Friday, 26 October 2012

Robin Fawcett On Modulation

On 25/10/12 on the Sysfling list in an email titled 'on the modal operator', Sharon Shaw (肖祎) asked:
how do you classify the phrase "be necessary" and "be allowed to" in the modal system? do they belong to the modal operator ?

to which Robin Fawcett replied:
A good question!  As it happens, I can provide you with an answer - or rather, the two answers that are needed, because the two cases are different.  Notice that we can say "Ivy is allowed to eat that" but not "Ivy is necessary to eat that " - though we can say "Ivy is necessary (to/for the success of this project)".
Let us take "be necessary" first.  I'm guessing, but I suspect that Sharon may be thinking of examples such as 
"It is necessary (for Ivy) to be there."  
This is an example of the construction that is best described, in functional terms, as an "evaluative enhanced theme" construction.  (This is in contrast with the "experiential enhanced theme" construction, as throughly researched and described in Huang 1996 and 2003.)  Thus the experiential meaning of "It is necessary (for Ivy) to be there" is the same as that of "For Ivy) to be there is necessary", and in both cases the analysis is that "(For Ivy) to be there" is a Subject/Carrier and "necessary" is a Complement/Attribute.  
But in our example the Performer has chosen to make the evaluative Attribute "necessary" the Theme, and to "enhance" it by preceding it by the experientially empty Subject "it" (followed by a form of "be").  See Chapter 10 of Fawcett 2009 for a full description of this construction and its meanings.    
What about Sharon's second example of "be allowed to"?   This is just one example of an area of the lexicogrammar that is seriously underdescribed (even in Quirk et al 1985).  It is the area of meanings and forms that are typically realized in a modal verb that expounds the (Finite) Operator - but which are also realized in expressions such as "be allowed to", "be able to", "be likely to", "be supposed to", "be going to" and "have got to" (to cite just the few examples that most big grammars refer to, typically as minor aberrations).  But such examples are merely the tip of a pretty large iceberg, and the description in IFG is, as Halliday himself says (1994: 362), "no more than a thumbnail sketch" (e.g. it surprisingly omits "be able to", "be obliged to" and "be inclined to").   (The relevant section in Halliday 1994 is pp 355-63, this being unchanged in Halliday and Matthiessen 2004: 616-25.)
The need for a fuller SFG description of this area was the reason why, in my contribution to the 'second round' of  festschrift volumes for Michael Halliday published in 1907 (Continuing Discourse on Language: a Functional Perspective, edited by Hasan, Matthiessen and Webster), I chose to provide a SFG account of this area of the lexicogrammar.  I wanted to contribute a description of an area that (i) was reasonably fully described in the Cardiff Grammar and - in the context of a volume that was intended as a tribute to Halliday - (ii) was not a direct alternative to the description in IFG.  Indeed, my hope was that it be a contribution to the common pool of SFG descriptions, and one that could be incorporated in any future revision of IFG.*
The statistics in Biber et al 1999 suggest that in most text types well over one in every 200 words is of the type illustrated here, so this is no way a fringe area of the lexicogrammar of English. It demands a description in its own right, rather than as an a"add-on" to an account of the modal verbs. So the full description of this area constitutes a major paper, which is probably to long to be accepted as an attachment on sysfling. So I attach the first section and, if you wish to have the full paper, complete with diagrams of how to analyse texts that contain such examples, please email me and I will send you a copy.
As I say in the above attachment, a description of this area of meanings and forms requires us to recognize more elements in the structure of the clause than are used in IFG , because more that one such trio of Auxiliary + Auxiliary Extension + Infinitive may occur in the same clause, e.g.
"He is bound to going to be willing to attend the meeting (if she asks him nicely)."
In the Cardiff Grammar these are all treated as elements of the clause rather than the "verbal group" - a unit that many systemicists - and not only those who use the Cardiff Model - have long since rejected. For a brief summary of the reasons for treating the elements of Halliday's "verbal group" as elements of the clause and for references to the the publication that sets out the full argument, see Footnote 27 on pp 49-50 of Fawcett 2008 (which also provides a readable introduction to the Cardiff Grammar's method of clause analysis). Alternatively, if you wish to use Halliday's verbal group you can simply treat these elements as elements of it.

Blogger Comment:

Neither of Sharon's examples includes elements that function as Modal operators in the verbal group.  As Halliday & Matthiessen (2004: 592) point out:
the semantic domain of modality is construed in more than one place in the grammar; for example, it is construed by clauses such as I suppose and it is possible, by verbal groups with finite modal operators such as may and by adverbial groups and modal adverbs such as perhaps.
In it is necessary, the semantic domain of obligation is realised grammatically as an attributive clause; while in Ivy is allowed to eat that, obligation is realised grammatically as a verbal group complex (see Halliday & Matthiesssen 2004: 513) — which is, of course, why such cases are not discussed in their section on the verbal group.  (Fawcett's error on this latter point may be the source of Mick O'Donnell's similar confusion; see Mick O'Donnell On Process Types.)

See also: 
Robin Fawcett On The Verbal Group 
Qualities Of Projection

Monday, 1 October 2012

David Rose On Cause

On 30/9/12, on the sysfling list, Sabiha Choura wrote:
I have met the following difficulty while analysing my corpus:
Adiposity and low aerobic fitness in children are associated with a clustering of cardiovascular risk factors.
Is the process "are associated" an identifying circumstancial process expressing accompaniment or an identifying intensive process?

To which David Rose replied:
Cause = time x obligation
Obligation is gradable, so is cause
may do : will do : must do ::
are associated with : arise from : are caused by

Blogger Comments:

[1] Viewed from 'round about', the clause is intensive, not circumstantial, since it has an implicit Assigner, and the system of ASSIGNMENT has the feature 'intensive' as an entry condition.  Click here for a transitivity analysis.

[2] Cause is not "time x obligation"; cause and time are features of ideational construals, whereas obligation is  a feature of interpersonal enactment.  The ideational and interpersonal metafunctions are distinct perspectives on language: language viewed in terms of its function of construing experience vs language viewed in terms of its function of enacting interpersonal relations.

[3] "Obligation is gradable" — so is the greyscale from white to black.  This red herring exemplifies the type of logical fallacy concerned with fallacies of relevance: attempts to prove a conclusion by offering considerations that simply don’t bear on its truth.

[4] May and will are low and median values, respectively, of probability (modalisation) not obligation (modulation).

[5] The meanings are associated with and arise from are not low and median values of are caused by.  See, for example, correlation does not imply causation.