"Identifying Things By Comparing Their Order"

Martin & Rose (2007: 167-8):

So things can be identified by comparing the intensity of their qualities, with words like better and best. They can also be identified by comparing their quantity, with words like most, more, fewer, less; so much, so little:

Spiritual murder is more inhumane than a messy, physical murder. 

What's wrong with him?

Could he have changed so much?

And they can be identified by comparing their order:

As an eighteen-year-old, I met a young man in his twenties.
An extremely short marriage to someone else failed all because I married to forget
More than a year ago, I met my first love again through a good friend.

Helena refers to him as her first love to distinguish him from someone else she later married. Other resources for identifying things by their order include first second, third; next, last; preceding, subsequent, former, latter.

Tutu also uses comparison to identify things:

the application should be dealt with in a public hearing
unless such a hearing was likely to lead to a miscarriage of justice

Here such a refers to a particular class of hearing (a public one), and no other.

Blogger Comments:

[1] To be clear, the function of comparative reference is not to identify things by comparing their qualities or quantities, but to create cohesion in the text by presuming information from the text itself. Halliday & Matthiessen (2014: 632-3):

Comparative reference items function in nominal and adverbial groups; and the comparison is made with reference either to general features of identity, similarity and difference or to particular features of quality and quantity.

[2] As previously noted, words like best and most are superlatives, not comparatives, and do not make comparative reference.

[3] To be clear, this is an instance of structural cataphora — since the presumed information is provided within the same nominal group — and so, not an instance of cohesive reference.

[4] To be clear, this is not an instance of comparative reference, since what's wrong with him provides no frame of reference by which so much makes a comparison.

[5] To be clear, Numeratives do not serve as comparative reference items, because they do not make a comparison in general terms of identity, similarity or difference, or particular features of quantity or quality; see Halliday & Matthiessen (2014: 632).

[6] To be clear, merely distinguishing one person from another does not, in itself, constitute an instance of comparative reference. For example, Helena could have distinguished her 'loves' using Classifiers, such as  'English' vs 'Boer', neither of which is a comparative reference item.

[7] To be clear, the function of comparative reference is not to identify things, but to create cohesion in the text by presuming information from the text itself.

[8] To be clear, the comparative reference item such functions cohesively by presuming a standard of reference in the preceding text: public.

Comparative Reference

Martin & Rose (2007: 167):

Now that we’ve looked at how various things are identified (concrete objects, abstractions, institutions, and things that people say) we need to look at ways that these things can be compared. Helena uses several comparative references in some parts of her story:

I finally understand what the struggle was really about.
I would have done the same had I been denied everything.
If my life, that of my children and my parents was strangled with legislation.
If I had to watch how white people became dissatisfied with the best
and still wanted better
and got it.

She begins by presenting the struggle as though we all know what she means, and then refers to it as the same. Later she presents what white people already had as the best, and identifies what they still wanted as better. She doesn’t need to say what is better, because words like better and best are resources for comparative reference, like same, other, else.

Blogger Comments:

[1] To be clear, the function of reference — personal and demonstrative co-reference — is not simply to identify "things", since this is an experiential relation that can be achieved in many ways, most obviously through identifying process clauses. What makes reference a textual resource is its function of creating cohesion in the text by presuming information from the text itself.

[2] Likewise, the function of comparative reference is not simply to compare "things" in the construal of experiential meaning, but to create cohesion in the text by presuming information from the text itself. Halliday & Matthiessen (2014: 632-3):

Whereas personals and demonstratives, when used anaphorically, set up a relation of co-reference, whereby the same entity is referred to over again, comparatives set up a relation of contrast. In comparative reference, the reference item still signals ‘you know which’; not because the same entity is being referred to over again but rather because there is a frame of reference – something by reference to which what I am now talking about is the same or different, like or unlike, equal or unequal, more or less. Comparative reference items function in nominal and adverbial groups; and the comparison is made with reference either to general features of identity, similarity and difference or to particular features of quality and quantity.

[3] To be clear, words like best are superlatives, not comparatives, and do not make comparative reference; see [2] above.  (The the of the best is homophoric (self-referencing), which is a type of exophoric reference, and so: not textually cohesive.)

[4] To be clear, in SFL Theory, the reference function of the in the struggle — if not anaphoric to earlier in the text — is homophoric, and so: not textually cohesive.

[5] To be clear, the same does not refer to the struggle. This is to treat comparative reference as if it were co-reference; see [2] above.

[6] To be clear, the comparison made by the comparative reference item better is with the best — that is, the author says that white people wanted, and got, better than the best.  Here, the best is the frame of reference by which the reference item better makes a comparison.

Mistaking A Verbal Process For A Reference Relation

Martin & Rose (2007: 166-7):

Another important aspect of the Act’s specialised tracking resources is its elaborate naming system:

• sections 1, 2, 3…49
• sub-section (1), (2), (3)
• paragraphs (a), (b), (c)
• sub-paragraphs (i), (ii), (iii)
• sub-sub-paragraphs (aa), (bb), (cc).

This allows the authors to refer exactly to certain paragraphs later (or earlier) in the document. So for example, section 3, subsection (3), paragraph (d) of Chapter 2 refers to sections 5(d) and 28(4)(a):

3 (3) (d) the investigating unit referred to in section 5(d) shall perform the investigations contemplated in section 28(4)(a)

Blogger Comments:

[1] To be clear, the system of IDENTIFICATION (Martin 1992: 93ff) is presented as 'reference as semantic choice', and in this chapter, IDENTIFICATION is said to be 'concerned with tracking participants (p155ff). In these examples, no references are made, since there are no reference items that presume identities recoverable elsewhere in the text, and no participants are tracked, since there are no processes to be participated in.

