Monday, 21 August 2017

Pageviews by Countries

Graph of most popular countries among blog viewers
United States
United Kingdom
South Korea

Wednesday, 5 April 2017

David Rose On "Lexical Meanings And Lexical Relations In Grammatical Analysis"

… I am also intrigued by the roles of interpreting lexical items in the grammatical analysis. For example, the first clause is interpreted as receptive circumstantial identifying, subclassified as Cause:purpose, from synonymy of ‘purpose’ with the lexical item ‘aim’ and the embedded clause ‘to (do)’ 
The primary clauses are attributive, in which each 'experimental method’ is classified as 'an isothermal process’ or 'an isochoric process’. That seems to be the function of the non-specific Deictic “an”, to indicate a class of items.
What’s particularly intriguing to me is the extent of assumptions of lexical meanings and lexical relations in grammatical analysis. This is most apparent in Halliday’s discussion of relational processes in IFG, which I found baffling until I started recognising the lexical assumptions implicit in his categories.  Has anyone else experienced this?

Blogger Comments:

[1] These false claims proceed from a misunderstanding of the theoretical dimension of delicacy (but see also point [2] below).  The subtypes of process are located intermediate between the most general systems and the most delicate systems that specify individual lexical items.  Halliday & Matthiessen (1999: 198-9):
… we can differentiate both processes and participants into finer and finer subcategories, until we reach a degree of differentiation that is associated with the choice of words (lexical items). Note that it is not (usually) the lexical items themselves that figure as terms of the systems in the network. Rather, the systems are systems of features, and the lexical items come in as the synthetic realisation of particular feature combinations. 

[2] This misunderstands the reasoning behind the grammatical analysis of the clause (shown here).  The clause construes a relation of identity between an embedded clause complex of cause: purpose (to investigate the properties of air…) and a nominal group of cause: purpose (the aim of this experiment); see Halliday & Matthiessen (2014: 500ff) on 'nouns of expansion'.

[3] The two 'primary' clauses are embedded and identifying, not attributive (as detailed here).  Each identifies (encodes) one experimental method.  The nominal groups with the non-specific Deictic an function as Identifier Token, not Attribute.

Wednesday, 22 March 2017

Jim Martin Inventing A New Tense

Here's one though that Halliday's [tense] system doesn't cope with:
if I had have done it

Blogger Comment:

Ignoring the if makes it clear that this is simply an instance of someone misinterpreting
I'd have done it
I had have done it 
instead of
I would have done it
The tense system provides the means of interpreting such instances — this one would be: "past-in-present-in-past" — and thus, the means of understanding why they are anomalous, and thus, why they are unlikely to scale the cline of instantiation and find a place in the general system of potential.

Googling 'had have' will demonstrate the infrequency of such instances.

For the discourse semantics of verbal groups (Martin 1992), see here, here and here.

Friday, 24 February 2017

John Bateman On The Interpersonal Metafunction In Mathematics

John Bateman wrote to sysfling on 24 February 2017 at 10:10 in reply to the query 'To what degree is TENOR important in mathematical discourse?':
one of the well known kinds of proofs for more advanced mathematics is 'proof by intimidation': sounds pretty interpersonal to me! :-) 
And as soon as you move down to educational contexts, you'll want to be hitting your tenor variables right. Since all metafunctions in less-grammaticised semiotic modes are in any case discourse interpretations, I'm sure you'll be able to find something... 
I'd be with Yaegan though in doubting that the technical resources employed in mathematics have any inherent interpersonal organisationalthough certain variations,
'x(1-y)' compared to 'x times (1 - y)'
may move in that direction. But is that what you meant?

Blogger Comments:

[1] The use of the word 'down' here — from genre to tenor — indicates that Bateman is using Martin's (1992) model* in which genre and register are misconstrued as context strata instead of functional varieties of language (a point on the cline of instantiation).  For some of the many theoretical misunderstandings on which Martin's model is based, see the arguments here.

[2] The use of the word 'discourse' here — instead of 'semantics' or 'meaning' — indicates that Bateman is using Martin's (1992) model* of discourse semantics.  For some of the many theoretical misunderstandings on which Martin's model is based, see the arguments here.

