A Fatal Theoretical Flaw And Misrepresenting Du Bois (1980)

Martin (1992: 142):

As noted in 3.1 above, first mentions of participants in discourse are commonly phoric.  Exophoric and homophoric first mention is quite common; parts and possessions are typically introduced via possessive Deictics; and bridging is favoured where the identity of a participant can be indirectly presumed.  As DuBois (1980:254) comments, "the definite article is the unmarked member of the pair.  When consciousness is not focussed on the task of introducing characters, it is the unmarked the which is uttered, whether or not the initial mention is in fact identifiable."  It may be that Dubois' data has been biassed in this way by its re-telling context — that of speakers recounting the plot of a short film they have just seen.  But it's certainly true that the principle of introducing participants phorically where possible is a very general one across a wide range of English texts.

Blogger Comments:

[1] To be clear, by this, Martin simply means that the first occurrence of an instantial participant is usually realised by a nominal group whose Deictic is realised by the definite article (phoric = Martin's presuming reference) rather than the indefinite article (non-phoric = Martin's presenting reference).

[2] It will be seen that, in Martin's model, where the identification is between participants, possessive Deictics wrongly identify possessors with possessions, as demonstrated by:

Jim gets a lot of things wrong, and so do his students.

Here Jim and his students, on Martin's model, are necessarily identified as the same participant. This demonstrates most clearly the absurd consequences that arise from mistaking nominal groups for reference items.

[3] As previously demonstrated, Martin's "bridging" confuses lexical cohesion with reference.

[4] On the one hand, this deliberately misrepresents Du Bois (1980: 254), whose view on the relative frequencies of definite and indefinite articles in this regard derives from a corpus of one million words: 

In either such case (premature introduction or late introduction), the initial mention receives the definite article instead of the expected indefinite article. This suggests that the definite article is the unmarked member of the pair. When consciousness is not focussed on the task of introducing characters, it is the unmarked the which is uttered, whether or not the initial mention is in fact identifiable. There are other reasons to consider the unmarked, among them sheer frequency. In a corpus of one million words of written English, the definite article was by far the most common word, with 69,971 occurrences; whereas the indefinite article (a ~ an) had only 26,984 occurrences (Kucera and Francis, 1967:5).

On the other hand, it should be pointed out that Martin adopted the same strategy as Du Bois (1980), in as much as the data for this chapter — text [3:88] (p147) — is a child's retelling of a published story.  A significant difference, however, is that, in almost every case in the chapter's exposition, Martin has ignored this text and instead used his own constructed examples involving characters from that text.

Mistaking Nominal Groups For Reference Items

Martin (1992: 141-2):

Again, the dependency relations in the chain are worked out in accordance with the principle that anaphoric items are related backwards once to the item last realising or presuming the information they presume; thus [in 3:79] the boy in the fifth clause and the dog in the sixth are shown to presume they in the fourth.

[3:79]

The boy and the dog woke up
and saw that the frog was missing.
with the boy leading the way [sic]

They headed off for the woods,
with the boy leading the way
and the dog following behind him.

Blogger Comments:

This continues the mistaking of nominal groups for reference items, on the basis of theoretical misunderstandings previously identified in this chapter review.  In the fourth clause of this text constructed by Martin, the reference item they makes anaphoric personal co-reference to first-mentioned boy and dog.  In the fifth clause, it is the reference item the that makes anaphoric demonstrative co-reference to first-mentioned boy, and in the sixth clause, it is the reference item the that makes anaphoric demonstrative co-reference to first-mentioned dog.

Two Theoretical Problems With Reference Chains

Martin (1992: 140-1):

At its simplest, a reference chain consists of two items, one presuming, one presumed.  For display purposes, the nominal group realising a given participant in a text will be used to name that realisation of the participant in the reference chain. […] For longer texts, these chains can simply be extended, with presuming items shown to be as well presumed; these then form the systemic equivalent of the "participant lines" in the Hartford stratificationalists' reticula […].  The basic descriptive strategy underlying chains formed in this way is to take each phoric item back once to the item which last realised or presumed the information that needs to be recovered.

Blogger Comments:

[1] This notion of reference chains, as linked instantial participants sharing the same identity, is not consistent with what the model is purported to be: 'reference as semantic choice'.  This is because, in SFL, reference is the grammatical system that is the resource for referring (personal and demonstrative co-reference and general and specific comparative reference).  In order to be a semantics of reference in SFL, the meanings realised by the grammatical system of reference need to be modelled, specifying how such semantic choices are realised at the level of lexicogrammar.

[2] Reference chains as linked instantial participants sharing the same identity, are not structures that realise a system of referring.  Moreover, reference chains are not structures that realise any of the systems provided in this chapter (pp 102, 105, 106, 110, 112, 114, 118, 120, 124, 126), and, naturally, none of these systems provide realisation statements specifying how their features are realised as structures.