[2] To be clear, 'naming' is reference in the sense of ideational denotation: the assigning of wordings to meanings.

[3] To be clear, here Martin & Rose confuse a verbal Process (referred) in the experiential structure of an embedded clause with the textually cohesive relation between reference item and referent.

Spatial And Temporal Reference

Martin & Rose (2007: 166):

Another example of specialised reference is the tracking device therewith, which refers to a specific ‘location' in the text. This is used to keep things open, to refer generally to the processes that have to be undertaken to establish the Commission and Committees and empower them:

and for the said purposes to provide for the establishment of a Truth and Reconciliation Commission
and to confer certain powers on, assign certain functions to and impose certain duties upon that Commission and those Committees
and to provide for matters connected therewith.

Reference to location in space (here, there) and time (now, then) is also found in non-specialised discourse. It is used by Tutu to refer to restorative justice (here):

I contend that there is another kind of justice, restorative justice, which is characteristic of traditional African jurisprudence.
Here the central concern is not retribution or punishment but, in the spirit of ubuntu

As with therewith in the Act, identifying by location in space or time is a little more general than using a demonstrative. It treats discourse as a region of meaning that we can be oriented to, as opposed to a collection of people and things we pick out and name.

Blogger Comments:

[1] To be clear, in this instance therewith serves the same function as 'with that', and as such, makes anaphoric specific demonstrative reference to a Truth and Reconciliation Commission. The authors' mistaken notion that therewith refers to a location derives from giving priority to form (there-) over function in their analysis — the opposite of the SFL approach.

[2] To be clear, the temporal demonstratives now and then function 'conjunctively rather than referentially' (Halliday & Matthiessen 2014: 632).

[3] To be clear, in this instance, here also serves as a conjunctive Adjunct of matter. Halliday & Matthiessen (2014: 620):

Here cohesion is established by reference to the ‘matter’ that has gone before. As noted earlier, many expressions of matter are spatial metaphors, involving words like point, ground, field; and these become conjunctive when coupled with reference items. …

||| Without chlorine in the antarctic stratosphere, || there would be no ozone hole. ||| (Here “hole” refers to a substantial reduction below the naturally occurring concentration of ozone over Antarctica.) |||

[4] To be clear, here and there are demonstratives — locative demonstratives — and the reference they make is demonstrative reference.

[5] This attempted hedge is invalidated by the exemplifying texts, wherein:

• there(with) specifically refers to a Truth and Reconciliation Commission, and
• here specifically refers to (another kind of justice,) restorative justice.

It also flatly contradicts the authors' opening claim — [1] above — that therewith refers to a specific ‘location' in the text.

The Authors' Claim That '(Afore)said' Is A Specialised Version Of 'The'

Martin & Rose (2007: 165-6):

In legal and administrative discourse quite a lot of pressure is put on identification resources in order to be precise. This includes some specialised features which we can see in the Act which established the Truth and Reconciliation Commission. The word said for example is used alongside the to refer precisely to what has just been said, specifying dates:

... the nature, causes and extent of gross violations of human rights committed during the period from I March 1960 to the cut-off date contemplated in the Constitution
... acts associated with a political objective committed in the course of the conflicts of the past during the said period

And specifying purposes:

To provide for the investigation and the establishment of as complete a picture as possible of the nature, causes and extent of gross violations of human rights committed and for the said purposes to provide for the establishment of a Truth and Reconciliation Commission, a Committee on Human Rights Violations, a Committee on Amnesty and a Committee on Reparation and Rehabilitation;

The words said or aforesaid are specialised versions of the, specifying that the identity presumed can be found in the preceding text.

Blogger Comments:

[1] To be clear, identification (reference) does not increase the precision of the meaning being made, because it involves the use of reference items whose identity may or may not be resolved by the reader.

[2] To be clear, the words said and aforesaid are not specialised versions of the, most obviously because they occur with the rather than in place of it. More importantly, what they have in common is that all serve as Deictic elements in the nominal group. Halliday & Matthiessen (2014: 373, 374):

In addition to the Deictic element we have just discussed, there may be a second Deictic element in the nominal group, one which adds further to the identification of the subset in question. We will refer to these as post-Deictic or Deictic₂. The post-Deictic identifies a subset of the class of ‘thing’ by referring to its fame or familiarity, its status in the text, or its similarity/dissimilarity to some other designated subset.

As Deictic elements, the and said specify the Thing of their nominal groups — period and purposes. However, only the determiner the serves as a reference item that specifies a recoverable identity. This can be demonstrated by removing the determiner from the nominal groups in question:

• said period
• said purposes.

Here it is clear that said does not present a recoverable identity (cf the period — what period?). Once again, Martin & Rose have confused deixis with reference.

The Contraction And Expansion Of Meaning

Martin & Rose (2007: 165):

In abstract discourse such as Tutu’s argument, this kind of reference to what was just said is very common to refer to a point that’s just been made, possibly to evaluate it. What was said previously is typically tracked with demonstratives (this, that):

For some it has been so traumatic that marriages have broken up.
That is quite a price to pay. 