[3] Every mathematical equation is a proposition realised by a declarative clause, and structured as Subject^Finite/Predicator^Complement, whether realised in the graphology peculiar to the field of mathematics or in the unspecialised graphology of an individual language.  A sample structural analysis can be viewed here.

[4] The variation here is textual (mode), not interpersonal (tenor).

* Postscript: A critical examination of Bateman's review of Martin (1992) will be the subject of a new blog: Thoughts That Didn't Occur.  At first glance, it appears that Bateman has failed to notice any of the 2000+ theoretical inconsistencies in Martin's work (identified here).

Friday, 17 February 2017

David Rose On Genres As Features In Systems

David Rose wrote to sys-func and sysfling on 11 February 2017 at 07:42:
Can I reverse Andrés’ proposition, so that 'Bakhtin’s concepts of Dialogism, Polyphony and Heteroglossia’ are explained through SFL? To do so requires the hard yards of describing genre and register theoretically as semiotic systems. Then we can insist that genres are not ‘conventions’ but features in systems, with distinct structural realisations, but plenty of latitude for interstratal realisation in field, tenor and mode systems.
Perhaps one reason for missing register is the paucity of our work on these systems, compared with genre and language systems (and other modalities). Could this be a useful place to look for the next generation of SFL scholars?

Blogger Comments:

[1] For glosses of Bakhtin's notions of dialogism, voice (polyphony = many voices) and heteroglossia, see here.

[2] As the terms 'genre' and 'register' suggest, these are types of semiotic systems — prototypically: types of linguistic systems.  That is why they are theorised in SFL as a point of variation on the cline of instantiation of language, rather than as strata of context more abstract than language.

[3] insist:  'demand something forcefully, not accepting refusal'.

[4] The opposition 'features vs conventions' is a false dichotomy.  In SFL theory, genres (text types) vary by the frequencies of feature selection — in systems at the level of semantics and lexicogrammar.  It is the feature frequencies that distinguish one text type from another that corresponds to any notion of a text type as a 'convention'.

The "system" of genre that Rose is promoting is merely a taxonomy of genres, with a structural realisation associated with each genre.  The system does not generate a genre — in the way that a clause system generates a clause — and the associated text structures are misconstrued as context instead of semantics.

[5] The absurd claim here is that a stage of a genre, such as a narrative, is realised, say, in the tenor relation between the teller of the story, a parent, and its audience, a child.

[6] Let's hope the next generation of SFL scholars are able to understand the theoretical architecture, or at the very least, are able to tell the difference between context and language as text type.

Thursday, 16 February 2017

David Rose On Genre And Register Not Being Language

David Rose wrote to sys-func and sysfling on 11 February 2017 at 07:42:
I’d like to assume Estela’s *position* is the stratified context she mentions, of genre configuring register realised as language (and other modalities). This stratification is missing from much work on genre/language relations (within and without SFL), which has the effect of putting all the explanatory load on genre and language. One result is the ongoing battles over variations in genre that are really variations in field, tenor and mode. Another is the assumption that finding similar language patterns associated with different genres means that genres are somehow ‘mixed’.

Blogger Comments:

[1] The "stratified context" model (Martin 1992), in which genre and register are misconstrued as context instead of language variants, derives from misunderstandings of both stratification and instantiation, as demonstrated in great detail here and here.

[2] This misrepresents Martin's model of stratified context.  Genre doesn't "configure" register; on Martin's model, genre is realised by register.  Martin posits genre and register as different levels of symbolic abstraction.  In SFL theory, genre (text type) and register are the same phenomenon — functional varieties of language — viewed from opposite poles of the cline of instantiation.

[3] In SFL theory, register isn't "realised" as language; it is the culture-as-semiotic-system that is realised as language.  Registers are (diatypic varieties of) language — different language variants that realise different context variants (situation types).

[4] In SFL theory, the relation between genre and language is modelled in terms of the cline of instantiation.  At the system pole of the cline is language as potential, at the instance pole is language as text, and in the middle of the cline is language as genre.

