Classifying Expansion Type On The Basis Of Form

Martin (1992: 317-8):

At clause rank, the TRANSITIVITY roles which have not been considered are Agent and Beneficiary (including Client, Recipient and Receiver).  These roles are intermediate between extensions and enhancements, and are accordingly realised both with and without prepositions in clause structure:

Agent:         +preposition   The ship was stolen by Ford.
                    –preposition   Ford stole the ship. 

Client:         +preposition   He bought the present for Trillian.
                    –preposition   He bought Trillian the present. 

Recipient:   +preposition   He gave the drink to Zaphod.
                    –preposition   He gave Zaphod the drink. 

Receiver:    +preposition   He told the story to Arthur.
                    –preposition   He told Arthur the story.

In line with the proposals developed above, the prepositional realisations of these intermediate roles can be taken as enhancing and the non-prepositional realisations as extending.

Blogger Comments:

 [1] No argument is provided as to why Agent and Beneficiary rôles are "intermediate between extensions and enhancements" — it is merely asserted as a fact.  In SFL theory, the expansion relation between the Nucleus and both the Agent and the Beneficiary is enhancement, not extension (Halliday & Matthiessen 1999: 175, 219), not least because

• the participant Agent (external cause) is agnate with the circumstance Manner: means;
• the participant Client is agnate with the circumstance Cause: behalf;
• the participants Recipient and Receiver are agnate with the circumstance Location: directional.

[2] The reason for the variation in participant realisation — as prepositional phrases or nominal groups — has nothing whatsoever to do with expansion type.  Halliday & Matthiessen (2004: 295-6):

… the choice of ‘plus or minus preposition’ with Agent, Beneficiary and Range … serves a textual function. … The principle is as follows. If a participant other than the Medium is in a place of prominence in the message, it tends to take a preposition (i.e. to be construed as ‘indirect’ participant); otherwise it does not. Prominence in the message means functioning either

(i) as marked Theme (i.e. Theme but not Subject) or

(ii) as ‘late news’ — that is, occurring after some other participant, or circumstance, that already follows the Process.

In other words, prominence comes from occurring either earlier or later than expected in the clause; and it is this that is being reinforced by the presence of the preposition. The preposition has become a signal of special status in the message.

Again, function is here being classified according to form (the view 'from below' of formal linguistics) instead of form being classified according to function (the view 'from above' of functional linguistics). 

Misconstruing Enhancement As Elaboration And Elaboration As Extension

Martin (1992: 317):

This interpretation of the experiential grammar in terms of the general logico-semantic relations of elaboration, extension and enhancement is summarised in Table 5.13.

Table 5.13. Elaboration, extension and enhancement across clauses and groups
Elaboration	Extension	Enhancement
process = range	process + medium + range:entity	process x circumstance
take shot (take a shot)	shoot deer (shoot the deer)	shoot field (shoot in the field)
classifier = thing	epithet + thing	thing x qualifier
parking lot (a parking lot)	new car (a new car)	car park (the car in the park)
event = particle	event + event	event x quality
shoot up (shoot up)	try shoot (try to shoot)	shoot carefully (shoot carefully)

Blogger Comments:

As this table is a summary of the preceding discussion, see previous posts for more detailed arguments.

[1] The relation between the Classifier parking and the Thing lot is enhancement, not elaboration.  It is a lot for the purpose of parking.

[2] To the extent that the preposition up construes direction, the relation between the verb shoot and the preposition within the phrasal verb is enhancement: location, not elaboration.

[3] Expansion relations obtain between the Nucleus (Process/Medium) and other participants and circumstances, not between the Process and Medium.

[4] The relation between the Epithet new and the Thing car is elaboration, not extension.  The Epithet provides further detail about the Thing.

[5] The expansion relation obtains between the Nucleus and the circumstance of Manner: quality, not  within a verbal group between Event and "Quality".

Misconstruing Projection As Enhancement

Martin (1992: 317):

The same prepositional phrases used to enhance Processes can also be used to enhance Things, functioning as Qualifiers in nominal groups:

CIRCUMSTANTIAL QUALIFIERS
the restaurant at the end of the universe
the race through the galaxy
the noise from the engine room
the ship like Ford's
the present for Zaphod
the story about Trillian

 Blogger Comment:

The relation between the Thing and Qualifier is projection: matter, not enhancing expansion.

