials does not really arise.

Blogger Comments:

[1] This invites the misinterpretation that such linguists also:

• misconstrue genre (language sub-potential) as a stratum of context (cultural potential), and
• misconstrue semantics (text structure) as genre (text type).

[2] To be clear, this is field, in Martin's misconstrual of Pike's sense of the term (as system rather than structure type).

[3] To be clear, as the quote makes plain, Hasan's generic structure potentials provide a generalised formula for a range of semantic structures of a given text type (genre).

[4] Hasan did not make systems of generic structural potentials because to do so would have been inconsistent with the rest of the architecture of SFL theory, since it would have confused the system pole of the cline of instantiation (semantic stratum potential) with the middle of the cline (genre/text type/register).

Misunderstanding Stratal Relations And Confusing Text Type (Genre) With System (Potential)

Martin (1992: 560):

Once posited as potentially divergent patterns of realisation, particulate, prosodic and periodic aspects of generic structure are not difficult to find.  More problematic is the question of how to model and integrate them once observed.  In particular this raises questions about the role played by genre in mapping field, tenor and mode choices onto each other in ways which capture teleological aspects of generic structure.  Current formulations of genre as system do not resolve these issues, but are nevertheless pursued below.

Blogger Comments:

[1] This is not in the least problematic.  The modes of realisation are the models, and the text, as semantic unit, provides the domain of application, just as the clause does in the lexicogrammar.

[2] This is a non-sequitur.  The modelling and integration of structure types is logically distinct from what follows; see [3].

[3] Translating into SFL theory, this becomes:

• the rôle played by text type
• in configuring contextual features
• in ways that capture purpose-directed aspects
• of the semantic structure of different text types.

In the first instance, this misunderstands the relation between contextual configurations and text type (genre).  Different text types realise different configurations of field, tenor and mode.

In the second instance, field, tenor and mode features do not "map onto each other"; selections from the three systems specify a given context (situation or situation type).

In the third instance, the notion of teleology is as redundant for text structure as it is for clause structure.  This confuses purpose with function.  In functional linguistics, structures have functions.

[4] The formulation of genre as system confuses subpotential (text type) with potential (system).  That is, it confuses a point of variation midway along the cline of instantiation with the system pole of that cline.  As such, such a formulation cannot resolve these non-issues — merely add to the confusion and theoretical inconsistency.

A Transparently False Claim

Martin (1992: 559):

Not in passing that wave patterns are very commonly reflected in the labelling given to beginning and end stages across genres, reflecting the peaks and troughs of prominence that open and close a text; a number of these are reviewed in Table 7.17.

Table 7.17 Culminative prominence and generic structure

	beginning (initial prominence)	end (final prominence)
Mitchell (1957)	Auctioneer’s Opening	Conclusion
Hasan (1977)	Greeting	Summary ^ Finis
Hasan (1985)	Sale Initiation	Purchase Closure
Hasan (1984)	Initiating Event	Finale
Ventola (1979)	Greeting	Leave-taking ^ Goodbye

Blogger Comment:

As the table demonstrates, this claim is wholly bogus.  Not one of the labels "reflects" the notion of a wave peak or trough.

Misrepresenting Halliday

Martin (1992: 559):

The final pattern of realisation to be considered is the culminative one, which in the grammar derives from the textual metafunction.  As developed in Chapter 6, Sections 3.2-3, culminative patterns structure discourse at several levels beyond the clause.  As Halliday puts this in Thibault (1987: 612):

Textual meanings typically give you the periodic movement which is so characteristic of discourse at all levels; everything from the smallest waves to the very large ones.  In other words, their is a hierarchy of periodicity, and that comes from the textual metafunction.

Blogger Comment:

This gives the very false impression that Martin's work is consistent with Halliday's theory, and invites the misinterpretation that Halliday might endorse both Martin's "beyond the clause" claim here and even his model of discourse semantics.

