Misconstruing Ellipsis Of Repetition As Relational Hyp(er)onymy

Martin (1992: 299):

Depending on the field the relationship between hyponym and hyperonym has different names.  These are commonly realised through Pre-Classifier structures (for which see Martin 1988): a class of noun, a make of car, a breed of dogs etc.  And they are used cohesively as well: Like my new car? — Yes, what make is it?  A cohesive relationship between one of these items and the hyponym it relates to a hyperonym will be referred to as relational hyponymy/hyperonymy.  Typical realisations include: class, kind, type, form, breed, make, sort, style, species, order, family, variety, genre, grade, brand, caste, category.

Blogger Comments:

[1] In SFL theory, this nominal group function is termed 'extended Numerative: variety', which is the cross-classification of 'collective' and 'type'.  See Halliday & Matthiessen (2004: 333).

a	make	of	car
Numerative: variety			Thing
Premodifier	Head	Postmodifier

[2] The cohesion in the response includes ellipsis of an entire clause, and anaphoric personal reference of it to my new car.  But most pertinent here is the ellipsis of the potential lexical repetition of car.

do	you	like	my new car
Finite	Subject	Predicator	Complement

yes	I	do	like	your new car
mood Adjunct: polarity	Subject	Finite	Predicator	Complement

what make of car	is	it
Subject	Finite	Complement

[3] The notion of a 'relational hyponymy/hyperonymy' therefore arises from not recognising a cohesive relation.

Misconstruing Meronymy As Hyponymy

Martin (1992: 297):

Composition taxonomies organise people, places and things in a given field with respect to part/whole rather than class/subclass relations.  An illustrative taxonomy is outlined in Fig. 5.13, drawing once again on the field of music.  The less delicate systems [Fig. 5.13] focus on the organisation of the swing bands… .  The more delicate systems [Fig. 5.14] break down the players into sections as they were typically organised on stage.

Blogger Comments:

[1] Figure 5.13 starts out as a composition taxonomy — e.g. 'management' and 'personnel' are parts of the whole 'swing band' — but then construes 'female' and 'male' as parts of the whole 'vocalists'; and 'vocalists' and 'players' as parts of the whole 'members'.  That is, it misconstrues hyponyms (types of x) as meronyms (parts of x).

The continuation of the taxonomy in Figure 5.14 is entirely hyponymic — despite being presented as meronymic; e.g. 'rhythm', 'horns' and 'reeds' are misconstrued as parts of the whole 'players'.

[2]  Halliday & Matthiessen (1999: 92):

While the choice of ‘level’ in hyponymic elaboration is the choice in delicacy of categorisation, the choice of level in a meronymic taxonomy is the choice in delicacy of focus.  The focus is typically on the whole (i.e. the most inclusive region within the meronymy) even if a specific part is particularly important.

[3] This misconstrues (what is not) a composition taxonomy as a system network.

Misconstruing Field As Language

Martin (1992: 295):

The point of the network is simply to illustrate superordination from the point of view of a field specific taxonomy which would underlie statements like Cool is a kind of American improvised modern popular music performed in jazz clubs in New York and Los Angeles in the late 50's and early 60's.

Blogger Comments:

[1] This evinces that language (statements) describing social activities are the source of field networks.

[2] This misconstrues context (field) as lying under semantics (statements).

Misrepresenting Superordination (Hyponymy)

Martin (1992: 294-5):

Taxonomic relations are of two main types, depending on the kind of taxonomy they reflect.  Superordination relations reflect taxonomies based on subclassification (the "is a" relation); composition relations reflect taxonomies based on parts to wholes (the "has a" relation). … Some examples are presented in Table 5.4.

Table 5.4. Examples of superordination and composition
	superordination	composition
‘people’	player–first seed	team–player
‘places’	line–service line	court–line
‘things’	official–referee	racquet–strings
‘actions’	hit–volley	–
‘quality’	excellent-agile	–

Blogger Comments:

[1] This is the distinction between delicacy (hyponymy) and composition (meronymy).  The general sense of hyponymy is 'be a kind of' whereas for meronymy it is 'be a part of'.  The logical organisation of the former is the expansion type: elaboration, while the logical organisation of the former is the expansion type: extension.  Both of these expansion types cross-classify with the distinction between identity and class membership (see Halliday & Matthiessen 1999: 144-6).

