Thursday, 5 December 2013

David Rose On Lexis, Grammar And Instantiation

On 4/12/13, David Rose wrote on the Sys-func and Sysfling lists:
PS My own view is we need take Ockham's Razor to the 'grammarian's dream', and handle lexical choices as realising discourse semantic features. Hasan showed the possibility of drawing system networks from grammatical to lexical features in one small region of relation process types, but it has gone no further as such networks soon become impossibly complex, and self-contradictory. To me it seems simpler to treat lexis and grammar as co-instantiating. (Actually I think that is what linguists do tacitly with their examples all the time.)

Blogger Comments:

[1] There is no need to take Ockham's Razor to the 'grammarian's dream' of lexis as most delicate grammar, since on the stratificational model, the lexicogrammatical stratum — from grammar to lexis — is already construed as realising the semantic stratum.  On the other hand, we could use Ockham's Razor to excise 'discourse' from 'discourse semantics' because the term is redundant.

[2] On the complexity involved in elaborating the lexicogrammar, Halliday & Matthiessen (2004: 46) write:
It would take at least 100 volumes of the present size [ie of IFG] to extend the description of the grammar up to that point [of maximum delicacy] for any substantial portion of the vocabulary.
[3] On the SFL model, grammar and lexis are the same phenomenon, lexicogrammar, viewed from different ends of the scale of delicacy.  Given that the process of instantiation is the selection of features and the execution of realisation statements, the instantiation of grammatical systems and lexical systems is the same process viewed from different ends of the scale of delicacy.

The notion of 'co-instantiation', like 'de-instantiation' and 'distanciation', betrays a misunderstanding of the theoretical valeur of instantiation.

Tuesday, 3 December 2013

Mick O'Donnell On Process Types And Projection

On 3/12/13, Mick O'Donnell wrote on the Sys-Func and Sysfling lists:
One thing I will say: IFG 3rd ed. (2004) was a result of extensive addition by Christian of Halliday's IFG 2.  One thing I have noticed is that Christian tends more to notional coding than Halliday, swayed more by semantic similarity, while Halliday is more grammatical in his coding
For instance, there is a new example in IFG3 involving "talk .. .about" being a verbal process, while the previous page (and IFG 1) list"talk" as behavioural.
In any case. the BNC lists one example of "mourn" with a projection:
"The steward was not left to mourn that his bottles found no custom"
Admittedly, one instance is not substantial. But this sentence reads ok to me at least.
The principle as I understand it is that the verb has to have the POTENTIAL to project verbally to be considered mental or verbal. Otherwise it may be behavioural (as in "talk").

Blogger Comments:

[1] Matthiessen, like Halliday, advocates taking a trinocular perspective on the grammar, looking at it 'from above', 'from roundabout' and 'from below'. This is in contradistinction to "notional coding".

[2] The verb talk can serve as verbal process or behavioural process, depending on how it functions in a clause.  O'Donnell again makes the error of taking the verb itself as the point of departure for determining the function of an element of clause structure.

[3] This example is a single clause with an embedded fact — (the fact) that his bottles found no custom — not a clause nexus of projection.

[4]  The principle that O'Donnell "understands" is entirely of his own making, and one source of his confusion.  There are subtypes of both mental and verbal processes that either don't have the potential to project, as in mental processes of perception and emotion, or are relatively unlikely to project, as in 'targeting' verbal processes.

Mick O'Donnell On Perception Verbs

On 3/12/13, Mick O'Donnell wrote on the Sys-func and Sysfling lists:
IFG3 lists the following perception verbs:
perceive, sense, see, notice, glimse, hear, overhear, feel, taste, smell
And note these can all project (with reservations on the last two):
I perceived that he was uneasy.
I sensed that he was uneasy.
I saw that he was uneasy.
I noticed that he was uneasy.
I glim[p]sed that he had the Joker.
I heard that you won.
I overheard that you like cheese.
I felt that I should leave.
? I smelt that she had been here
? I tasted that she had used garlic

Blogger Comments:

There are two problems here.  