Amnesty is not given to innocent people or to those who claim to be innocent. It was on precisely this point that amnesty was refused to the police officers 

Once amnesty is granted,
and this has to happen immediately

The advantage of this kind of tracking is that stretches of meaning can be packaged up to play a new role as the argument unfolds. In the following passage, for example, Tutu packages up the effect of amnesty in order to expand on its consequences for civil damages (this means that...). These consequences are in turn packaged up to be evaluated (that is...) and identified (it is...):

The effect of amnesty is as if the offence had never happened, since the perpetrator's court record relating to that offence becomes a tabula rasa, a blank page.
This means... that the victim loses the right to sue for civil damages in compensation from the perpetrator.
That is indeed a high price to ask the victims to pay,
but it is the price those who negotiated our relatively peaceful transition from repression to democracy believed the nation had to ask of victims.

This kind of tracking of what was said is called text reference. As we’ve seen it is used to turn big meanings into smaller, more manageable ones, so that we can then make some more meanings with them. Meanings contract, in other words, so that new meanings can expand. The text is breathing, as the argument moves along.

Blogger Comments:

[1] To be clear, in terms of SFL Theory, each of these is a genuine instance of cohesive anaphoric reference. However, in terms of the authors' model, each is inconsistent with the notion of participant identification, since each of the referents is either a figure or a sequence, not a participant.

[2] To be clear, the unacknowledged intellectual source of this theorising is Halliday & Hasan (1976: 52), who, unlike Martin & Rose, make a distinction between text reference and extended reference.

[3] To be clear, it is not the meanings that become "smaller", but their realisation in wording, as when the meaning of a referent (marriages have broken up) is realised in wording by a reference item (that) which then serves as an element in another clause (that is a high price to pay).

[4] The text is farting, as the argument moves along.

Identifying What People Say: Text Reference

Martin & Rose (2007: 164-5):

Beyond abstractions, it’s possible to track things people say. Helena refers to her prayers for example as all my questions and heartache:

‘God, what's happening? What's wrong with him? Could he have changed so much? Is he going mad? I can't handle the man anymore! But, I can't get out. He's going to haunt me for the rest of my life if I leave him. Why, God?' Today I know the answer to all my questions and heartache.

And Tutu refers to the question he’s just asked as this:

So is amnesty being given at the cost of justice being done?
This is not a frivolous question, but a very serious issue

Blogger Comments:

[1] To be clear, even in the authors' own terms (p155), the system of identification is concerned with tracking participants. For 'things people say' to qualify as participants, they must be construed as participating in a Process, most congruently as the Verbiage of a verbal Process.

[2] To be clear, the only reference item in the nominal group Qualifier to all my questions and heartache is the determiner my, which refers (non-cohesively) to the speaker, not to her prayers. Moreover, in terms of the authors' model, Helena's prayers are not a participant, but an extended quote of several figures:

‘God, what's happening? What's wrong with him? Could he have changed so much? Is he going mad? I can't handle the man anymore! But, I can't get out. He's going to haunt me for the rest of my life if I leave him. Why, God?'

And trivially, heartache is not a 'prayer'.

[3] To be clear, in terms of SFL Theory, this is a genuine instance of cohesive anaphoric demonstrative reference. However, in terms of the authors' model, it is inconsistent with the notion of  participant identification, since So is amnesty being given at the cost of justice being done? is a figure, not a participant.

Comparing Abstractions

Martin & Rose (2007: ):

Comparison can also be used to distinguish types of abstractions, for example kinds of justice:

Further, retributive justice — in which an impersonal state hands down punishment with little consideration for victims and hardly any for the perpetrator — is not the only form of justice. 

I contend that there is another kind of justice, restorative justice, which is characteristic of traditional African jurisprudence.

Blogger Comments:

[1] To be clear, this is reference in the sense of ideational denotation: a nominal group (form) realising a clause participant (function). The nominal group retributive justice does not include a reference item that presumes a recoverable identity.

[2] To be clear, this is a genuine instance of comparative reference. Halliday & Matthiessen (2014: 633):

In comparative reference, the reference item still signals ‘you know which’; not because the same entity is being referred to over again but rather because there is a frame of reference – something by reference to which what I am now talking about is the same or different, like or unlike, equal or unequal, more or less. Comparative reference items function in nominal and adverbial groups; and the comparison is made with reference either to general features of identity, similarity and difference or to particular features of quality and quantity.

Presenting Less Concrete Things

Martin & Rose (2007: 164):

Less concrete things such as agencies (special forces) and abstractions (price, marriage, amnesty) are identified similarly to objects:

We're moving to a special unit
After about three years with the special forces, our hell began. 

Our freedom has been bought at a very great price.
But to compute that price properly 

An extremely short marriage to someone else failed all because I married to forget
After my unsuccessful marriage, I met another policeman. 

It was on precisely this point that amnesty was refused to the police officers who applied for it for their part in the death of Steve Biko.

Blogger Comments:

[1] To be clear, all four non-specific nominal groups "reference" in the sense of ideational denotation; that is, they realise participants. None of the four include reference items whose identity is to be recovered elsewhere. The function of the three non-specific determiners is deictic, not referential.

[2] To be clear, the nominal group here is actually an extremely short marriage to someone else with to someone else serving as Qualifier.

Presenting "Plural Mass" Participants

Martin & Rose (2007: 163-4):

There are a few ways of introducing plural participants (things or people). One way is to use the plural with no determiner:

In the upper abdomen were twenty-five wounds.
These wounds indicated that different weapons were used to stab him…

For presenting participants, there is the plural of ‘a’, namely ‘some’:

they had some friends over
he had some milk for Helena

English uses the indefinite plural ‘some’ with things that can be counted like friends, and things that can’t like milk. Things like milk are ‘masses’. We can package them and then count the packages (two bottles of milk), but we can’t count a ‘mass’ (*two milks).