[5] In SFL theory, "variations in genre" is linguistic variation according to text type, whereas "variations in field, tenor and mode" is contextual variation.  In SFL theory, different genres (text types) realise different contextual configurations (Hasan) of field, tenor and mode features.  In Martin's model, it is the opposite: different genres are realised by different field, tenor and mode features, with these features being misconstrued as register.  In Martin's model, text type (genre) is realised by register, and neither are language.

[6] This is a non-sequitur.  This assumption is not a result of ignoring Martin's confused model of stratified context.   For the notion of 'mixed genres', see the previous post.

Wednesday, 15 February 2017

John Bateman On 'Genre Mixing'

John Bateman wrote to sys-func and sysfling on 10 February 2017 at 23:57: 
So, yes, one can well get a variety of 'mixes' when we cross strata, and perhaps the more abstract, the more this is possible (semiotically because the body or weight of distinctions instantiated at the less abstract strata begins to be sufficiently substantial as to be able to 'stand against' other collections of instantiations)Isn't this just heteroglossia? Doesn't sound too contentious.
As a text unfolds, there is then sufficient material to realise syndromes of choices that appear to realise several 'genres' either simultaneously or within locally distinct portions of the text as a whole. Taking the weather-climate metaphor: some texts may have patchy weather, perhaps even microclimates, generically. This shouldn't lead to theoretical difficulties.

Blogger Comments:

[1] This misunderstands stratification.  There is no "crossing of strata" during logogenesis.  In SFL, strata are different levels of symbolic abstraction, not interacting modules.  Since "going up the strata" cannot occur, it cannot increase the likelihood of a text being attributed to more than one text type ("genre mixing").

[2] To be clear, on the SFL model, the way that genres differ is in terms of the frequencies of feature selections (if viewed from the instance pole of the cline of instantiation, as text types), or in terms of the probabilities of feature selections (if viewed from the system pole, as registers).

[3] In SFL, the criterial stratum for genre classification is semantics. This is because genre/text type/register is a semantic concept.  Halliday in Halliday & Hasan (1985: 38-9):
A register is a semantic concept.  It can be defined as a configuration of meanings that are typically associated with a particular situational configuration of field, mode, and tenor.  But since it is a configuration of meanings, a register must also, of course, include the expressions, the lexico-grammatical and phonological features, that typically accompany or REALISE these meanings.
[4] This is not heteroglossia.  Heteroglossia refers to the same wording realising different meaning in different contexts.  Bakhtin (1981: 428):
The base condition governing the operation of meaning in any utterance. It is that which insures the primacy of context over text. At any given time, in any given place, there will be a set of conditions — social, historical, meteorological, physiological — that will insure that a word uttered in that place and at that time will have a meaning different than it would have under any other conditions; all utterances are heteroglot in that they are functions of a matrix of forces practically impossible to recoup, and therefore impossible to resolve.
[5] This misconstrues instantiation as stratification, following Martin (1992).  In SFL, the relation between text and genre is instantiation, not realisation.  Some of the theoretical problems with Martin's construal of genre as a stratum of context are detailed here.  It can be briefly noted, for instance, that if genre is modelled as a stratum of context, then a text is not an instance of a genre, since text is an instance of language, not context, and genre is modelled as context, not language.

[6] This misunderstands the weather-climate metaphor that is used to explain the cline of instantiation.  On the basis of that metaphor, the meteorological counterpart of a text that "mixes genres" is weather that "mixes weather types".  That is to say, the meteorological counterpart of a text that fits the statistical profile of more than one genre is weather that fits the statistical profile of more than one weather type.

Tuesday, 14 February 2017

David Rose Confused By Mixed Genres

So where does our model afford ‘mixing’ 
Between features in systems? That would produce a new valeur, so a different feature

Between features and structures in axis? Again, a different (mixed?) structure realises a different feature
Between systems at different strata? OK, here we have lots of variation (grammatical metaphor is one example among many)
Here’s an advantage of modelling context as a stratified semiotic 
Genre is a semiotic system, whose features are realised axially by various structural configurations (e.g. staging)
Change the structural configuration sufficiently and we have a different valeur in the system, i.e. a different genre
But one generic structure can be realised interstratally by many variations in register

Again, field, tenor and mode are semiotic systems, with similar sets of constraints and potential for variation

And again their structures can be realised interstratally in variable ways, including modalities other than language

Blogger Comments:

[1] The notion of 'mixing' here refers to instances that 'mix genres'.  On the SFL model, this means texts that can be ascribed to more than one text type.  As explained in the previous post, text types vary by the frequency of feature selections, and so text types are more like fuzzy sets than discrete categories, with texts showing degrees of membership to text types.