To be clear, Halliday does not equate discourse with semantics, and does not model it as a stratum of symbolic abstraction.  For example, Halliday & Matthiessen (1999: 512) define discourse as:

… the patterned forms of wording that constitute meaningful semiotic contexts.

Halliday (2008: 78) also clarifies this meaning of 'discourse' by relating it to 'text':

I do make a distinction between these two; but it is a difference in point of view, between different angles of vision on the phenomena, not in the phenomena themselves. So we can use either to define the other: “discourse” is text that is being viewed in its sociocultural context, while “text” is discourse that is being viewed as a process of language.

Halliday & Matthiessen (2014: 601) further clarify the semogenic dimension of 'discourse/text':

It is helpful to have a term for this general phenomenon – i.e. the creation of meaning in the course of the unfolding of text. We shall call it logogenesis, with ‘logos’ in its original sense of ‘discourse’ … 

Confusing Text Type With Text Structure

Martin (1992: 559):

Given the subversion of Labov and Waletzky's Complication/Resolution distinction by the non-discrete realisation of Evaluation, recognising simultaneous dimensions of experientially and interpersonally oriented structure at the level of genre (as with Halliday's 1985 analysis of the English clause) seems a more promising approach than the unidimensional modelling undertaken by Labov.

Blogger Comment:

This continues the confusion of text type (genre) with text structure (semantics).  This is not a promising approach, in terms of theoretical consistency.

Misconstruing Prosody

Martin (1992: 557):

Martin (1992a) treats one aspect of interpersonal meaning at the level of genre as a kind of generic prosody (with high, median and low values) mapped onto a particulate structure of the kind reviewed in 7.3.1.1 above …

Blogger Comments:

[1] Cf Martin (1992: 556):

The difficulty here lies in Labov's equation of interpersonal meaning (language) with Evaluation (genre) and his failure to distinguish clearly between particulate and prosodic perspectives on generic structure.

[2] Prosody and particulate are alternative modes of structural realisation.  A prosody is not particulate; a particulate realisation is not prosodic.  Anything "mapped onto" a particulate structure is, by definition, not prosodic.

Distinguishing Interpersonal Meaning From Evaluation

Martin (1992: 556):

Rather problematically for this definition of Resolution [Labov and Waletzky (1967: 39)], the realisation of Evaluation turns out to be potentially non-discrete.  There may be more than one Evaluation in a narrative … .  The non-discrete realisation of Evaluation makes locating the major focus of the Evaluation difficult, which in turn calls into question the use of a localised Evaluation to define the Resolution.  The difficulty here lies in Labov's equation of interpersonal meaning (language) with Evaluation (genre) and his failure to distinguish clearly between particulate and prosodic perspectives on generic structure.

Blogger Comments:

[1] This is a contradiction in terms.  Martin claims Evaluation is potentially non-discrete — i.e. not separate — but equates this with there being more than one.  By definition, for there to be more than one, each must be a discrete realisation.

[2] The claim here is that the connected realisation of Evaluation entails that there is more than one of them, and that this makes it difficult to choose which of the "connected realisation" can be used to define the Resolution stage of a narrative.  See [1].

[3] The "difficulty" here derives:

• firstly, from Martin's misunderstanding of the term 'non-discrete',
• secondly, from his opposition of interpersonal meaning and Evaluation, and
• thirdly, and more seriously, from his claim that types of language (genres) and the linguistic evaluations made in such text types are not language.

[4] Perhaps this failure can be forgiven, on the grounds that Labov was not a student of Halliday and worked in a discrete linguistic tradition.

Prioritising Structure Over System

Martin (1992: 553):

The advantage of the dynamic perspective is that choices can be conditioned by the point reached in a text's development.  Keeping in mind Firth's comment that "The moment a conversation is started, whatever is said is a determining condition for what, in any reasonable expectation, may follow" (1935/1957: 31-2), this is an important perspective to keep in mind.