However with plural things or with masses we also have the option of presenting participants without ‘a’ or ‘the’:

I put garden shears through his neck
They were shot and massacred with AK47s
they poured acid on his face.

Plural things are presented with the ending ‘-s’, while plural masses are presented with neither an ending nor a determiner.

Blogger Comments:

[1] To be clear, this is reference in the sense of ideational denotation only: a nominal group realising a participant. The nominal group twenty-five wounds does not include a reference item whose identity requires recovering elsewhere.

[2] Again, this is reference in the sense of ideational denotation only: a nominal group realising a participant. The non-specific determiner some does not serve as a reference item, because it does not specify a recoverable identity.

[3] To be clear, this concerns the deictic function of non-specific determiners, which do not serve as reference items. Cf Halliday & Matthiessen (2014: 368):

[4] Again, this is reference in the sense of ideational denotation only: a nominal group realising a participant.

[5] To be clear, here Martin & Rose inform their readers, who are necessarily literate in English, the most common way to mark plural number in English nouns. (Cf. also mice, geese, men, children, loaves, appendices etc.)

[6] Here, having informed the reader that "we can’t count a ‘mass’ (*two milks)", Martin & Rose nevertheless count them ("plural masses"). To be clear, mass nouns, like plural nouns, can be "presented" with a determiner, as the authors' own example some milk demonstrates. Again, this is deixis, not reference. Cf Halliday & Matthiessen (2014: 369):

Presenting Concrete Objects

Martin & Rose (2007: 163):

In Chapter 3 we looked at different kinds of entities that can participate in figures, including people, objects, institutions and abstractions, as well as figures that function like things. Each of these different kinds of entities can be identified in different ways. 

Concrete objects, that we can touch, taste, hear, see or feel, are identified pretty much like people. They are introduced indefinitely, and then tracked with determiners like the or pronouns like it:

We used a yellow portable Robin generator to send electric shocks through his body when we put the generator on his body was shocked stiff…

they started to take a plastic bag then one person held both my hands down and the other person put it on my head. Then they sealed it so that I wouldn't be able to breathe and kept it on for at least two minutes…

Blogger Comments:

[1] To be clear, the notion of 'figure' derives from the ideational semantics of Halliday & Matthiessen (1999) — not from Martin's (1992) discourse semantics — wherein the "different kinds of entities that can participate in figures" are elements, a different order of phenomenon from figures. "Figures that function like things" are instances of grammatical metaphor where what would congruently be realised as a clause is instead realised as a nominal group (the congruent realisation of an element).

[2] This claim, unsupported by evidence, is falsified by every single text, during the course of which, any inanimate object is initially realised by a specific nominal group — i.e. one featuring the determiners the, this, that, its, or the pronoun it.

Moreover, the realisation of concrete objects as non-specific nominal groups is reference in the sense of ideational denotation, not reference in the textual sense, since non-specific nominal groups do not include a reference item whose identity is recoverable elsewhere, since non-specific determiners do not specify.

The Apparent Anomaly Of Using Presuming Reference For First Mentions Of People

Martin & Rose (2007: 162-3):

Another apparent anomaly is the use of what looks like presuming reference the first time a character is mentioned. Some examples of this include the use of ‘the’, and a proper name, for first mentions:

He was popular with all the 'Boer' Afrikaners.
I can understand if Mr F. W. de Klerk says

What’s going on here is that Helena assumes that her readers will know who she’s talking about. The identity of the Boer Afrikaners will obviously be known to a South African audience, and the same for the name of a former prime minister. The point here is that speakers/writers make assumptions on the go about what listeners/readers can and can’t be expected to know. If someone's identity is as good as given then presuming reference is used, even where that character hasn’t been mentioned before.

Blogger Comments:

[1] To be clear, in SFL Theory, this is known as the homophoric reference (Halliday & Hasan 1976: 71), which is exophoric reference that is not dependent on the specific situation of the text. Like all exophoric reference, it is not textually cohesive.

[2] To be clear, proper names reference people in the sense of ideational denotation only, and lack reference items whose identities are resolved through textual relations with referents.

[3] To be clear, the so-called "anomaly" here is the use of a specific nominal group (presuming reference) for the first mention of a person — anomalous because it contradicts the authors' model that non-specific nominal groups (presenting reference) are deployed for the first mention of a person.

However, there are fundamental problems with both the authors' model and their explanation of the apparent anomaly. On the first point, non-specific nominal groups — such as a carpenter and someone — do not include reference items that present recoverable identities; non-specific nominal groups only 'reference' in the sense of ideational denotation (realising participants).

On the second point, exophoric the is not the only resource of presuming reference; Martin & Rose additionally list the following items: this, that, I, me, you, he, she, it, we, us, they, them. That is, every single first mention of the speaker as I is also an anomaly, on the authors' model, since it deploys presuming reference instead of presenting reference.

Instances That Neither Present Nor Presume Participant Identities

Martin & Rose (2007: 162):

Usually, presenting reference is used when we first mention a person, and presuming reference is used for second or subsequent mentions. But in English this doesn’t always hold. For example, what looks like presenting reference is used to describe Helena’s first love, even after we know who he is:

he was an Englishman

Similarly her second love is described indefinitely after he is introduced:

(He was) Not quite my first love, but (he was) an exceptional person.