[1] The first problem is O'Donnell's continual mistaking of embedded clauses for projected ones, as demonstrated in previous critiques.  As a general guide, whenever that can be replaced by the fact that, we know we are dealing with a 'fact' clause, and so with a clause that is embedded as a constituent within another clause.

[2] The second problem is the analytical mistake of using the verb itself as the point of departure for determining the process type of a particular clause.  A specific verb can serve as various types of Process, as when see functions as a mental process of cognition, rather than perception.

Mick O'Donnell On Mental Processes

On 3/12/13 Mick O'Donnell wrote on the Sys-func and Sysfling lists:
As for your "get a bird's eye view" being mental, if it can't clausally project, it can't be mental, if we take Halliday's criteria seriously:
* I got a bird's eye view that it was beautiful
Note however:
I got the idea that you were here.
...that potentially projects, so could be mental.  An alternative analysis would place the "that" clause as a postmodifier to "idea", which would dismiss the mental analysis.
Blogger Comments:

[1] If we do indeed "take Halliday's criteria seriously", then the ability to project is not a necessary criterion for mental processes.  Only the 'higher' mental processes of cognition and desideration have the potential to project; the 'lower' mental processes of perception and emotion cannot project, though they can range over (embedded) pre-projected facts, as in He heard (the fact) that you were ill.

(the fact) [[ that you were ill ]]
Process: mental
Phenomenon: metaphenomenon

[2] Again, this is not a projection nexus — but it is a mental clause.  The embedded Postmodifier does not "dismiss the mental analysis":

the idea [[ that you were here ]]
Process: mental
Phenomenon: metaphenomenon

As demonstrated elsewhere here, O'Donnell does not understand the distinction between projection nexuses and simple clauses with embedded projections.

Monday, 25 November 2013

John Bateman On Doing Transitivity Analyses By Swapping Verbs

To give any support for interpretations where transitivity comes out looking like a classification of verbs is going to get us into trouble and potentially confuse (or mislead) those just beginning, surely. 
So, to get a dig in at Brad, doing transitivity analyses by swapping verbs is, well, well down that path

Blogger Comments:

On the one hand, this is the opposite of what is true, since, if two verbs are interchangeable in a clause, then it is the Process function of the clause — that either serves — that is being identified, rather than "a classification of a verb".

On the other hand, there are instances where "swapping verbs" is recommended as the 'best strategy' for transitivity analysis. Halliday & Matthiessen (2004: 233):
With a verb other than be it is clear which is Token and which is Value, since … this can be determined by the voice: if the clause is ‘operative’, the Subject is Token, whereas if the clause is ‘receptive’, the Subject is Value. … With the verb be one cannot tell whether the clause is ‘operative’ or ‘receptive’; the best strategy for analysing these is to substitute some other verb, such as represent, and see which voice is chosen.

Thursday, 17 October 2013

Mick O'Donnell On Projection And Clause Complexing

On 17/10/13, Mick O'Donnell wrote to the sys-func and sysfling lists:
Truing [sic] to get my head around Halliday's treatment of projection as clause complexing:
a) Many believe that John is right.
b) That John is right is believed by many.
If (a) Is a clause complex, how do I analyse (b) which seems to be the passive form of the same?

Blogger Comment:

(a) is not a clause complex.  It is a clause simplex with an embedded projection (i.e. a 'pre-projected fact') serving as Phenomenon, and (b) is its receptive agnate.


From 'roundabout': (a) is agnate with Many believe the fact that John is right

From 'above': the projection was not created by the process of believing.  It was projected by a previous figure of a senser sensing or sayer saying (i.e. 'pre-projected'), and the process of believing ranges over it as a metaphenomenon.

See a transitivity analysis here.

See why (ranking) projections are not clause constituents here.

