Mick O'Donnell On Semantics And Grammar

Mick O'Donnell replied to David Rose on Sys-func on 11/7/22 at 17:57:

I would not think any Hallidayan would ever posit a 1-to-1 mapping between Semantics and Grammar. That would invalidate the need for Semantics as a separate stratum. …

So, the mapping between Semantics and Grammar always needs to be seen as a many-to-many mapping.

- language provides multiple forms for realising a particular meaning.

- a particular form can express alternative meanings

(but yes, "Construing Experience" can be read (misread?) to suggest a one to one mapping between semantic process type selection and Transitivity selection in the grammar. Something I and other have argued against in the past. And not something that I think was intended by the authors, Christian and MAKH).

Blogger Comments:

[1] To be clear, it is metaphor that motivates semantics as a stratum of content. Halliday & Matthiessen (1999: 237):

If the congruent pattern had been the only form of construal, we would probably not have needed to think of semantics and grammar as two separate strata: they would be merely two facets of the content plane, interpreted on the one hand as function and on the other as form.

[2] Here O'Donnell misconstrues the distinction between semantics and grammar as a distinction between meaning and form. To be clear, the distinction in SFL Theory is between meaning and wording, where wording is form (the rank scale) interpreted in terms of its function of realising meaning; e.g. verbal group (form) as Process (function).

[3] This is misleading. On the one hand, PROCESS TYPE is a grammatical system within the larger system of TRANSITIVITY (Halliday & Matthiessen 2014: 219, 354-5), so, in this sense, the notion of a 'one to one mapping' of 'process type selection' to 'transitivity selection' is nonsensical.

On the other hand, in the absence of grammatical metaphor, the process type served by a verbal group is congruent with its semantic value. Moreover, in the case of grammatical metaphor, the incongruence is not between one process type and another, but between a process and some other element, as when a Process is realised grammatically by a Range participant, as he had a bath.

[4] To be clear, O'Donnell misunderstands the system of PROCESS TYPE, largely because he fails to recognise verbal group complexes. See, for example:

Mick O'Donnell On Process Types [1]

Mick O'Donnell On Process Types [2]

Mick O'Donnell On The Number Of Process Types

Mick O'Donnell On Process Types

Mick O'Donnell On The Difference Between 'Should Not Fail To See' And 'Should See'

David Banks On Process Type Criteria

David Banks replied to David Rose on SYSFLING on 6 Mar 2022, at 20:40:

With all due respect:

A happy baby makes a happy mother.

You say this must be identifying, so the baby and the mother refer to the same entity - that can't be right.

It's simple present — I see no reason why a simple present cannot be material: physical actions are material whether they are habitual or not.

"'I walk on the beach every day".

And you say it's reversible: this is perhaps true in the pragmatic sense that if the mother is happy the baby probably will be too, but that's not what the clause says, and "A happy mother makes a happy baby" is not the same thing as "A happy baby makes a happy mother". What about 

"A sunny day makes a happy tourist"?

What have I missed?

Blogger Comments:

[1] This is misleading, because it is not true. Halliday & Matthiessen (2014: 276):

In the ‘identifying’ mode, some thing has an identity assigned to it. What this means is that one entity is being used to identify another: ‘x is identified by a’, or ‘a serves to define the identity of x’. Structurally we label the x-element, that which is to be identified, as the Identified, and the a-element, that which serves as identity, as the Identifier.

Cf. the identifying clauses:

• Peter owns the house and 
• asbestos dust causes mesothelioma

[2] This misses the point. It is not a question of whether or not the simple present tense can be used in material clauses, but whether or not it is the marked or unmarked option. Because the simple present tense is the marked option for material clauses, but the unmarked option for relational clauses, it is one diagnostic tool for distinguishing these two process types.

[3] To be clear, the 'reversibility' is one of voice:

• A happy baby makes ('produces') a happy mother [operative]
• A happy mother is made ('produced') by a happy baby [receptive]

and it distinguishes (reversible) identifying relational clauses from (non-reversible) attributive relational clauses. 