[2] The claim here is that the person 'first seed' is a subclassification of the person 'player'.  Strictly speaking, 'first seed' identifies one of a group of players.  That is, it identifies a part of a whole.  The relation is therefore not one of superordination/delicacy/hyponymy.

[3] The claim here is that the quality 'agile' is a subclassification of the quality 'excellent'.  However, the two qualities are not even of the same type, since 'agile' is a behavioural propensity, whereas 'excellent' is an evaluation.  On the model of Halliday & Matthiessen (1999: 211), 'agile' is a quality of expansion, and 'excellent' a quality of projection.  The relation is therefore not one of superordination/delicacy/hyponymy.

Confusing Context, Semantics And Lexicogrammar

Martin (1992: 293-4):

The network for lexical relations developed below will be organised around these three different aspects of the contextual structure of activity sequences. … At primary delicacy, the system for lexical relations is presented as Fig. 5.10.

Blogger Comments:

There are two major violations of the stratification hierarchy here.

[1] On the one hand, lexical relations (lexicogrammatical stratum) are presented as being a system of a higher stratum (discourse semantics).

[2] On the other hand, while the system is presented as located on the discourse semantic stratum, it is actually composed of the features that have been argued to be located at the stratum of context: (taxonomic, nuclear and activity).


Summary:

The system proposed for the discourse semantic stratum is concerned with the stratum below (lexicogrammar) and composed of features purported to be of the stratum above (context).

Misrepresenting Congruent Vs Incongruent

Martin (1992: 293):

The message part may or may not be realised by a single lexical item; congruent and incongruent realisations are exemplified in Table 5.3. 

Table 5.3. Simple and complex lexicalisation of message part
MESSAGE PART	CONGRUENT (1 lexical item)	INCONGRUENT (1+ lexical items)
‘person’	champion	tournament winner
‘place’	court	playing area
‘thing’	draw	playing schedule
‘quality’	outstanding	first class
‘action’	smash	hit an overhead
‘quality’	well	out of sight

Blogger Comments:

[1] In SFL theory, the distinction between congruent and incongruent realisations is the distinction between non-metaphorical and metaphorical expressions.  When the semantics and the lexicogrammar are in agreement (congruent), there is no grammatical metaphor; when the semantics and the lexicogrammar are not in agreement (incongruent), there is grammatical metaphor.

Here the distinction is not used in the SFL sense, but instead used merely to distinguish whether the semantic unit is realised by one word or more.

[2] There are two layers of confusion here.  On the one hand, the elements listed for the discourse semantic unit, the message part, are those that have been previously identified as the field variables that the message part realises.  On the other hand, these elements are not field variables, on the SFL model, but semantic elements in a description of social activities.

Misconstruing Stratification: Semantics Realising Semantics

Martin (1992: 293):

For a given field, the message part realises (i) one of the features taxonomising people, places and things, or (ii) one of the actions configuring with people, places and things and entering into activity sequences, or (iii) one of the qualities associated with people, places, things and actions.

Blogger Comments:

The proposal here is to present a model of field-specific semantics.  The method of doing so is to interpret field as semantic descriptions of social activities and to have those semantic descriptions realised by a discourse semantic unit, the message part.  That is, the level of cultural context is misconstrued as a level of language, semantics, and this semantics is then realised by discourse semantics.

This violates the theoretical principle of stratification as distinct levels of symbolic abstraction, not only because it confuses context and language, but also because it distributes the same level of symbolic abstraction, semantics, over two distinct strata.

In SFL theory, the ideational meaning realising a specific field is known as a semantic domain — the semantic profile of a field (Halliday & Matthiessen 1999: 323).

Misconstruing Experiential As Constituent Of Logical

Martin (1992: 293):

On the basis of this characterisation of field, the discourse semantic unit underlying lexical item and entering into cohesive lexical relations can be set up.  Since it is an experientially defined unit, it will be referred to as a message part, to bring out its metafunctional relationship with CONJUNCTION.

Blogger Comments:

[1] The characterisation of field confuses context with language by construing field as a semantic description of social activity.  That is, it violates the principle of stratification.

[2] This confuses the syntagmatic axis (structural unit) and the paradigmatic axis (lexical item).

[3] This places the higher level of symbolic abstraction (discourse semantic) below the lower level (lexical).  That is, it inverts the stratification of content.