Reclassifying Function According To Form

Martin (1992: 316-7):

It is sometimes argued that Manner adverbs should be treated as circumstances in light of the fact that they all have alternative prepositional phrase realisations.  Note however that in order to be expressed as a prepositional phrase, these qualities of events must be nominalised, making the circumstantial realisation the marked one:

MANNER ADVERBS : INCONGRUENT PREPOSITIONAL PHRASES ::

Ben ran quickly : Ben ran with considerable speed ::

Ben won easily : Ben won with ease ::

Ben slept fitfully : Ben slept in fits and starts :

Ben winced painfully : Ben winced in great pain

The prepositional realisations will be taken here as enhancing Processes, and the adverbial realisations as enhancing Events.

Blogger Comments:

[1] In SFL theory, priority is given to the view 'from above', the meaning realised by the wording, not to the view 'from below', how the wording is realised at lower ranks.  Here function is being classified according to form, rather than form classified according to function.  That is, the account is formal rather than functional.

[2] The argument here is that a "marked" realisation of a circumstance of Manner: quality (a prepositional phrase) is reason to treat the "unmarked" realisation (an adverbial group) as something different — as an element of the verbal group.  By the same logic, a "marked" realisation of Theme is reason to treat an "unmarked" realisation of Theme as something other than Theme.

See yesterday's post for a theoretical explanation for the different realisations of Manner: as adverbial group or prepositional phrase.

[3] This continues the earlier confusion of markedness and congruence.  See previous clarification  (30/4/15) here.

Relocating A Subset Of Manner Circumstances To The Verbal Group

Martin (1992: 316):

Adverbial realisations of manner are probably better treated as enhancements of events (at verbal group rank) rather than of processes (clause rank).  This brings out the semantic continuity at group rank between modifying nouns and modifying verbs:

EPITHET°THING : EVENT°MANNER ADVERB ::

careful player : play carefully ::

hungry child : eat hungrily ::

fast track : run fast ::

intense pressure: write intensely

etc.

Blogger Comment:

The proposal here is to treat meaning differently on the basis of how it is realised in wording.  That is, it is theorising that gives priority to the view 'from below'.  In SFL theory, priority is given to the view 'from above': the meaning being realised in wording.

As might be expected, it creates a host of theoretical inconsistencies, as well as making the theory less parsimonious.  Two examples:

• while 'manner' realised by adverbial groups — quality, comparison and degree — is treated as realised in the function structure of the verbal group, 'manner' realised by prepositional phrases — means, quality, comparison and degree — is treated as realised in the function structure of the clause, despite the fact that they all 'construe the way in which the process is actualised' (Halliday & Matthiessen 2004: 267);

• the experiential structure of the verbal group is made more complicated, just for this subset of instances, adding an extra element — one that is realised by a rank-shifted adverbial group embedded in the structure of the verbal group.

Halliday & Matthiessen (2004: 267) explain the reason for the different realisations of Manner as follows:

Manner comprises four subcategories: Means, Quality, Comparison, Degree … Means is close to the participant rôle of Agent and Comparison is like a participant in a clause with the same type of process, whereas Quality and Degree are like features of the process itself. These differences in status are reflected in realisational tendencies: Means and Comparison tend to be realised by prepositional phrases, whereas Quality and [Degree] tend to be realised by adverbial groups.

Reducing All Verbal Group Complex Relations To Extension

Martin (1992: 315-6):

With verbal groups, extension combines events.  Halliday (1985: 255-69) categorises a wide variety of verbal group complexes with respect to both expansion and projection.  He reserves the category of extension for conation… 

For purposes of lexical cohesion analysis however, all verbal group complexes will be treated as involving extension here, in order to bring out proportionalities of the following kind:

 ELABORATION : EXTENSION : ENHANCEMENT ::

(phrasal verb : verbal group complex : event x quality ::)

look into : keep looking : look carefully ::

run into : attempt to run : run quickly ::

see through : happen to see : see clearly ::

go over : promise to go : go reluctantly

etc.