[4] See above.

David Rose On The Realisation Of Process Type

David Rose replied to Johanna Salomonsson on Sysfling on 4 Mar 2022, at 09:58:

Two points about SFL that could help your readers…  

1. It differs from traditional grammars that gloss each word and morpheme. In SFL, process type is a clause rank system realised by relations between lower rank units, such as verbs. The whole clause is material or relational. (This is sometimes confused by labelling verbal groups as process types.)

2. Analyses are multiperspectival. In Halliday’s model, process type is a transitive perspective on a clause, and the ergative perspective is complementary.

As you show, your second examples are caused attributive relations, in which Attributor is additional Agent from ergative perspective. The complementary perspectives help to show why clauses such as Tom’s ‘We painted the wall red’ look similar in agency, but differ in process type.

Blogger Comments:

[1] To be clear, this is nonsense. Systemically, PROCESS TYPE is a system of the clause, and a clause consists of groups, including the verbal group, each of which is the entry condition to its systems. Structurally, the Process of a clause is realised by a verbal group.

"Relations between verbs", on the other hand, refers to relations between lexical items, each item being specified by the most delicate lexicogrammatical features. Rose's misunderstanding, therefore, is that a grammatical system is realised by a lexical set.

[2] To be clear, what Rose refers to here is the widespread practice of classifying verbal groups in terms of  PROCESS TYPE, instead of considering the function of each group in a clause. 

But, on the other hand, 'Process' is the function label of a verbal group in a clause.

[3] To be clear, the "second examples" in the discussion were:

A happy baby makes a happy mother

Your happy Glaxo baby will make you a happy mother

but only the second of these is an attributive clause with an Attributor. The first is an identifying clause, with no agency:

That is, the clause identifies a happy baby with a happy mother; it does not construe a happy baby as a member of the class a happy mother, as would an attributive clause.

This is a fact that escaped all interlocutors in this Sysfling discussion, including Jim Martin.

Postscript:

Ten hours after this post was published, David Rose correctly reanalysed the clause in question as identifying. See the next post.

David Banks On Process Types

After Wendy Bowcher wrote to SYSFLING on Thu, 3 Mar, 21:08:

I wonder if anyone can help out with the following:

(1) identifying what type of process 'thrive' is in the following clause. It seems to me to be Relational, but it's also kind of Material with a dash of Behavioural. :)I've given the clause within its cotext. And the clause comes from an historical advertisement.

"Your child will thrive on Glaxo when all else fails because it is a natural substitute for breast milk"

(2) And I just want to check the process type of 'make' in the following clauses - I have analysed these as Relational (attributive, intensive) ('cause to be something'), but I'm not 100% sure:

"A happy baby makes a happy mother"

"Your happy Glaxo baby will make you a happy mother"

David Banks replied to SYSFLING on Thu, 3 Mar, 21:16:

I would have no hesitation in analysing your first example (thrive) as material: it's about growing and increasing in good health, so about physical change. 

And I would have analysed your second example (make) as material too: it's causal, concerned with making the mother better, so again to do with physical change.

 

Blogger Comments:

[1] Transitivity analyses of these clauses are provided here.

[2] To be clear, 'physical change' is not an exclusive feature of material processes, as demonstrated by the attributive clause the baby turned into a pig. Halliday (1994: 120):

Halliday & Matthiessen (1999: 159-165) provide a helpful discussion on material doing and relational being as complementary perspectives on change; (op. cit.: 161):

Jim Martin On Process Types

Jim Martin replied to Mick O'Donnell on SYSFLING on 10 Apr 2021 at 08:58:

It would be unusual to have a ‘process type’ which in effect crossclassifies the others:

They allowed us to leave/chat/ask/believe/be/have…

A case would need to be made.