Blogger Comment:

Importantly, but apparently unknown to Martin, this reflects one crucial difference between Firthian Linguistics and the (Neo-Firthian) Systemic Functional Linguistics of Firth's student, Halliday.  For Firth, consistent with his quote above, entry conditions to systems of choice are specified syntagmatically, whereas for Halliday, entry conditions to systems of choice are specified paradigmatically.  Systemic Functional Linguistics, as the name implies, prioritises paradigmatic system over syntagmatic structure.

Misrepresenting Hasan On Generic Structure Potential

Martin (1992: 551-3):

In the same paper, Martin (1985) cautions against another bias inherited by genre theory from grammarians — namely that towards a synoptic as opposed to dynamic form of representation and interpretation.  The basic problem here is that in order to attain generality, grammarians have tended in various ways to abstract away from linear progression (to divorce in other words functional order from sequence).  Structure potentials such as those used by Hasan, and to an even greater extent the system/structure realisation cycle suggested by Martin, separate order from sequence in such a way that choices leading to a particular text structure bear no direct temporal relation to the unfolding of that text in real time.  To the extent that aspects of the unfolding of text are conditioned by linear progression, this perspective breaks down.

Blogger Comments:

[1] In Systemic Functional Grammatics, sequence during instantiation is specified by realisation statements of features, such as declarative Subject ^ Finite.

[2] This misunderstands the relation between order and sequence.  Sequence is one type of order; that is, sequence is a hyponym of order.  To speak of separating a hyponym from its superordinate category is a nonsense — cf separating lorikeet from bird.

[3] This is manifestly untrue with respect to Hasan's generic structure structure potentials.  Hasan (1985/9: 61) specifies the order of obligatory elements in the sense of 'sequence':

So the obligatory elements of text 4.1 are SR, SC, S, P and PC in that order.  This can be displayed as SR ^ SC ^ S ^ P ^ PC, with the sign ^ showing the order of elements.

[4] This continues the earlier confusion of syntagmatic structure with logogenesis (the unfolding of text at the instance pole of the cline of instantiation).

Misinterpreting Pike

Martin (1992: 548-9):

Halliday (1978, 1979) associates different types of structural realisation with different metafunctions. … In addition Halliday points out the correlation between these different metafunctional perspectives on realisation and Pike's work on particle, wave and field. … Halliday correlates constituency realisation with the particulate perspective, culminative realisation with wave theory and prosodic realisation with field.  The last of these correlations seems to be stretching a point since for Pike field theory is not so much concerned with non-discrete realisations as with "intersecting properties of experience" that "cluster into bundles of simultaneous characteristics which together make up the patterns" of experience (Pike 1982: 13); Pike sums up his perspectives when introducing field as follows:

We have discussed the way in which elements can be viewed by themselves as particles, or as waves smearing into some kind of continuum whose prominent parts make up nuclei.  Now we turn to sets of relationships which occur when units are linked to one another by their presence in some larger system.  A total set of relationships and of units in these relationships we call a field.  (1982: 30)

The component of systemic theory which would appear to correspond most closely to this articulation of field is system, not prosody.  For this reason the discussion will be pursued here at the level of genre under the headings of particle, prosody and wave.

Blogger Comments:

[1] Significantly, Martin's model of discourse structure is inconsistent with the SFL association of metafunction and structure type.  On the one hand:

• the structure of the logical system, conjunction, is modelled as messages (particles) related by types of expansion, and
• the structure of the experiential system, ideation, is modelled as lexical strings of message parts (particles),

but on the other:

• the structure of the interpersonal system, negotiation, is modelled as constituency (particles) — exchange and move — rather than as prosody, and
• the structure of the textual system, identification, is modelled as reference chains of participants (particles), rather than as wave.

[2] While it is true that Pike's notion of field does not correlate with the notion of prosody, it is not true that it correlates with the notion of system in SFL theory.  Martin has simply misinterpreted Pike's notion of 'system' as Halliday's notion of 'system'.  Pike is concerned with complementary perspectives on a sequence of items; that is, in SFL terms: syntagmatic, not paradigmatic relations.  Pike (1993):

The observer can choose to look at a sequence of items as if they were separate, in order to talk about phonemes, words, etc., but he also needs to be able to talk about them in terms of fusion, and he doesn't understand them ultimately unless he sees them in relation to a patterned system. Hence particle, wave, and field.

