David Rose On The Tension Between Lexis And Grammar

David Rose wrote to sys-func on 21 Oct 2018 on 10:08 a message titled 'Whale snack?':

Headline in Sydney Morning Herald :)
Whale snack stops challenge feast or famine myth: Australian Museum

and on 24 Oct 2018 at 09:46:

... a revealing illustration of tension between lexis and grammar. To make sense of the grammar you have to know the field connoted by the lexis! So you have to read the article. What an ingenious hook :)

https://www.smh.com.au/environment/conservation/whale-snack-stops-challenge-feast-or-famine-myth-australian-museum-20181017-p50a8r.html

Blogger Comments:

[1] The instance does not illustrate tension between lexis and grammar.  A genuine example of "tension" between lexis and grammar would be the use of a lexical item like giggle as a projecting verbal Process, as in:

"You're stupid!" the child giggled.

In such cases, the lexical choice adds a behavioural feature to a verbal Process.

[2] In SFL, the way to analyse any clause is to take a trinocular approach, shunting between the grammar and the view 'from above' (the meaning being realised) and the view 'from below' (the realisation of the wording at lower ranks).  The difficulty in this clause lies in its grammatical constituency, even though there is only one possible reading, as shown here.

[3] Here Rose repeats Martin's misunderstanding of the theoretical term 'field'.  In SFL, 'field' refers to the ideational dimension of context: the culture as semiotic system; field is not a dimension of language.  When Martin and his students use the term 'field', they are usually referring to a domain, the semantic correlate of a contextual field (Halliday & Matthiessen 1999: 323).

This confusion is further complicated by the fact that Rose, following Martin, misunderstands field as a dimension of register, where register, a context-specific subpotential of language, is misunderstood as context, instead of language.

Evidence of Martin's (1992) misunderstandings of SFL theory can be viewed here; evidence of Martin & Rose's (2007) misunderstandings of SFL theory can be viewed here.

David Banks On "Complex Verbal Groups"

In answer to the query:

1. In a structure like: "Honey, don't forget to pay the bill before the week is over" should I say that 'forget to pay' is a complex verbal group? O should I say that 'forget' is a mental process that projects the clause complex 'to pay the bill before the week is over'? 

2. If I compare these two uses of "want":

1. "I want to visit the Vatican"
2. "I want Mary to visit the Vatican"

I'd say that in number 1, 'want to visit' is a complex verbal group, where 'want' shows inclination phase, while in number 2, 'want' is a process that projects 'Mary to visit the Vatican' am I right?  

3. Finally, a case similar to the one with 'forget': in a clause like "that afternoon, we decided to go over the wall" should I say that "decided to go" is a complex VG or should I say that "decided" is a mental process that projects "to go over the wall"

 David Banks wrote to sysfling on 5 October 2018 at 19:47:

1. It seems to me that you can only say that it is a complex verb group if there is only one process involved (e.g. continue to pay). Here, it seems obvious that forgetting and paying are quite separ[a]te processes, so it is not a complex verb group. 

2. It does not seem logical to have completely different analyses for these two examples. The fact that English "understands" the subject of "visit" in 1, to be the same as the subject of "want" does not alter the fact that it is basically the same structure as 2. 

3. The answer to this is the same as for No.1.

Blogger Comments:

[1] Trivially, a verbal group complex is not a complex verbal group.  The former is a complex of verbal groups, and a theoretical term, whereas the latter is a single verbal group deemed to be "complex".

[2] To be clear, it is important to distinguish between Process (clause rank) and Event (group rank).  A verbal group nexus construes two Events as one Process; examples provided by Halliday & Matthiessen (2014: 569, 574, 580, 587) include:

• decides to write
• would like to paint
• forget to do
• remind to do
• want to do
• ask to do
• anticipate doing
• profess to do.

That is, the instances in question, forget to pay and decided to go, can be interpreted as verbal group complexes that construe two Events as a single Process.

[3] The two different analyses correspond to the preferred analyses of Halliday & Matthiessen (2014: 584-6), who provide the reasoning on which the analytical distinction is based.

But, to be clear, the two instances do not have "basically the same structure", since, if both are interpreted as projecting clause nexuses, the second projected clause includes a topical Theme, an Actor and an explicit Subject, and so a Mood element, whereas the first has none of these.