Thursday, 10 October 2013

David Rose On Epistemology

On 10/10/13, David Rose wrote on sysfling:
We have what is claimed to be a 'mimetic' theory… that describes language "by mimicking its crucial properties” (Halliday 1996), a claim that the theory is more than just a 'model'.
It shouldn't therefore be surprising that the list exchange is as much interpersonal as ideational, if the theory can describe how institutional power relations are realised in texts, alongside the categories and relations of the theory.
Conversely, a claim that the architecture of the theory is independent of institutional power relations would contradict the metafunctional dimension of the theory. From the interpersonal perspective of the theory, such a claim would have to be interpreted as a strategy to manipulate institutional power relations.
In order to understand and perhaps reconcile our apparently incommensurate versions of the theory, Im sure we have to honour its breadth, and admit our own varying institutional positions. These include the focus of our specialisations on particular strata, as discussed earlier, alongside our personal affiliations (and antipathies) to supervisors, colleagues and institutions.
If we could acknowledge but tease apart the personal from theoretical, perhaps it would be possible to admit that genre, register and discourse semantic theory may indeed mimick the properties of the system itself, although they may be unnecessary for the institutional work of many linguists.

Blogger Comments:

[1] Halliday's claim does not assert that 'the theory is more than just a model'.   Ideationally, language is an evolved theory of experience, and a linguistic theory is a designed theory of language; not just a model — a model.  Given that language and linguistic theories are both semiotic systems, they both share the properties of semiotic systems, and it is in that sense that a theory which construes language as a semiotic system is mimetic ("mimics its properties").

[2] To claim that the (ideational) architecture of a theory — think of general relativity or quantum mechanics, for example — is dependent on (interpersonal) "institutional power relations" is manifestly absurd.  Being ideational, the architecture depends on agreement with experience and logical consistency.  It is the architecture of religious dogma that potentially relies on "institutional power relations"; for example, because the Catholic Church included the Aristotelian cosmogony in its dogma, the Pope had Galileo imprisoned and Giordano Bruno burnt at the stake for supporting cosmogonies that threatened their "institutional power".  Cf Halliday & Matthiessen (1999: 618):

… the ontogenetic perspective shows that in fact our experience is being ongoingly reconstrued and recategorised as we grow from infancy to maturity.  This is the outcome of processes taking place in human history — evolutionary events that are at once both material and semiotic, and that cannot be reduced to either purely physical processes driven by technology or purely discursive processes driven by ideology.
[3] It is precisely those of us who are outside the "institutional power relations" who are free to make public our reasoning without fear of recrimination, and who have nothing to gain from "manipulating" such relations, even if we were motivated to do so, or if such a thing were possible.

[4] Rose's use of admit and acknowledge harks back to a previous post in which he judged those who didn't agree with him as obstinate (and himself as capable);  see an Appraisal analysis here.

[5] "Genre, register and discourse semantic theory" may well mimic the properties of language in some way or other.  The problem with them is that theorising them as strata is inconsistent with the rest of the theory in which they are placed — and it is the rest of the theory that gives these theories their "institutional power".  It is clearly this last point that Rose is so afraid of — the loss of "institutional power" — given his continual raising of the issue.

John Bateman Misunderstanding Stratification And Realisation

On 10/10/13, John Bateman wrote on the sysfling list:
I'm afraid we know rather less than we might hope (or think?) on some of these issues though.
However [quoting the blog post below],
by definition, the relation between different levels of abstraction is realisation. That is what the theoretical architecture means. The reason why it is difficult to apply realisation to register and genre as strata is that they are not levels of abstraction higher than the strata of language.
This is circular unfortunately; even though I might agree with it :-) 'The theoretical architecture' is then one of strata related by realisation. Realisation is not appropriate to register and genre because register and genre are not strata. Register and genre are not strata because that is not what 'the' theoretical architecture says.
I agree that terms *within* theories should be used in ways that are consistent within those theories; but across different theories such consistency may not be present. And I think we are dealing with different theories here (in a technical sense of different organisational principles, different predictions, different descriptions).

Blogger Comments:

[1]  See appraisal analysis here.

[2] The argument is demonstrably not circular — it is merely misunderstood.  The theoretical architecture is an hypothesis for managing the complexity of language.  It has to be self-consistent, otherwise any theory building on its design will collapse.  The stratified model is a means of parcelling out the complexity by distributing it over different levels of symbolic abstraction, and different levels of abstraction are related by realisation.  The reason they are related by realisation is because, in terms of the grammar, the verb 'realise' serves as an intensive identifying process that relates two levels of abstraction, Token and Value; the lower level Token realises the higher level Value.  (This is theory turned back on itself.)