[3] Halliday (1977, 1978: 188) uses the term 'prosodic' rather than 'field' in relating types of grammatical structure to metafunctions.  Martin (1992) is merely following Halliday while implying otherwise.

The consideration of structure types with regard to genre suggests that text type is about to be confused with semantic structure.

[4] Having introduced Halliday's association of structure types with metafunctions, Martin will now consider the structure types in relation to genre, a stratal system that he does not differentiate in terms of metafunction. 

Confusing First And Second Orders Of Experience

Martin (1992: 546-7):

In this connection it is useful to compare three of Barthes's examples of sequences with three of Hasan's examples of text structure (only her obligatory elements will be considered here):

sequences (Barthes 1966/1977):

‘telephone call’:

telephone ringing ^ picking up the receiver ^ speaking ^ putting down the receiver (1966/1977: 101)

“having a drink”:

order a drink ^ obtain it ^ drink it ^ pay for it (1966/1977: 101)

“offering a cigarette”:

offering ^ accepting ^ smoking ^ lighting (1966/1977: 102)

text structures (Hasan 1977, 1984, 1985/9):

“medical appointment making”:

Identification ^ Application ^ Offer ^ Confirmation (1977: 233)

“service encounter”:

Sale request ^ Sale compliance ^ Sale ^ Purchase ^ Purchase closure (1985/1989: 60)

“nursery tale”:

Initiating event ^ Sequent event ^ Final event (1984: 80)

At a glance it might appear that Barthes is analysing action whereas Hasan is analysing text (cf. Barthes's telephone call and Hasan's appointment making).  But the opposition is by no means as simple as this.  Most of Barthes's nuclei involve interlocutors speaking, and those which do not are easy to render linguistically in narration (as they have been in Barthes's own account).  So a simple opposition between verbal and non-verbal action will not do.

Blogger Comments:

[1] Both Barthes and Hasan are analysing texts, instances of language, at the level of semantics.  The difference between them, in terms of SFL theory (Halliday & Matthiessen 2014), is that Barthes approach is more "bottom-up", staying close to the sequences of figures that are realised grammatically as clauses, whereas Hasan's approach is more "top-down", viewing from context.

[2] The interlocutors in Barthes' data are characters in narrative texts.  That is, they are verbal projections of the writer.  The confusion here is in terms of orders of experience: first-order (writers, speakers, readers, listeners etc.) vs second-order (characters in stories etc.).

Misconstruing Activity Sequence (Semantics) As Field And Schematic Structure (Semantics) As Genre

Martin (1992: 546):

Hasan's association of obligatory elements of text structure with field makes it critical to consider the relationship between activity sequence (field) and schematic structure (genre) here.

Blogger Comment:

[1] For Hasan (1985/9: 62), the obligatory elements of text structure are the elements that define the genre (text type):

So, by implication, the obligatory elements define the genre to which a text belongs;

[2] This is a non-sequitur; see [1] and [3].

[3] One problem here is that activity sequence is not field and schematic structure is not genre.

An activity sequence is a semantic construal of experience, whether a mental construal of visual experience or a verbal construal in a spoken or written text.  Field, on the other hand, is 'what is going on' and the subject matter when interlocutors are speaking to each other.  The confusion is along the stratal dimension: semantics (activity sequence) vs context (field).

Schematic structure is the structure of a text as a semantic unit.  Genre, on the other hand, is text type — that is, register viewed from the instance pole of the cline of instantiation.  The confusion is along the dimension of instantiation: potential (semantic) vs subpotential (genre).

In short, in purporting to consider the relationship between field and genre, Martin is actually considering the relation between two ways of construing semantic units.  This will be borne out in the following blog posts.