Helena even describes herself as a farm girl, after presuming her own identity twice with my:

My story begins in my late teenage years as a farm girl …

The reason for these apparent anomalies is that these indefinite expressions are being used to describe or classify people, not to identify them. These kinds of expressions are discussed in more detail in Chapter 4. They include classifying figures:

He was an Englishman
It was only a means to the truth

And classifying roles:

She lived as a farm girl.
He worked as a policeman.

Since they don’t actually identify people, we will set these expressions aside here as far as identification is concerned.

Blogger Comments:

[1] To be clear, from the perspective of SFL Theory, presenting reference is non-specific deixis misunderstood as reference. The claim here is that the first mention of a person in a text is usually as a non-specific nominal group — that is, one featuring any the determiners a, an, one, some(one), any(one). This absurd claim is falsified, in the authors' own terms, by every text in which the first mention of the speaker is I and the first mention of anyone else is by their name, since both are presuming rather than presenting.

[2] To be clear, the non-specific determiner an does not reference a recoverable identity. The reference item in this instance is he, which refers anaphorically to a young man in his twenties, who is ascribed the Attribute an Englishman by means of the attributive clause in which he serves as Carrier.

[3] Again, the non-specific determiner an does not reference a recoverable identity. The "indefinite description" is the non-specific nominal group an exceptional person that realises the participant; that is, this is reference in the sense of ideational denotation, not textual meaning.

[4] To be clear, the non-specific determiner a does not reference a recoverable identity. The "description of her" as a farm girl construes the speaker in terms of Rôle: guise; that is, this is reference in the sense of ideational denotation, not textual meaning.

[5] To be clear, "the reason for these apparent anomalies" confuses the purported textual function of non-specific nominal groups (reference) with their experiential function (Attribute, Rôle) in clause structure. In this sense, the confusion is one of metafunction.

However, this metafunctional confusion is further complicated by the fact that non-specific nominal groups do not reference in the textual sense, since they do not include a reference item that presumes a recoverable identity. Non-specific nominal groups only reference in the sense of ideational denotation, that is: in the sense of realising a participant.

That is to say, from the perspective of SFL Theory, the confusion here is between

• the experiential function of a non-specific nominal group in realising a clause participant, and
• the experiential function of a non-specific nominal group in a structural configuration.

Introducing A Few Basic Terms

Martin & Rose (2007: 162):

We can also now introduce a few basic terms for the words that English uses for identifying people and things. Of course words like I, she, it, my, his are pronouns. Words like a and the are known as determiners, since they ‘determine’ whether we can assume an identity or not; ‘a’ is an indefinite determiner, while ‘the’ is a definite determiner. Words like this, that, these, those are known as demonstratives, since they ‘demonstrate’ where to find an identity, ‘near’ with this or ‘far’ with that.

Blogger Comments:

Halliday & Matthiessen (2014: 75) list the word classes that are assumed by SFL Theory:

see also Halliday & Matthiessen (2014: 365-7).

[1] In SFL Theory, only I, she, it are pronouns; my and his are determiners.

[2] In SFL Theory, a is a non-specific determiner, and serves no referential function, whereas the is a specific determiner of the sub-class demonstrative.

[3] In SFL Theory, this, that, these, those are specific determiners of the sub-class demonstrative.

[4] To be clear, the proximal/distal distinction of deixis marks proximity to speaker, not "where to find an identity" in terms of reference, as demonstrated by the demonstrative reference in:

now tell me, this guy you were talking about…
now tell me, that guy you were talking about… .

The Resources For Identifying People

Martin & Rose (2007: 161):

Let’s sum up the resources we’ve seen for identifying people, and add a few more in Table 5.3. 

Blogger Comments:

To be clear, from the perspective of SFL Theory:

1. 'Presenting' reference is not reference, but non-specific deixis, which, by definition, does not present a recoverable identity. The identities said to be presented are the participants realised by the nominal groups in which the non-specific determiner functions. That is, this is reference in the sense of ideational denotation.
2. 'Presuming' reference is a mixture of demonstrative reference (the, this, that), cohesive personal reference (she, he, it, they, them), non-cohesive personal reference (I, me, you, we, us), and reference in the sense of ideational denotation (Helena).
3. 'Possessive' reference is a mixture of cohesive personal reference (his), non-cohesive personal reference (my), and reference in the sense of ideational denotation (Helena's).

Comparative Reference and Possessive Reference

Martin & Rose (2007: 161): 

However, comparative reference and possessive reference are a little different, because they can be used in nominal groups which both present and presume. So another policeman and someone else both present a new person, at the same time as they presume the person they are compared with. The ‘an’ part of another presents a new person, but the ‘other’ part compares him with someone we already know. Likewise someone presents a new person, but else compares him with someone we already know. With possessive reference, my first love presumes someone we already know. However, all my girlfriends presents new people, even though my refers to someone we already know: the narrator, Helena.

Blogger Comments:

[1] To be clear, in comparative and personal ("possessive") reference, it is the reference item — not the nominal group — that presumes a recoverable identity. Moreover, in comparative reference, the domain in which reference items occur also includes the adverbial group — see Halliday & Matthiessen (2014: 626) — which is unaccounted for in this rebranding of Halliday & Hasan's (1976) grammatical reference as Martin's discourse semantic identification.

[2] To be clear, this misconstrues the deictic function of determiners (an, some) in nominal groups as a referential function. Moreover, since these are non-specific determiners, they do not reference a specific identity to be recovered.

[3] To be clear, the people said to be presumed and presented here are the participants realised by the nominal groups my first love and all my girlfriends, respectively. That is, this is not textual reference, but reference in the sense of ideational denotation. This contrasts with the textual reference of my, which, as acknowledged, refers to the speaker, Helena, and as such, does not function cohesively (Halliday & Matthiessen 2014: 628).