However, a functional theory gives priority, in grammatical reasoning, to the view from above: to system and function, rather than to structure and form (Halliday & Matthiessen 2014: 49).  From above, the two instances differ in terms of the interpersonal meaning being realised.  In the first clause complex, I want to visit the Vatican, the desiderative Process projects an offer, whereas in the second, I want Mary to visit the Vatican, it projects a command.  Halliday & Matthiessen (2014: 584) explain:

We saw in Chapter 7, Section 7.5.4.2 that a mental process of desideration projects an exchange of the goods-&-services type, i.e. a proposal. If the Subject of the projection is the same as that of the mental process clause, the proposal is an offer, as in she wants to do it; if the two are different, then the proposal is a command, as in she wants you to do it. In the first type, the Subject is not repeated, but is carried over from the desiderative clause. (It can then be made explicit by a reflexive, as in she wants to do it herself.)

Moreover, the second clause complex is also an instance of an interpersonal metaphor of modality, being the metaphorical counterpart of Mary should visit the Vatican; see Halliday & Matthiessen (2014: 693).

Mick O'Donnell On "Main And Support Verbs"

In answer to the query:

1. In a structure like: "Honey, don't forget to pay the bill before the week is over" should I say that 'forget to pay' is a complex verbal group? O should I say that 'forget' is a mental process that projects the clause complex 'to pay the bill before the week is over'?

 Michael O'Donnell wrote to Sysfling on 5 October 2018 at 18:40:

There are arguments for both analyses, here. 

I would argue for grouping "forget to pay" and "forget that..." as both instances of mental processes.

1. if you follow your precedent and treat "She forgot to pay" as a complex verbal group, then would you do the same with "She SAID to pay at the counter". So, where do you draw the line in determining whether the first verb is the main verb or the support verb? 

2. In Construal Analysis (analysing a document to see what participant roles particular entities fill), it would probably be beneficial to treat the participant as both construed Sensor and Actor. Alternatively, in "he started to think", I don't see the value of recording a participant role for "starting". 

3. In a case like:

A: Did you pay him?
B: I forgot. 

...where the second is elliptical for "I forgot to pay him", it seems strange to say "forget" is not the main verb. It is certainly the main activity reported. 

None of these arguments are conclusive. I think the issue of where the dividing line lies between aspectual verbs and projecting verbs is very blurred, with different linguists placing the involved verbs on different sides. I have always gone with the criteria that if the verb involves mental or verbal action (semantically), then this should be recognised, allowing it to group with the other cases where the projection takes the form of a that-clause, a present participle clause, or a nominal group.

Blogger Comments:

[1] Trivially, a verbal group complex is not a complex verbal group.  The former is a complex of verbal groups, and a theoretical term, whereas the latter is a single verbal group deemed to be "complex".

[2] To be clear, while 'forget (that…)' is unambiguously an instance of a mental Process, 'forget to pay' can alternatively be analysed as a verbal group complex that construes two Events, 'forget' and 'pay' as a material Process; see, for example, Halliday & Matthiessen (2014: 574).

[3] This is a non-sequitur, since the two instances differ markedly.  In She forgot to pay, a proposition is mentally projected with the same Subject as the projecting clause; whereas, in She said to pay at the counter, a proposal is verbally projected with a different Subject from the projecting clause.

[4] To be clear, in terms of SFL theory, the question is whether

• the verbal groups form a complex to realise a single process in a clause simplex, or
• each verbal group realises a different process in a clause complex.

What O'Donnell calls "the main verb" is the verb serving as the Event of the verbal group that serves as the Process of a projecting clause in a clause complex; what O'Donnell calls "the support verb" is the verb serving as the Event of the projecting verbal group of a verbal group complex.  In focusing on verbs, O'Donnell is focusing on form (the view from below) instead of function (the view from above).

[5] Here O'Donnell is largely consistent with Halliday ± Matthiessen (1985, 1994, 2004, 2014) — though unwittingly so, since he presents the analysis as his own.

[6] For reasoning with regard to the distinction between projection in clause complexes and projection in verbal group complexes, see Halliday & Matthiessen (2014: 584-92).

[7] Here O'Donnell misinterprets the question of what rank complexing takes place, clause or group, as a distinction between "aspectual and projecting verbs".  Moreover, this distinction reduces the vast array of potential relations between (what are) verbal groups in verbal group complexes to time phase elaboration ("aspectual") and projection.  The systemic potential of verbal group complex is given in Halliday & Matthiessen (2014: 584-92) as:

[8] To be clear, the issue is not what linguists do, but whether analyses are consistent with the theory said linguists are using.

[9] This criterion is oblivious to the question of what rank, clause or group, the "mental or verbal action" is to be "recognised".