So, for theoretical constructs to be proposed as strata in this model, they must be consistent with its defining principles; the proposed stratum must be more abstract than the stratum below it — it must be realised by the stratum below — and it must be less abstract than the stratum above — it must realise the stratum above.  If this principle is violated when proposing strata, the model becomes internally inconsistent.  I have given reasons several times why placing register and genre is this model makes the model internally inconsistent.

[3] We are indeed dealing with different theories with different organisational principles etc.  That is the point. Halliday and Martin don't mean the same thing by 'register' or by 'context', and genre is largely not theorised on SFL principles.  When Martin's register and genre are inserted into Halliday's stratification model, it creates an internally inconsistent theory.

John Bateman Misunderstanding Realisation And Instantiation

After Annabelle Lukin wrote on sysfling on 8/10/13:
The key issue is whether you think it matters if you treat situation and culture as in a realization relation (Martin) or in a relation of instantiation (Halliday). I continue to be persuaded by Halliday’s view on this matter. Given that these are fundamentally different relations,
John Bateman replied on 9/10/13:
realisation and instantiation are indeed necessarily fundamentally different. I don’t see that we can have one without the other in any sensible fashion, and just what inter-stratal realisation is, as indicated above, certainly needs far more work to say that it does or does not apply. Some relation of its kind will probably be necessary, otherwise there is no connection between strata—so we may as well call it realisation. This will involve complex patterns of instantiation and who knows what else. I wouldn’t want to give any of those dimensions of description up before we’ve tried many more detailed models out. I doubt if philosophising on the possible inter-relationships of the dimensions of the theory is going to take us all the way (although it helps of course).
Blogger Comments:

[1] The issue that Lukin raises is not a matter of "giving up" either realisation or instantiation as dimensions of the theory.  It is a matter of which theoretical dimension applies in the relation between context of culture (Martin's 'genre') and context of situation (Martin's 'register').

[2] By definition, the strata represent different levels of symbolic abstraction, and, by definition, the relation between different levels of abstraction is realisation.  That is what the theoretical architecture means.  The reason why it is difficult to apply realisation to register and genre as strata is that they are not levels of abstraction higher than the strata of language.  The term 'register' denotes a functional variety of language, and as explained elsewhere on this site, functional varieties are not higher levels of symbolic abstraction.  If the term 'register' is not used in the sense of a functional variety of language, why use that term?  The problem with theorising genre as a stratum is more complex, because it is largely not theorised in ways that are consistent with SFL theory as a whole — for example, there is no metafunctional differentiation — and components of the genre model are scattered across the theoretical architecture of SFL; some suggestions on this can be read here.

[3] Relations between strata do not involve "complex patterns of instantiation" because, by definition, instantiation is not a relation across strata; instantiation is the relation between the system and instance, on each stratum.

[4] There is no need to philosophise on "the possible inter-relationships of the dimensions of the theory"; all that's needed in this regard is an understanding of the dimensions of the theory, as theorised.

Wednesday, 9 October 2013

David Rose On The "Discovery" Of Theoretical Constructs

On 2/10/13 David Rose wrote on sysfling:
… the discovery that the contexts of language are also stratified as register and genre.
and on 4/10/13:
Halliday's major specialisation has always been grammar, and one of his key discoveries (yes) was the metafunctionality of grammar in relation to field, mode and tenor.

Blogger Comment:

It's not necessary to delve into the history of philosophical thought on epistemology to discount Rose's claim that theoretical constructs are "discovered" as if material phenomena; the theory of experience that has evolved in the grammar of English is instructive enough.

Semiotic phenomena — metaphenomena — such as 'stratified context' and 'metafunction', are created by being projected onto the semiotic plane by a senser sensing (ideas) or by a sayer saying (locutions).  The sense in which semiotic phenomena are 'discovered' is restricted to the sensing of ideas or locutions that have already been created by a senser sensing or a sayer saying (i.e. pre-projected facts).