Basic Resources For Introducing And Tracking People

Martin & Rose (2007: 160-1):

The participant identification resources we’ve been looking at so far are summed up in Table 5.2, with ways of introducing participants on the left and ways of tracking them on the right. 

So on the left we have resources that introduce us to people; and on the right we have resources which tell us who we already know. Technically, we can say that resources that introduce people are presenting reference, and those that track people are presuming reference. Words like a, an and someone are used for presenting reference. Words like the, that, he, we and names like Helena are used for presuming reference.

Blogger Comments:

[1] To be clear, the presenting reference resources that introduce people (a, an, some) do not function as reference items because, as non-specific determiners, they do not present a recoverable identity. As previously explained, the authors' misunderstanding arises from confusing the deictic function of determiners in nominal groups with the referential function of determiners in cohesion. (The "introduced" person is necessarily the Thing of the nominal group that the determiner, as Deictic, sub-classifies as non-specific.)

[2] To be clear, the presuming reference resources that track people confuse reference in the sense of ideational denotation (Helena) with non-cohesive textual reference (my) and cohesive textual reference (his, he, that, the).

As previously explained, the speaker does not need to keep track of participants because she already knows who she is talking about. This is especially so in the case of my — which Martin & Rose strategically omit from the final sentence — since this refers to the speaker herself.

However, the authors clearly think that the identity at stake is Helena's first love, rather than Helena herself, which demonstrates their confusion of deixis with reference, since the relation between my and first love is deictic, not referential.

Identifying Participants Using Possessive "Pronouns"

Martin & Rose (2007: 160):

Another important resource for identifying participants is possessive pronouns. These pronouns (my, your, her, his, its, our, their) work like a, some, the, this, that, these, those, to tell us which participant we are talking about. In her story, Helena introduces her girlfriends, their police friends and the Africans' leaders in this way:

all my girlfriends
and three of our friends
their leaders

As well as people, the possessions and parts of people can all be presented and presumed with this resource, for example, his throat, my head. 

There are actually two identities in these expressions; one is realised by the possessive pronoun (e.g. my) and the other by the thing that is ‘possessed’ (e.g. girlfriends). The possessive pronoun always presumes an identity, but the thing that is ‘possessed’ may or may not have been previously mentioned.

Blogger Comments:

[1] To be clear, in SFL Theory, these are known as 'possessive determiners' (Halliday & Matthiessen 2014: 382, 623, 627).

[2] To be clear, only 3rd person forms function cohesively as reference items (Halliday & Matthiessen 2014: 628). For the cohesive component of 'exclusive' we/our/ours, see Halliday & Hasan (1976: 49-50).

[3] To be clear, non-specific determiners, like a and some, do not presume a recoverable identity, because they are non-specific. Their inclusion in Martin's system of identification — 'the semantics of reference' — derives from his confusion of nominal group deixis with cohesive reference.

[4] To be clear, the identities presumed by my, our and their are not the participants introduced. This discrepancy derives from confusing nominal group deixis, where a determiner sub-classifies the Thing of a nominal group, with cohesive reference, where a determiner signals a recoverable identity elsewhere in the text.

[5] To be clear, the relation between people and their parts is one of lexical cohesion, meronymy, not reference.

[6] To be clear, the two "identities" in such nominal groups are (a) that which is referred to by the possessive determiner and (b) that which the Thing realises. Only the determiner refers in the sense of textual reference; the Thing of a nominal group refers in the sense of ideational denotation. The confusion here is one of metafunction.

"Comparing People"

Martin & Rose (2007: ):

We have seen that participants can be referred to as different from others, with another or someone else. These kinds of resources compare one participant with another, and so are known as comparative reference. Comparative reference may involve simple contrast, or numbers such as first, second and superlatives such as best, better:

my first love
someone else
another policeman

In English, unlike many languages, we tend to insist on signalling whether we are presenting or presuming every time a participant is mentioned. However comparison is optional; we just use it when we need to.

Blogger Comments:

[1] To be clear, the unacknowledged sources of comparative reference are Halliday & Hasan (1976: 76-87) and Halliday (1985: 294-5). Moreover, the function of comparative reference is not merely to compare one participant with another, but to create cohesion through a relation of contrast. Halliday (1985: 294):

[2] To be clear, ordinatives and superlatives do not serve as comparative reference items, and Martin & Rose provide no argument in support of their claim that they do. Moreover, reference items are not restricted to the domain of the nominal group. Halliday & Matthiessen (2014: 633) classify comparative reference items as follows:

[3] Trivially, better is not a superlative, but non-trivially, as a comparative adverb, it does serve as a comparative reference item.

[4] To be clear, the authors' 'presenting reference' is concerned with their notion of 'introducing participants', and their 'presuming reference' with their notion of 'tracking participants'. However, the distinction here is not one of reference, but one of deixis: non-specific ('introducing/presenting') vs specific ('tracking/presuming') — and it is this distinction that "we tend insist on signalling" in English.

This is shown by the fact that non-specific determiners do not refer, and by the fact that Martin & Rose are concerned with the identity realised by the nominal group in which the determiner functions, rather than with information elsewhere in the text that would resolve the identity referred to by a genuine reference item.

[5] To be clear, the 'optionality' of comparative reference, in this context, lies in the fact that it is not a feature of deixis.  Presumably, when 'we need to use it', it ceases to be 'optional'.