When not involved in what IFG treats as a verbal group complex, these verbs typically take a complement realising a figure (or one of Hao’s activity or semiotic entities):

They allowed the application/petition/objection/extradition/extract…

I have tried to make a case for Tagalog relationals crossclassifying material, verbal and mental processes, most clearly perhaps in this in this paper:

Relational processes in Tagalog: A Systemic Functional perspective. (J.R. Martin & P. Cruz) in K. Rajandran & S.A. Manan (eds.), Systemic Functional Linguistics: South East Asian perspectives. Universiti Sains Malaysia Press. 2019. 225-251.

Blogger Comments:

[1] Here Martin misunderstands O'Donnell (see previous post). O'Donnell interpreted these types of instances as projection nexuses — clause complexes — not as clause simplexes in which one type of process "cross classifies" another type.

[2] This is misleading, because it is untrue. On the one hand, the Complements of Martin's examples are nominalised processes, and as as such, realise elements not figures. On the other hand, apart from its function in verbal group complexes, and such metaphorical variants, the verb allow typically serves as a verbal Process, as in 

She allowed ['admitted'] that the penalty appeared too harsh for the crime

Bush allowed ['conceded'] that he himself could never support Trump

or a material Process, as in

I'll allow ['allocate'] you just ten minutes.

[3] To be clear, the reason why Martin introduces the irrelevant notion of cross classification here is simply to provide a pretext for promoting his own paper, in which he claims to provide the "case that needs to be made". The function of not clarifying the notion of 'cross classifying process types' is to bamboozle the theoretically naïve.

Mick O'Donnell On Process Types

Michael O'Donnell wrote to SYSFLING on 9 Apr 2021 at 21:35:

In my own practice, I have added a 7th process type modal-process, to cater to this type of process. For example, in English:

He allowed me to smoke.

They required us to teach long hours.

Verbs such as allow/permit/authorise, require/oblige, all allow projection. Yet they are not verbal in the sense that they do not necessarily involve communication, no addressee, etc. They are not mental, in that they are externalised in the real world, not just mental activity.

One could say that these last exclusion criteria are notional, while process type classification should be based on grammatical reactance. But in practice, our process type classification has always been somewhat of a hybridisation of semantic classification linked to common grammatical reactance. Halliday actually groups together various subtypes of verbal process, such as "she blamed him ..." (which do not project), the central verbals (say/tell etc,) which do project, and sometimes the non-projecting verbs such as talk/grumble etc. which in some issues of IFG are verbal. The only thing these all have in common is the shared notional involvement of verbal activity, or expanding out, communicative activity. But in the case of these "modal" processes, there is no necessary explicit communicative activity.

I haven't seen Halliday dealing with these verbs explicitly, but I may have missed it.

Blogger Comments:

[1] Here O'Donnell mistakes allowed and required for Processes of a clause in a clause complex, and on the basis of that misunderstanding proposes a new 'modal' PROCESS TYPE. In SFL Theory, each of these is the primary verbal group of a verbal group complex that serves as a material Process:

In hypotactic verbal group complexes, it is the (last) secondary verbal group that realises the PROCESS TYPE. Halliday & Matthiessen (2014: 568):

It is the secondary group, or last secondary group if there is more than one, that realises the process type of the clause, e.g. [material:] she seemed to mend it, [behavioural:] she seemed to laugh, [mental:] she seemed to like him, [verbal:] she seemed to tell us, [relational:] she seemed to be nice.

[2] To be clear, there is no projection in either of O'Donnell's examples, both of which are clause simplexes, not complexes. The relation between the verbal groups in each nexus is enhancement. See Halliday & Matthiessen (2014: 580, 582-4) on modulation in enhancing causative verbal group complexes.

[3] To be clear, this argumentation is irrelevant to the analysis of the two examples; see [1] and [2] above.

[4] To be clear, this misrepresents SFL methodology. Analysis requires taking a trinocular perspective, while giving priority to the view 'from above'. Halliday & Matthiessen (2014: 48-9):

We cannot expect to understand the grammar just by looking at it from its own level; we also look into it ‘from above’ and ‘from below’, taking a trinocular perspective (Halliday, 1978: 130–131; 1996). But since the view from these different angles is often conflicting, the description will inevitably be a form of compromise. …

Being a ‘functional grammar’ means that priority is given to the view ‘from above’; that is, grammar is seen as a resource for making meaning — it is a semanticky kind of grammar. But the focus of attention is still on the grammar itself.