[2] The fundamental disadvantages of modelling context as stratified genre and register are:
  • genre and register are varieties of language, not systems of context;
  • context is the culture as semiotic, not varieties of language;
  • genre and register are not different levels of symbolic abstraction (strata), but different angles on a point of variation on the cline of instantiation;
  • strata are systems (semantics, lexicogrammar, phonology), not varieties;
  • a text is an instance of language, not an instance of context — whether stratified or not.  
For more detailed argumentation, see here, here, or here.

[3] The genre "systems" in Martin (1992) and Martin & Rose (2008) are not systems of feature options, but genre taxonomies of types, and do not include realisation statements.  From the perspective of SFL theory, the stages of genre are semantic structures varying according to text type; see Hasan (1985: 64-9) on 'generic structure potentials'.

[4] This confuses different options in the systems of field, tenor and mode with different variants — different configurations of field, tenor and mode — and misrepresents these contextual systems as diatypic varieties of language.  See Hasan (1985: 55ff) on contextual configurations.

[5] No systems of field, tenor or mode provide realisation statements specifying syntagmatic structures.  The realisation of such structures, in terms of stratification, would be semantic structures.

Monday, 13 February 2017

John Bateman Re-Inventing Genre

If it is not recognisable as an instance of one genre, then it must be another genre.
I disagree with this position.

indeed; this would be genre à la old-style literary criticism (probably being ungenerous even to them). The story can be 'made to be true' trivially by defining genre in such a way that it is a unique label that can be attributed to any text, but that is an unnecessary modelling move that weakens any account and has a dubious relationship to properties of the relevant semiotic system(s).
Since genre is a discourse phenomenon (if it's anything), attributions of genres are attributions of hypotheses concerning relevant conventions of interpretation.  The lexicogrammar and semantics are well capable of carrying patterns attributable to multiple sources of conventional interpretation. Any instance of language use can draw on and signal relevance of a variety of genres simultaneously or across logogenesis. Models are required that do not promote 'mixing' as something special, but as the norm.  Otherwise genre change over time becomes more difficult to account for than necessary, rather than falling out of the model as a prediction. Surprised that one should consider anything else these days...

Blogger Comments:

[1] It is Bateman's de novo theorising on genre that is unnecessary.  In SFL theory, a genre is Hasan's (1985) term for a text type — that is, a genre is a register viewed from the instance pole of the cline of instantiation.  Every text, by definition, is a type of text, even if it is the only token of its type.

[2] On the SFL model, where genre variation is a point on the cline of instantiation, rather than misconstrued as a stratum of context, genres vary by the frequencies of feature selection at the stratum of semantics, together with those of the strata below that realise semantic selections.

If genre variation is viewed from the system pole of the cline, then it appears as register variation. Registers vary by the probabilities of feature selection at the stratum of semantics, together with those of the strata below that realise them.  This means that registers shade into one another, since they may share some probabilities, and it means that a given text (instance) can be attributed to different registers — and so: different genres — depending on which features are focused on as criterial.  Halliday & Matthiessen (2014: 73):
In general, like any other language, English needs to be interpreted and described as an assemblage of varieties – varieties that are differentiated along different dimensions, with fuzzy boundaries.
It is the probabilistic interpretation of variation that provides a window on the evolution of language and its subpotentials, as Halliday & Matthiessen (2014: 73-4) point out, taking the perspective of lexicogrammar:
the system of lexicogrammar is probabilistic in nature, and probabilities vary across varieties of English – dialectal, codal and registerial varieties. If we include probabilistic information in the description of the lexicogrammar, we also pave the way for interpreting the system as one that is always in the process of becoming, not one that is in a frozen state of being: the evolution of language involves gradual changes in probabilities, over long periods of time but also over much shorter periods.