Confusing Text Comprehension With Text Production

Martin & Rose (2007: 159-60):

Helena didn’t name her first love, although she did make up a name for herself, as introduced by Tutu:

a woman calling herself Helena

The name gives us a useful way of referring to Helena, although in her story of course she relies on pronouns (I, my; we, our).

Another tracking resource is ‘the’, which Helena uses later to refer to her second love:

I can't handle the man anymore!

Blogger Comments:

[1] To be clear, a name is an instance of ideational denotation, not textual reference.

[2] Here again Martin & Rose confuse the meaning potential of a speaker/writer — the language that constitutes the data for linguistic theorising — with the comprehension strategies of a listener/reader/analyst. In the text, it is the speaker/writer who refers, not the listener/reader/analyst.

[3] To be clear, in SFL Theory, personal reference items are restricted to the 3rd person (non-interactants). Halliday & Matthiessen (2014: 628):

This is because the SFL model of reference is a system of textual cohesion, not ideational denotation. Halliday & Hasan (1976: 51):

[4] To be clear, the speaker, Helena, does not need to keep track of the participants in her story, since she demonstrably knows who she is talking about. This is the task of her audience, and so does not constitute a resource of the speaker, which is what linguistic theory models.

Misrepresenting Text Comprehension As Text Production

Martin & Rose (2007: 158-9):

The minor players aren’t mentioned again in the story; but Helena’s first and second loves are. Her first love is tracked as follows:

As an eighteen-year-old, I met a young man in his twenties.
He was working in a top security structure.
It was the beginning of a beautiful relationship.
We even spoke about marriage,
A bubbly, vivacious man who beamed out wild energy.
Sharply intelligent.
Even if he was an Englishman
he was popular with all the 'Boer' Afrikaners.
And all my girlfriends envied me.
Then one day he said he was going on a 'trip'.
'We won't see each other again... maybe never ever again.'
I was torn to pieces.
So was he.
An extremely short marriage to someone else failed all because 1 married to forget.
More than a year ago, I met my first love again through a good friend.
I was to learn for the first time that he had been operating overseas
and that he was going to ask for amnesty.
I can't explain the pain and bitterness in me
when I saw what was left of that beautiful, big, strong person.
He had only one desire - that the truth must come out.

Once he is introduced as a young man, the main strategy for tracking his identity is with pronouns, which refer to him ten times on his own (he and his), and twice together with Helena (we). He’s also identified twice as a kind of person: my first love, and that beautiful big, strong person.

Blogger Comments:

[1] To be clear, it is the listener/reader — or discourse analyst — that has to keep track of an identity through a text, not the speaker/writer, since the speaker/writer already knows who they are talking about. That is, the notion of tracking participants misrepresents a listener/reader strategy for the meaning potential of the speaker/writer.

[2] To be clear, in terms of SFL Theory, this confuses two distinct types of cohesion: reference (that) and lexical cohesion (love, person).

Introducing Other People

Martin & Rose (2007: 158):

Moving on to her second love, Helena introduces him as follows:

I met another policeman.

Another identifies him in two ways. First an is indefinite like ‘a’, so we know we can’t assume this identity; secondly other tells us that he is different from the first policeman. These two meanings (indefiniteness and difference) are fused together as another. …

A third man whom she briefly marries is introduced indefinitely as someone, i.e. nobody we know; and is also distinguished from her first love as someone else:

An extremely short marriage to someone else

Blogger Comments:

[1] This is self-contradiction. To be clear, if no identity is recoverable, then there is no identification.

[2] To be clear, other and else are markers of comparative reference, where what is presupposed is another referent of the same class. Contrary to Martin's model of identification, the 'semantics of reference', the domain of comparative reference is not restricted to nominal groups. Halliday & Matthiessen (2014: 633):

Introducing People

Martin & Rose (2007: 157-8):

Helena’s story has three main characters: Helena and her first and second loves. Helena is introduced to us by Tutu:

The South Africa Broadcasting Corporation's radio team covering the Truth and Reconciliation Commission received a letter from a woman calling herself Helena (she wanted to remain anonymous for fear of reprisals).

And Helena in turn introduces her two loves; her first love is introduced as follows:

I met a young man in his twenties.

How are these people introduced? As we can see, the basic strategy here is to use ‘a’ to introduce a woman and a young man. The word ‘a’ tells us that this is someone whose identity we can’t assume. When we can’t assume an identity it is ‘indefinite’, so a participant with ‘a’ is traditionally known as ‘indefinite’. On the other hand ‘the’ tells us that we can assume an identity, so a participant with ‘the’ is traditionally known as ‘definite’. …

When she meets her first love again, it is through a friend, who is also indefinite:

I met my first love again through a good friend.

Blogger Comments:

To be clear, introducing an 'indefinite' participant whose identity can not be presumed is neither participant identification nor textual reference. In terms of identification, the participant is not specified, since the indefinite article is a non-specific determiner.  In terms of reference, non-specific determiners do not serve as reference items because they do not present an identity as recoverable. Halliday & Hasan (1976: 33):

What is essential to every instance of reference whether endophoric (textual) or exophoric (situational) is that there is a presupposition that must be satisfied: the thing referred to has to be identifiable somehow.

On the other hand, the referent through which the identity of a reference item is recoverable can be marked as indefinite or definite, as demonstrated by the fact that the reference item he can refer to a singer or the singer.

Basic Choices In Identification

Martin & Rose (2007: 157):

In simple terms then, what we see here are a range of resources for introducing participants into a discourse and for keeping track of them once there. We can express this as a set of choices, first between introducing participants and tracking them; and second, within tracking, between pronouns, names, and entities with ‘the’. These choices are shown in Figure 5.1. (Note that this is not drawn as a system network, as we need to expand the discussion before we’re ready to draw the system for identification.)