[5] This is misleading, because it is not true. In the first edition of IFG (1985), the verbs talk and grumble do not feature at all. In the second edition (1994), these verbs are explicitly listed as exemplifying 'near verbal' behavioural processes. Halliday (1994: 139):

The two verbs are again listed as exemplifying behavioural processes in both the third edition (2004: 251), and the fourth (2014: 302).

[6] To be clear, what (genuine) verbal processes have in common is the symbolic exchange of meaning. Halliday & Matthiessen (2014: 303, 304):

‘Saying’ has to be interpreted in a rather broad sense; it covers any kind of symbolic exchange of meaning … verbal processes might more appropriately be called ‘symbolic’ processes.

[7] To be clear, O'Donnell's notion of "modal" processes derives from a misunderstanding; see [1] above.

[8] When subsequently alerted to Halliday's (1994: 285Iff) analysis by Bartlett, O'Donnell replied:

Totally right, Tom. p287 actually gives examples of force/oblige/allow as causative structures.

Not that I will give up on my own approach.

David Rose On Process Type Features

David Rose wrote to Sysfling on 12 Oct 2017 at 21:08:16:

… Fourth, as Mick points out there are consistent (but also fuzzy) relations between the grammatical categories and domains of experience denoted by certain verbs (for example). I don’t think labels like behavioural, mental and so on are semantic. Rather they denote general fields of experience, beyond language. The grammar organises these fields, while lexical items are far more variable. 

A question I’m interested in is where do these relations come from? I mean our experience is so complex, fluid and variable. It seems a miracle that the grammar organises it as neatly as it does, despite the fuzz.

Blogger Comments:

[1] This misunderstands the theory.  In SFL, the process type features are theorised as both semantic and lexicogrammatical.  Halliday & Matthiessen (1999: 503):

… we treat transitivity both within semantics (the paradigmatic and syntagmatic organisation of figures) and within lexicogrammar (the grammar of transitivity): it is a system construed within the content plane of language — both in the ideational component in the lexicogrammar and in the ideation base. This two-stratal approach to transitivity makes it possible to model the resource of grammatical metaphor and is fundamental to work on multilingual systems for generating text.

[2] On the one hand, this can be read as consistent with SFL theory, and so, as contradictory of Rose's previous statement.  On this reading, process type features, as ideational semantics do realise (Rose's 'denote') field, in the sense of 'field' as ideational context (the culture as semiotic system).  But this is not Rose's meaning here — as suggested by the contradiction — since Rose follows Martin's error in treating 'field' as a dimension of register.

On the other hand, this can be read as inconsistent with the epistemological assumptions on which SFL theory was developed.  On this reading, 'fields of experience beyond language' refers to the experience that is construed as ideational meaning.  On the SFL model, meaning is located within semiotic systems ("immanent"), rather than being transcendent of them.

In claiming that process type features denote (realise) 'fields of experience beyond language', Rose is (a) ascribing meaning to the domain outside semiotic systems, and (b) construing this domain as more abstract meaning than the meaning of language.  See also Philosophical Realism.

[3] On the one hand, this can be read as consistent with SFL theory if 'fields' is understood as the experience that is transformed into meaning, and if 'organises' is understood as 'construes as meaning'.  But this is not Rose's meaning here.  For Rose, 'fields' refers to (transcendent) meanings outside language, and it is the rôle of the grammar to organise them.

[4] This misunderstands the relation between grammar and lexis.  In SFL, lexical items are the synthesis of features of the most delicate systems of the lexicogrammar (just as phonemes are the synthesis of phonological features).

[5] This puzzlement arises from the epistemological inconsistencies identified above in [2] and [3].  The 'relations' arise as ideational meaning, construed of experience, in the logogenesis, ontogenesis and phylogenesis of meaning potential.