Sunday, 12 February 2017

John Bateman Denying The Existence Of Text

well, *actually* there is no such thing as text. There's just variations of patterns of pressure gradients in the air and contrasts in brightness in the visual field....

Blogger Comments:

Bateman claims that variations of patterns of pressure gradients in the air and contrasts in brightness in the visual field exist, but that text does not.

no such thing as text

Process: existential
comment Adjunct: factual

variations of patterns of pressure gradients in the air and contrasts in brightness in the visual field

Process: existential

mood Adjunct: counterexpectancy: limiting

On the one hand, the claim here is that only phenomena in the material domain of experience can be ascribed to the set of existents — a view that owes something to Galileo's distinction between primary and secondary qualities, refined further by Descartes' distinction between res extensa and res cogitans.  On the other hand, it is a reductionist interpretation of that view, since it reduces the existence of such material phenomena to (virtually) their lowest level of organisation, ignoring all higher levels.

Galileo's distinction assumes that meaning is transcendent — that it is not confined to semiotic systems.  In contrast, SFL theory models meaning as immanent — confined to the domain of semiotic systems. Variations of patterns of pressure gradients in the air and contrasts in brightness in the visual field, just as much as texts, are construals of experience as meaning.

Saturday, 11 February 2017

David Rose Denying The Existence Of Linguistic Theory

*Actually* there’s no such thing as genres, or registers, or grammar. There's just text. What we like to call genres, or cultural practices, or semiotic acts, or grammatical features… are just recurring instances of similar patterns in ongoing text, that we recognise as types and might give names to. 
Features at any stratum are nothing but records of past recurrences (usuality) that may predict future recurrences (probability) 
One social application of linguistics is to (try to) control future recurrences (obligation) 
How many recurrences does it take before we call it a feature?

Blogger Comments:

[1] Rose claims that text exists but that genres, registers and grammar* do not.

no such thing as genres, or registers, or grammar

Process: existential
comment Adjunct: factual


Process: existential

mood Adjunct: counterexpectancy: limiting

The assumption here is that only instances of perceptual phenomena can be ascribed to the set of existents.  Halliday (2008: 13):
Observed from close up, language appears in the guise of text, instances of spoken or written discourse that can be perceived by the senses — that can be heard or seen. Observed from a distance, language appears as a potential, an open ended network of possibilities with certain statistical properties and having certain kinds of interrelationship with its eco-social environment.
The validity of the proposition can be assessed by considering a couple of its implications:
  • Language–as–instance (text) exists, but language–as–potential does not.  (Cf weather exists, but climate does not.)  That is, only the instance pole of the cline of instantiation "exists"; a speaker's potential to instantiate texts in logogenesis does not "exist"; the potential that is established in individuals in ontogenesis does not "exist"; the potential that evolves in the species does not "exist".
  • Texts about genres, registers and grammar exist, but their subject matter does not.  A text such as Genre Relations is concerned with things that don't exist.

[2] This is the view of registers/genres — as a midpoint on the cline of instantiation — from the instance pole only; see Halliday (2008: 81-2) here.  Rose, however, follows Martin in misconstruing genre and register as systems of context.

[3] Feature frequencies in instances reflect feature probabilities in the system of potential. From the system pole perspective, differences in feature probabilities characterise different registers; from the instance pole perspective, differences in feature frequencies characterise different text types (genres).

[4] "Nothing but" is the language of reductionism (a.k.a. "nothing buttery").

[5] This is prescriptivist pedagogy.  See the is–ought problem, also known as Hume's law, or Hume's guillotine.

[6] To be clear, the question is: 'how far up the cline of instantiation does a feature extend?  Is it only a feature of one text?  Is it also a feature of its text type?  Is it also a feature of the general system of potential?

* Note that, in order to avoid ambiguity, Halliday sometimes distinguishes between 'grammar' (data: the phenomenon modelled) and 'grammatics' (theory: the model of the phenomenon).