Blogger Comments:

[1] Extension — in common with projection, elaboration and enhancement — relates verbal groups logically in a verbal group complex.

[2] The logical grammatical structure of verbal group complexes is not a factor in analysing cohesive (textual nonstructural) relations between lexical items.

[3] To treat all verbal group complexes as involving extension — for whatever reason — is to miscategorise the instances that do not involve extension, and to misrepresent the category 'extension'.  Three of the four examples do not involve extension, and one does not even involve expansion:

• keep looking is elaboration: phase: time: durative
• happen to see is enhancement: modulation: cause: reason
• promise to go is projection: proposal: locution

[4] Any proportionalities that depend on a falsehood are themselves false.  Moreover, no argument is provided for the value of cross-categorising types of expansion with phrasal verbs, verbal group complexes, and clause fragments.

[5] As related by enhancement, these examples are each Process and Manner circumstance within a clause realising a figure.


General Observations:

• The concern here is purported to be discourse semantics, but the focus is merely on (rebranding) the grammar.

• The concern here is purported to be the experiential metafunction, but the focus is on (rebranding) logical relations.

Misconstruing Elaboration As Extension

Martin (1992: 315):

Similarly, whereas Classifiers elaborate the Thing, Epithets add qualities.  These may be of an experiential or attitudinal kind:

Table 5.12. Extension in the nominal group
NominalGroup
epithet	+	thing	epithet	+	thing
			(attitudinal)
red		car	ugly		car
big		lunch	chauvinist		pig
round		ball	pleasant		spot
speeding		bullet	beautiful		shot
hungry		puppy	greedy		kitten

 Blogger Comments:

[1] In every case, the expansion relation between the Epithet and Thing is elaboration, not extension.  This is demonstrated by the agnate relational clauses, all of which are intensive (elaborating), not possessive (extending): the car is red/ugly, the lunch is big, the ball is round etc.

[2] This is a Classifier, not an Epithet, as shown by the fact that, unlike Epithets, it can't be intensified as a very speeding bullet.

Misconstruing Enhancement & Projection As Extension

Martin (1992: 314-5):

Next, extension.  With extension distinct meanings are combined.  The contrast between Process°Range:process and Process°Medium structures in the clause, and Classifier°Thing and Epithet°Thing structures in the nominal group has already served to illustrate the difference between elaboration and extension at issue here.  Where Range:processes elaborate a Process, Range:entities (Halliday 1985: 134-5) and Mediums (Halliday 1985: 144-5) extend them as outlined in Table 5.11.

Table 5.11. Extension in the clause
Clause process	+	range:entity	process	+	medium
climb		mountain	shoot		terrorist
play		piano	hug		friend
cross		court	cook		rice
like		tennis	please		crowd (it pleased them that…)
see		play	strike		me (it struck me that…)
consider		text	convince		audience (it convinced them…)

Blogger Comments:

[1] Extension, like elaboration, enhancement and projection, is a logico-semantic relation.  It includes  the relations of composition (meronymy), possession and association.

[2] As previously demonstrated (here and here), it is not  the case that 'the contrast between Process°Range:process and Process°Medium structures in the clause, and Classifier°Thing and Epithet°Thing structures in the nominal group has already served to illustrate the difference between elaboration and extension at issue here'.  (Process and Medium are not related by extension; in the examples provided, both Classifier and Epithet elaborate the Thing.)

[3] Neither Range:entities nor Mediums extend a Process.  On the one hand, in SFL, expansion and projection relations obtain between the Nucleus (Process/Medium) and the rest.  On the other hand, the relation between the Nucleus and Range:entity depends on the process type.  In material processes, the relation is enhancement, not extension.  In mental and verbal processes, the relation is projection, not extension.  See Halliday & Matthiessen (1999: 175).

Thus, none of the six examples involve extension.  In the first three, the relation is enhancement,  not extension, whereas in the last three, it is projection, not extension.

[4] 'Tennis' and 'play' are processes, not entities.  Martin (1992: 311):

As Halliday (1985: 135) points out, "Tennis is clearly not an entity; there is no such thing as tennis other than the act of playing it."