Blogger Comments:

[1] To be clear, 'introducing participants into a discourse' is a speaker projecting experiential meaning, whereas 'keeping track of them once there' is a listener (or text linguist) processing the experiential meanings projected by the speaker.

[2] To be clear, this confuses what speakers do (introducing participants) with what listeners do (tracking participants in the speaker's text) and presents the confusion as choices of the speaker.

[3] To be clear, this confuses reference items (pronouns, 'the'), which serve a textual function, with experiential meanings (names, entities).

It will be seen in the posts that follow that Martin's model of identification ('the semantics of reference') confuses the referent, as experiential meaning, with the reference item that makes a textually cohesive relation. This results in each reference chain involving relations between referents — as experiential meanings — rather than involving textually cohesive relations between reference items and their referents.

Two Basic Misunderstandings That Invalidate The System Of Identification

Martin & Rose (2007: 156):

In order to make sense of discourse, one thing we need is to be able to keep track of who or what is being talked about at any point. When we first start talking about somebody or something, we may name them, but then we often just identify them as she, he or it. By this means our listener/reader can keep track of exactly which person or thing we are talking about, i.e. which participant in the discourse. There are many other ways of introducing participants into a discourse, and keeping track of them as we go, that we will explore in this chapter.

Blogger Comments:

To be clear, Martin's system of identification confuses the instantiation of experiential potential by a speaker (talking about persons and things), with the comprehension strategies of a listener (making sense of discourse, keeping track of what is being said), and presents the confusion as the instantiation of a speaker's textual potential.

That is, it is the speaker who introduces participants, but it is the listener who has to keep track of them. Systemic potential is a model of what speakers can mean, not a model of how listeners understand.

Moreover, participants are construals of experience (experiential meanings), not textual statuses assigned to experiential or interpersonal meanings (textual meanings).  Introducing participants is introducing experiential meaning; tracking participants is tracking experiential meaning.

As will be seen in later posts, this metafunctional misunderstanding derives from confusing the textual meaning of referring with the experiential meaning of the referent.

The Textual Discourse Semantic System Of Identification

Martin & Rose (2007: 155):

Identification is concerned with tracking participants — with introducing people and things into a discourse and keeping track of them once there. These are textual resources, concerned with how discourse makes sense to the reader by keeping track of identities.

Blogger Comments:

To be clear, Martin (1992: 93) glosses identification as 'reference as semantic choice', and it is indeed Halliday & Hasan's (1976) grammatical system of cohesive reference, misunderstood and rebranded by as discourse semantics. (Evidence here.)

The name of the system derives from a second intellectual source, the participant identification of the Hartford stratificationalists (Martin 1992: 95), which, as the term 'participant' makes clear, is concerned with experiential meaning, not textual meaning. It will also be seen in later posts that Martin's system of identification additionally confuses (non-structural) textual reference with the (structural) interpersonal deixis of the nominal group.

The Demonstrated Advantages Of Unpacking Ideational Metaphor

Martin & Rose (2007: 152):

Here is an example of ‘conjunction-as-quality' in Tutu’s argument:

Many of those in the security forces who have come forward had previously been regarded…as respectable members of their communities.

This could be unpacked as:

Many of those in the security forces who have come forward were regarded as respectable members of their communities before…they came forward.

How much we choose to unpack ideational metaphors in our analyses will depend on our purposes. We have shown two advantages of unpacking experiential and logical metaphors. One is that by paraphrasing highly metaphorical discourse in a more spoken form, we can show learners how it means what it does, and also design a curriculum that leads from more spoken to more written modes. Another is that we can recover participant roles and logical arguments that tend to be rendered implicit by ideational metaphor. This can be a powerful tool for critical discourse analysis — revealing implicit nuclear relations such as agency and effect, and implicit logical relations such as cause and effect.

Blogger Comments:

[1] To be clear, this (inelegant) attempt to unpack grammatical metaphor involves reconstruing the temporal Location of a Process — not a Quality — as a temporal relation between two Processes, the second of which is a repetition of the Process embedded as Qualifier:

Many of those in the security forces who have come forward	had	previously	been regarded as	respectable members of their communities
Carrier	Process:	Location	relational	Attribute

Many of those in the security forces who have come forward	were regarded as	respectable members of their communities	before	they	came forward
α			× β
Carrier	Process: relational	Attribute		Behaver	Process: behavioural

[2] This is misleading, because it misrepresents what the authors have demonstrated. To be clear, Martin & Rose have not shown these advantages of unpacking ideational metaphors, not least because, in focusing on individual words, they have confined their explication to elemental metaphors, and largely ignored the syntagmatic syndromes in which elemental metaphors appear, such as the reconfiguring of functions that is necessitated when a sequence is metaphorically reconstrued as a figure. Merely paraphrasing text does not provide a theoretical understanding of what is involved in grammatical metaphor, still less a theoretically-informed curriculum design.

Moreover, because of this narrow focus, Martin & Rose have not shown how to recover participants and logical relations made implicit by ideational metaphor. For example, the unpacking of metaphor, above, does not recover any implicit participants or feature of time. See also the previous illustrative examples:

• Unpacking Grammatical Metaphor: "Conjunction" As Process, and
• Unpacking Grammatical Metaphor: "Conjunction" As Circumstance.

[3] To be clear, this is presumably 'effective', a feature of the grammatical system of AGENCY.