This, like so many other confusions identified here, demonstrates the serious pitfalls of misunderstanding 'all strata make meaning' (semogenesis) as all strata are levels of linguistic meaning.

Misidentifying Metafunctions

Martin (1992: 508):

Mode refers to the rôle language is playing in realising social action.  Within register, it is the projection of textual meaning, and so is realised primarily through the textual metafunction in language.  Mode thus puts major systems such as TONALITY and TONICITY in phonology, and THEME and INFORMATION (clause), DEIXIS (nominal group), TENSE (verbal group) and SUBSTITUTION and ELLIPSIS (clause and group) in the grammar at risk, and because of their textual orientation impacts on all systems at the level of discourse semantics (NEGOTIATION, IDENTIFICATION, CONJUNCTION and IDEATION).

Blogger Comments:

[1] To be clear, Martin's 'register' is a misconstrual of culture (context) as a functional variety of language (register).  In SFL theory, mode is the "projection" of the textual metafunctional principle onto context, in modelling culture as a semiotic system, not the projection of textual meaning.

[2] Information is not a system of the clause.  It is a system of the information unit (Halliday & Matthiessen 2014: 115).

[3] Deixis is not a textual system of the nominal group.  It is an interpersonal system (Halliday & Matthiessen 2014: 387).  Deixis functions textually in the cohesive system of reference.

[4] Tense is not a textual system of the verbal group.  Tense is realised by the logical structure of the verbal group (Halliday & Matthiessen 2014: 398).

[5] Strictly speaking, substitution-&-ellipsis is not a system of the clause and group.  Rank systems are structural.  As a system of cohesion, it is non-structural.  The clause and the group are domains in which it operates.

The other cohesive systems — reference, conjunction and lexical cohesion — are also relevant here, but omitted because Martin has misconstrued them as discourse semantic systems of various metafunctions: identification (textual), conjunction (logical) and ideation (experiential), respectively.

[6] This exemplifies Martin's confusion between text as a semantic unit, and the textual metafunction as a dimension of semiosis.  See [5] above.

Confusing Context (And Semantics) With Text Type

Martin (1992: 505):

Where the models do differ substantively is with respect to Martin's suggestion that there is a network of relationships underlying register which relates text types to each other in ways they cannot be inter-related considered from the perspective of any one register variable.

Blogger Comment:

This misunderstands the SFL model with which Hasan's model is consistent, and with which Martin's model is not consistent.

In SFL theory, registers are text types (genres) viewed from the system pole of the cline of instantiation.  Text types realise situation types, whose metafunctional dimensions are field tenor and mode.  Text types differ in terms of the contextual features (field tenor and mode) they realise.

In Martin's model, a system of mode is relocated from context (misconstrued as register) to a new higher stratum (misconstrued as genre), with this higher system said to "underlie" the lower stratum.

Moreover, schematic structure — text structure — is located on the higher stratum of "context" (genre), instead of on the stratum of semantics.

Misunderstanding Stratification And Realisation

Martin (1992: 505):

The common ground between the two models lies in the correlation proposed between schematic structure and field, mode and tenor options; for both Martin and Hasan staging redounds with social context.  Keeping in mind that realisation is not theoretically directional in systemic models, there is nothing substantive in the fact that whereas for Hasan, choices in field, mode and tenor are realised by schematic structure, for Martin schematic structure is realised through these same components of register.

Blogger Comment:

This seriously misunderstands stratification and realisation.  Strata are relative levels of symbolic abstraction, with higher levels realised by lower levels.  Contra Martin, higher levels do not realise lower levels; e.g. lexicogrammar does not realise phonology.

In Hasan's model, context is realised by semantics —configurations of field, tenor and mode features are realised in text structure.

In Martin's model, semantics is realised by context — i.e. schematic (text) structure is realised by field, tenor and mode — with the complications that semantics is misconstrued as genre, and context is misconstrued as register.

It is manifestly untrue that this represents "common ground between the two models".