[4] This construes the experiential unit as something that is interrelated —cohesively, so: textually — rather than as something with internal structure; cf sequences, figures, elements in SFL theory.

[5] In SFL theory, a 'message' is a unit of the textual metafunction on the semantic stratum.

[6] This misconstrues the relation between the logical and experiential metafunctions as one of constituency.  On the one hand, logical units don't consist of experiential units; on the other hand, the relation between units in complexes is not constituency but interdependency.

Misconstruing First Order Field As Its Semantic Description [2]

Martin (1992: 292-3):

As a first approximation, these activity sequences can be broken down as follows:

i. taxonomies of actions, people, places, things and qualities

ii. configurations of actions with people, places, things and qualities and of people places and things with qualities.

iii. activity sequences of these configurations

This can be illustrated from the field of tennis as follows:

i. TAXONOMY — part/whole relations among game-set-match

ii. CONFIGURATION — Agent Process Medium structure player-serve-ball

iii. ACTIVITY SEQUENCE — player serve-opponent return-player volley

Blogger Comment:

This provides an ideational semantic model for analysing construals of social activities — while purporting to be a model of field.

In terms of SFL theory, the three components of this 'field' model correspond to the three orders of phenomena in ideational semantics: element, figure and sequence (Halliday & Matthiessen 1999).

Misconstruing First Order Field As Its Semantic Description [1]

Martin (1992: 292):

Halliday's notion of field will be developed here by defining fields as sets of activity sequences oriented to some global institutional purpose.  Examples include linguistics, tennis, cooking, wine making, gardening, dog breeding, film, architecture, sewing, car racing, philosophy, sailing, building, chess, war, politics, religion and so on. …

A given institution comprises a large number of different activity sequences, where these are realised linguistically through temporally ordered chains of Process and Medium with their attendant participant and circumstantial roles.  Tennis for example includes, among other activity sequences, playing coaching, commenting, analysing, viewing, training, club meetings, tournament presentations, interviews, grading, and so on.  The participants and processes in these are overlapping, not identical.  At the same time, each of these activities participates in an overall institutional focus — that of making it possible for people to enjoy in diverse ways a game played with strung raquets [sic] and a furry hollow ball.

Blogger Comments:

[1] This defines field only as first order field (what's going on), but exemplifies it largely in terms of second order field (what it's about).  That is, it is inconsistent in terms of orders of experience.

[2] This misconstrues (the linguistic realisation of first order) field as semantic descriptions of 'what's going on'. That is, if the field is 'someone predicting the weather', then the claim is that this is realised linguistically as participant^process^participant. In other words, the theoretical characterisation of the field is treated as text to be analysed semantically. On the SFL model, the field of weather prediction is realised in the language that realises the weather prediction.

[3] In SFL theory, 'chains of Process and Medium with their attendant participant and circumstantial roles' are modelled as sequences of figures, but they are not limited to temporal relations.  They involve the full range of expansion and projection types.  The depiction of 'activity sequences' as only 'temporally ordered' demonstrates that field is here misconstrued as semantic descriptions of temporally ordered social activities.

Confusing Context With Semantics

Martin (1992: 291-2):

Halliday (1985: 12) defines field of discourse as follows:

The FIELD OF DISCOURSE refers to what is happening, to the nature of the social action that is taking place; what is it [sic] that the participants are engaged in, in which the language figures as some essential component.

This is a useful starting point for interpreting the contextual semantics of experiential meaning in functional linguistics because the perspective is a social one, and because of the emphasis placed on language as action; field is not specifically tied to subject matter or topic, notions through which it can be discussed only in certain modes (it makes little sense for example when watching and listening to a rugby league training session to ask what the topic is; the appropriate question is "What is going on?").

Blogger Comments:

[1] The term 'contextual semantics' merges two distinct strata — two levels of symbolic abstraction — in the SFL model: context and semantics.  The former refers to the culture as a semiotic system, while the latter refers to a level within language.  It is a clear indication that the principle of stratification is not understood in the discourse semantic model.

[2] The concept of 'language as action' is distinct from 'field' referring to 'the nature of the social action taking place'. For the former, Halliday & Matthiessen (1999: 354) acknowledge a cline from ‘language in reflection’ to ‘language in action’:

In situations of the ‘language in action’ kind, where the discourse is a relatively minor component of the total activity, the grammar and the semantics are obviously less constructive of the whole than in a ‘reflection’ context …

[3] In SFL theory, field refers to both 'what's going on' and the subject matter.  The former is the first order field; the latter is the second order field (Halliday & Matthiessen 1999: 320-1).