Misrepresenting Elaboration

Martin (1992: 314):

As far as analysing lexical cohesion with specific fields in mind is concerned, all of the elaborating structures discussed above can be taken as realising single message parts.  This is the main source of incongruence between message part and lexical item as these units are being defined here.

Blogger Comments:

[1] In SFL theory, field specificity is irrelevant for analysing lexical cohesion — as it is for all non-structural resources of the textual metafunction in the grammar.

[2] As demonstrated in recent posts (since 10/10/15), many of the 'elaborating structures' did not involve elaboration, and genuine elaboration relations were misclassified as extension or enhancement.

[3] In SFL theory, 'incongruence' is a technical term that refers to a metaphorical relation between the two levels of content, semantics and lexicogrammar.  Here it is applied to a non-metaphorical (congruent) relation between a proposed experiential unit at the level of (discourse) semantics, and the output of the most delicate systems of the lexicogrammar.

Misconstruing Elaboration As Enhancement

Martin (1992: 314):

These pre-nominal structures contrast with qualifying ones in which facet, measure, isolation and type are coded as distinct participants; the latter can be treated as involving two message parts:

ONE MESSAGE PART : TWO MESSAGE PARTS ::
the top of the ridge : the ridge x on the top ::
a herd of buffalo : buffalo x in the herd ::
the biggest of the melons : the melons x that were the biggest ::
that kind of beer : beer x of that kind ::
the train's passengers : the passengers x on the train

Blogger Comments:

In the case of the defining relative clause that were the biggest and of that kind, the expansion relation between the Thing and the Qualifier is elaboration, not enhancement,.

Misconstruing Extension (Possession) As Elaboration

Martin (1992: 314):

Finally, possessive Deictics can be taken as elaborating where the relationship between the participant realised by the Deictic and that realised through the Thing are in part whole relation rather than an ownership one: the train's passengers or its wheels but not his passengers or its kittens.

Blogger Comment:

Logically, both composition (part/whole) and possession (ownership) are types of extension, not elaboration, in SFL theory.

Misconstruing Extension (Composition) As Elaboration

Martin (1992: 313-4):

With nominal groups, one further set of elaborations to consider are realised through Pre–Deictic, Pre–Numerative, Pre–Epithet and Pre–Classifier structures.  The first three of these code part/whole relations and the last class/subclass ones.  Like Classifier°Thing structures, from the point of view of field these function simply as grammatical resources for isolating particular parts or classes of people, places and things.

Pre–Deictic            the top of = the ridge, the back of = the garden

Pre–Numerative     a herd of = buffalo, a mouthful of = food

Pre–Epithet            the last of = the questions, the biggest of = the melons

Pre–Classifier        that kind of = beer, this sort of = thing

Blogger Comments:

[1] The part/whole relation — composition (meronymy) — is a type of extension, not elaboration.

[2] In SFL theory, these are now termed 'facet' expressions and classified as extended Numeratives.

[3] In SFL theory, these are now termed 'measure' expressions and classified as extended Numeratives.  The first example is an 'aggregate' expression, whereas the second is a 'portion 'expression.

[4] In SFL theory, this is an ordering Numerative or 'ordinative'.

[5] In SFL theory, these are now termed 'variety' expressions and classified as extended Numeratives.

[6] These are nominal group functions, not structures.

[7] Only the first two involve part/whole relations: 'last' is not part of the whole 'questions'; 'biggest' is not part of the whole 'melons'.

[8] The point of view of field is irrelevant to these grammatical functions.

Misconstruing Enhancement (Cause) As Elaboration

Martin (1992: 313):

The same verbs [go and come] are similarly elaborated at group rank.  The verbal group complexes involved specify types of coming and going with respect to leisure activities:

go/come = swimming, skiing, snorkelling, jogging, running, sailing, handgliding [sic], surfing, drinking, etc.

Blogger Comments:

The leisure activities listed do not 'specify types of coming and going'; for example, 'drinking' is not a type of 'coming and going'.  Instead, the leisure activities represent the cause (purpose or result) of coming or going.  The expansion relation is thus enhancement, not elaboration.