Mick O'Donnell On The Number Of Process Types

 Michael O'Donnell wrote to Sysfling on 12 October 2017 at 19:20:

I think part of the problem here is the assumption that all clauses possible within the English language can be explained as belonging to one of 6 different grammatical categories: material, behavioural, mental, verbal, relational, existential. 

I do agree that the bulk of the clauses of the English language conform to basic prototypes within these categories. However, the English language has a long history, and occasionally we find examples which are structurally different from other sentences. "It matters" seems to be one of these. 

Remember, Halliday's process types are supposed to be grammatical categories, linked together not only by common semantics, but also by commonality of form. 

And I have always felt it strange that the "present tense" test has always been the primary (and often only) grammatical criterion given for classifying clauses. 

I also do not accept the criterion that "IFG says so" is a valid criteria, unless IFG contains reasoning to support the categorisation. 

Looking purely structurally, "It matters" is closer in form to:

Life evolved.
Something happened.
It appeared.
He abstained.
He apologised.
He breakfasted.
etc.

Looking through a list 2,700 verbs that cannot take a complement (intransitive verbs), they seem to fall into 2 main semantic patterns:

a) Those that explain happenings (material, no agent): evolve, happen, appear.
b) verbal activity with no verbiage nor addressee: abstain, apologise, snigger, yammer, soliloquize, snitch, compromise

...and maybe a set of (rare) items that do represent semantically a relation to another participant which can be specified via a "to" circumstance:

He mattered (to me)
It bonded (to the wall)
They conformed (to the regulations)

We could then shove such cases under relational, but then it seems our unstated but primary grammatical criteria is: All clauses can be categorised under one of the six named process types. ...even when the verbs at issue reflect substantially distinct structural patterns. That to me is the problem. We are b[e]ing too simplistic, trying to fit the language into a box which is too small for it.

Blogger Comments:

[1] Here O'Donnell offers the historical depth of the English language as support for his contention that the most general types of process — motivated by the data — number more than six.

[2] As O'Donnell makes clear, his approach to process types proceeds from the perspectives of structure and form.  Systemic Functional Linguistic theory, on the other hand, as the name implies, is theorised from the perspectives of system and function.  Halliday & Matthiessen (2014: 49):

Being a ‘functional grammar’ means that priority is given to the view ‘from above’; that is, grammar is seen as a resource for making meaning — it is a ‘semanticky’ kind of grammar. But the focus of attention is still on the grammar itself. 

Giving priority to the view ‘from above’ means that the organising principle adopted is one of system: the grammar is seen as a network of interrelated meaningful choices. In other words, the dominant axis is the paradigmatic one: the fundamental components of the grammar are sets of mutually defining contrastive features.  Explaining something consists not of stating how it is structured but in showing how it is related to other things: its pattern of systemic relationships, or agnateness (agnation).

By adopting a different perspective on the data, O'Donnell is unwittingly using a different theory to interpret the data, thereby yielding interpretations at odds with the original theory.

[3] Strictly speaking, Halliday's process types are both grammatical and semantic.  The two are in agreement (congruent) except in the case of grammatical metaphor.  Halliday & Matthiessen (1999: 503):

… we treat transitivity both within semantics (the paradigmatic and syntagmatic organisation of figures) and within lexicogrammar (the grammar of transitivity): it is a system construed within the content plane of language — both in the ideational component in the lexicogrammar and in the ideation base. This two-stratal approach to transitivity makes it possible to model the resource of grammatical metaphor and is fundamental to work on multilingual systems for generating text.

[4] Here O'Donnell maintains his steadfast ignorance of the trinocular perspective that Halliday used in theorising the grammar; see previous posts here, here, and especially here.  The theory of process types arises from looking at the grammar (1) from above (what meanings are being realised?), (2) from roundabout (what other systemic variants are possible?), and (3) from below (how are they realised?).