[4] If the first order field ('what's going on') is watching and listening to a rugby league training session, then the second order field ('what the topic is') is the subject matter of what is said while doing so.  To paraphrase Halliday & Matthiessen (1999: 354): while watching and listening to a rugby league training session itself is not constituted of language, the activity of commenting on it is.  The subject matter is indispensable for modelling field.

Confusing Register (Language) With Context (Culture)

Martin (1992: 291):

As an alternative to approaching the problem simply from the perspective of grammar and lexis, one can turn to the level of context, and ask questions about appropriate units based on the structure of the register variable field.

Blogger Comments:

[1] The problem (of the syntagmatic extension of the lexical item) is a problem that only arises from misunderstanding the notion of 'lexical item'.  See previous post.

[2] In SFL theory, field is the ideational system, not structure, of context, not register.  Registers, on the other hand, are functional varieties of language use — text types — that differ in terms of the probabilities of systemic features being selected.

Misconstruing The Difference Between Lexical Item And Grammatical Word

Martin (1992: 290-1):

The distinction within a systemic model between lexical item and word means that in principle cohesion analysis is not tied to orthographic word boundaries.  Phrasal verbs for example can be taken as single lexical items. …

Just how far this notion of a lexical item could be pushed however, remains unclear. …

The problem is that if strong mutual expectancy is used to define lexical items, where does one stop?  If phrasal verbs are treated as single lexical items, then what about Process Range structures …

And if Process Range structures, such as these, what about Process and Medium …

Or Process and Circumstance (of location, with deixis-less destinations) …

The point is that the distinction between word and idiom is a gradient one, and that distinguishing word from lexical item does not determine where the line between the two is drawn.

Blogger Comments:

[1] This misrepresents the distinction between lexical item and (grammatical) word as a question of how many words make up a lexical item.  Halliday & Matthiessen (1999: 568):

The folk notion of the “word” is really a conflation of two different abstractions, one lexical [lexical item] and one grammatical [word rank].

[2] The view of lexical items presented here is from below (graphology) and the syntagmatic axis (word boundaries).  In SFL theory,  lexical items are specified paradigmatically, by combinations of the most delicate features of lexicogrammatical systems.

[3] In SFL theory, strong mutual expectancy is not used to define lexical items.  This misconstrues lexical items as a syntagmatic relation.  See [2].

[4] Here combined elements of grammatical structures are being used to raise doubts on the (misconstrued) notion of the lexical item.  The purpose of doing so is to justify an approach based on a misconstrued notion of 'field'.  See following posts.

[5] The distinction between word and idiom is irrelevant to the distinction between word and lexical item.  Again, idioms, in Martin's words 'frozen collocations', takes a syntagmatic perspective on the issue.

Inconsistencies Of Structure And Metafunction

Martin (1992: 290):

The problem of semantic units has been a recurrent one throughout Chapters 2, 3 and 4.  Stratifying the content pla[n]e makes it possible to distinguish move or message from clause, and participant from nominal group, although in a functional grammar, semantic units will obviously be founded on grammatical ones.  The issue arises once again here, with respect to the nature of the discourse semantic units lexical cohesion is analysed between.

Blogger Comments:

[1]  In SFL theory, units have internal structure: figures have elements of structure such as Senser, Process and Phenomenon.  In discourse semantics, on the other hand, the units participant and message don't have internal structure; instead, they are the elements that are related.  This inconsistency is necessary because, as previously explained, these discourse semantic structures are not actually structures.

[2] The discourse semantic unit move, on the other hand — inconsistent with the other discourse semantic units, message and participant — is presented as a unit with internal structure and one which realises elements of structure at a higher rank, exchange (op cit: 46-59).

[3] In SFL theory, a message is a unit of the textual metafunction at the level of semantics.  In discourse semantics, messages are the units that are said to be logically related to each other.

[4] In SFL theory, a participant is an experiential category.  In discourse semantics, participants are units of the textual metafunction — despite the fact that the textual metafunction is concerned with the organisation of the ideational and interpersonal metafunctions, not with the construal of experience (e.g. as participants).