Thursday, 3 October 2013

David Rose On Stratified Context

On 2/10/13 David Rose wrote on sysfling:
The historical materialist perspective that underpins Halliday's life work, and the theory we all work with, also suggests that our consciousness tends to be shaped by our position in the social division of labour, a view that Bernstein among many others has elaborated.
There is also a social division of labour within the field of production of SFL theory, organised along the lines of the theory itself. Some of us are specialist phonologists, others grammarians, others discourse analysts, and many others are specialists in fields beyond linguistics.
It is a matter of interest to me, and perhaps worth researching further, that the latter two groups have provided the bulk of contributions to the explosion of research generated over the last 30 years by the discovery that the contexts of language are also stratified as register and genre. On the other hand, of those among us who still seem to find most difficulty with accepting this concept, the first two groups seem to predominate.

Blogger Comment:

If the problem with Martin's model — the stratification of the context as register and genre — were merely the negative attitude of some linguists toward it — see an Appraisal analysis of Rose's attitude here — the proponents of the model would have nothing to fear.  It is the fact that there are valid reasons for rejecting the stratified model of context that guarantees it will not survive long in its current form.

It is not genre theory that is the problem — it stands alone as a distinct theory  — it is its construal with the SFL architecture as a stratum of context that is the problem.  For suggestions on how genre is scattered across the SFL architecture, see arguments here.

On the other hand, Martin's model of register is problematic even if it is not construed as a stratum of context — but that's another matter.

Wednesday, 2 October 2013

David Rose On Text Type [2/10/13]

On 2/10/13, David Rose wrote on sysfling:
In a cline of instantiation, both genre and register lie at the system end, and text type is the next step

 Blogger Comments:

[1] On the stratified model (Martin 1992) that Rose advocates, genre and register are distinct strata.  Instantiation is the relation between system and instance.  It is not a relation across strata.  So, in terms of Martin's stratification — to be consistent with the notion of instantiation — the cline for each distinct stratum would be the relation between:
  • genre as system and genre as instance
  • register as system and register as instance

[2] In Martin's stratified model, genre and register are strata of context.  Here Rose is construing text type — i.e. language instead of context — as a point on the cline of instantiation of context.

Rose's construal of text type is quite difficult to unpack, owing to the fact that:
  1. the model that Rose advocates is full of internal inconsistencies, and
  2. Rose's construal of that model is inconsistent with the model itself.

David Rose On Text Type [1/10/13]

On 1/10/13, David Rose wrote on the sysfling list:
One use of 'text type' is to avoid the word genre, by pedagogues who oppose explicit genre pedagogy, and linguists who oppose a stratified context model.
Another is for a node in an instantiation cline starting with genre at the system end, and reading at the instance end.
… genre - text type - text - reading
In other words, text type is more specific subset of texts of the same genre.

Blogger Comments:

[1] In SFL theory, the term 'text type' has a precise definition.  It is the same point on the cline of instantiation as register.  Text type is register viewed from the instance pole of the cline, just as register is text type viewed from the system pole of the cline.  

This is a separate issue from any "opposition" to Martin's stratified model of context.  As argued elsewhere on this site, Martin's stratified model of context is based on a misunderstanding of the meaning of both stratification and context in SFL theory.

[2] The cline of instantiation Rose presents above is that of Martin, with the terms 'system' and 'register' (strategically) omitted.  In Martin's model, register is conflated with genre, and both are points on the cline; ie genre is not at the system pole of the cline.  If genre is placed at the system pole, this cline is concerned only with instantiation on Martin's stratum of genre, thereby restricting the notions of text type, text, and reading to that stratum.

As argued elsewhere on this site, it is logically inconsistent to place 'reading' on this cline because the relation between reading and text is not the same as the relation between text and text type.

As also argued elsewhere on this site, Martin's cline of instantiation is logically inconsistent with his stratified model of context.  This is because his stratified model construes genre and register as more abstract than (the strata of) language — a construal inconsistent with the SFL notion of register — while his instantiation cline construes genre/register as a subpotential of (the system of) language.