[5] This is a statement about the competence of the most intrusive list members, not about the criteria used by Halliday, as set out in IFG.  For example, Halliday & Matthiessen (2014: 354):

[6] In citing IFG, an analyst is, of course, referring the reader to both the theoretical assumptions and the reasoning on which an analysis is based.

[7] Trivially, the clause it bonded (to the wall) is material, not relational, both on the basis of what it means (from above), and what would be the unmarked present tense (from roundabout).

[8] This risks a misunderstanding.  The six named process types are the most general categories only.  The total number of process types construed by the SFL approach is enormous, but won't be known until these most general types are elaborated into more and more delicate categories, to the point where each lexical item is specified.  Halliday & Matthiessen (1999: 198):

… 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).

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.

David Rose On Process Type Vs "Verb Type"

On 20/12/12 David Rose wrote on the sys- func and sysfling lists:

It is very common to confuse process type with type of verb, but they are at different ranks in the grammar.

This confusion is possibly not helped by expert discussions that often elide this distinction, for example of 'lexis as delicate grammar'.

Blogger Comments:

[1] process type is a functional system at the rank of clause, whereas "type of verb" is a type of form, like 'clause type' or 'group type', and as such, is not a functional system at word rank.

In SFL theory, the functional system that is realised by lexical verbs is the verbal group system of event type,  which is concerned with the temporal properties of verbs and not, for example, their potentiality in realising process type.  Halliday and Matthiessen (2004: 348):

… the system network of the verbal group is a network of systems representing contrasts that are purely grammatical in nature.  The only system that extends in delicacy towards distinctions that are realised lexically is the system of event type — the verbal group analogue of the thing type system in the nominal group.  This system is concerned with distinctions among verbs relating to their temporal properties (thus complementing the clausal system of process type, which is concerned with distinctions among processes relating to configurations of process plus participants).

[2] Expert discussions of 'lexis as delicate grammar' are not even concerned with the distinction between process type and "type of verb" — let alone elide it — and any discussion that confuses either function with form, or the rank scale with the scale of delicacy is manifestly inexpert.

Mick O'Donnell On Process Types [1]

In a discussion on the email list sfl_education on 23/2/12 concerning process types, Mick O'Donnell writes:

Cases where we don't really have a category: For me, we really need a 7th category for modal lexical verbs, e.g., I was required to leave (also allowed, permitted, obliged, etc.) I find it hard to say these are auxiliary verbs.

Blogger Comments:

[1] On the SFL model, examples such as  I was required/allowed/permitted/obliged to leave do not involve an auxiliary in a verbal group.  Instead, these involve a hypotactic verbal group complex.  Halliday & Matthiessen (2004: 513):

… there is a special set that only exist as causatives, where the meaning is simply that of agency: make, cause, force require, let, allow, permit, etc.  These admit of three degrees of modulation …

[2] O'Donnell's "I find it hard to say" — like "I cannot believe" — is an example of the logical fallacy known as the argument from personal incredulity.

Mick O'Donnell On Process Types [2]

Mick O'Donnell replied on 24 February 2012 08:27:

This doesn't change my point. Halliday and Matthiessen's analysis is one analysis of these phenomena. But if we can one minute assume that it is not the case that because they say something in IFG 3, then it must be true, then we can start getting back to doing real linguistics.

As I said in my email, I believe an alternative analysis of this phenomena, where require/allow/permit, etc. are treated as full verbs, and since they have grammatical patternings very distinct to material verbs, to create a new process type for them.

Blogger Comments:

[1] O'Donnell's point has no validity because it proceeds from an inaccurate rendering of the theory in question.

[2] For example, the verbs in these verbal groups are neither 'auxiliary' nor less than "full". In clauses such as 'I was required to leave', 'was required' is the alpha verbal group and 'to leave' is the beta verbal group in a verbal group complex.

[3] The distinct grammatical patterns reflect the fact that these verbs of modulation introduce causation into the clause by means of a verbal group complex, adding the feature of agency to the clause as a whole.

[4] Theory improvement proceeds from an accurate rendering of the theory in question.  The issue is accurate interpretation, not dogmatic belief.