Metafunctional Inconsistency

Martin (1992: 140):

Here, with reference structures, it is the dependency approach that will be pursued.  The reason for this is that reference structures are essentially covariate ones.  Where endophoric, phoric items depend semantically on the items they presume, but need not be grammatically related to them.  And presuming items can themselves be presumed.  The covariate structures deriving from phoric options in the IDENTIFICATION system will be referred to as reference chains (closely related to Hasan's identity chains; see Hasan 1984, 1985).

Blogger Comments:

[1] No argument has been made as to why these identity relations between instantial participants are  textual structures.  They are simply asserted to be so.

[2] Dependency relations are features of the logical metafunction that obtain between rank units in complexes (forms).  Here they are being used for the textual metafunction, rather than the logical, which is a major theoretical inconsistency.  The fact that they are not applied to rank units in complexes is a further (internal) theoretical inconsistency.

[3] To be clear, the relation between structure and system is realisation (elaboration + identity).

The Notion Of "Grammatical Metaphor Functioning As A Kind Of Discourse Process"

Martin (1992: 139-40):

The responses in [3:77] can be read as scaled according to the degree of "participanthood" the reference items entail.  So respects clausehood; but in reconstructing the meanings made in She saw them building a new school as reported locutions or ideas, facts and acts that and it are transforming text into thing.  This is another instance of grammatical metaphor, this time functioning as a kind of discourse process.  This point will be taken up again below in connection with internal conjunction (Chapter 4) and metalinguistic lexis (Chapter 5).

Blogger Comments:

[1] This confuses the experiential metafunction (degree of "participanthood") with the textual metafunction (reference items).  For what 'degree of participanthood' means in SFL theory, see Halliday & Matthiessen (1999: 167-72).

[2] The wording "respects clausehood" is chosen to mislead, since it disguises a theoretical inconsistency.  In SFL theory, these instances of so substitute for a clause — they do not refer to it (or "respect clausehood").  As previously explained, Martin misunderstands substitution as a type of reference, rebranding the misunderstanding "redundancy phoricity", and relocating it from the stratum of lexicogrammar to his stratum — misunderstood as a module — of discourse semantics.

[3] To be clear, the reference items that and it do not "reconstruct the meanings", they simply refer to them.

[4] As previously noted, in Martin's example, [3:77ii], the experiential functions of the reference item that are Verbiage and Phenomenon, not locution and idea.

[5] To be clear, the reference items that and it do not "transform text into thing"; they simply refer to referents that resolve their identity.

[6] To be clear, grammatical metaphor is a mismatch between strata, semantics and lexicogrammar, such that two meanings are made at once, one metaphorically and one congruently.  Martin's characterisation of grammatical metaphor as a "discourse process" not only misunderstands grammatical metaphor, but is also inconsistent with his own later (p401, Chapter 6) misunderstanding of grammatical metaphor as a 'texturing interface' between strata, the latter misunderstood as modules.

[7] This point was not taken up again: neither in the 111 pages of Chapter 4 nor the the 108 pages of Chapter 5, which is why Martin provides no page numbers or chapter subheadings.

∞

In the text examined in this and the previous post, Martin has done his best to make his theoretical misunderstandings and inconsistencies unintelligible, the two main techniques being:

• presenting the text examples and the discussion of them in reverse order, and
• switching from the experiential functions of reference items in (i) and (ii) to experiential functions of the referents in (iii-v).

Confusing Metafunctions And ConfusingTypes Of Cohesion

Martin (1992: 139):

To illustrate this [i.e. "constructing participants by referring to them"] consider the range of responses possible in [3:77]:

			PRESUMED AS
[3:77]	She saw them building a new school
i.	— If you say so.	— Do you think so?	locution/idea
ii.	— Who told you that?	— I can't believe that.	locution/idea
iii.	— That's impossible / It's not possible.		fact
iv.	— That's hard work / It's hard work.		act
v.	— That's what we needed / It's what we needed.		thing

Response v illustrates the pattern considered to this point: reference between participants — that and it presume a phenomenon, realised earlier as a new school.  In response iv however, that and it presume a macro-phenomenon — the act of building a new school…  .  Response iii takes this one step further, presuming a metaphenomenon — the fact that she saw them building a new school: cf. It's not possible [[that she saw them building a new school]].  In ii that also presumes she saw them building a new school as a metaphenomenon — but this time as a projected locution or idea, not a fact.  That contrasts with so in i; as Halliday (1985:234) points out, with verbal processes the distinction is between quoting (that) and reporting (so), whereas with mental processes the opposition is between assertion (that) and postulation (so).

Blogger Comments:

[1] This confuses the experiential function of that and it (participant) with the textual function of that and it (reference item).  The reference relation obtains between the reference item and the referent.

[2] To be clear, the reference items that and it do not refer to the referent whose experiential function is Thing (school), but to that which functions as the Goal (a new school) of an embedded clause serving as Phenomenon: macrophenomenon.

[3] Here Martin creates an inconsistency by switching from the experiential function of the referent to the experiential function of the reference item.  In doing so, he misconstrues the function of each referent, Verbiage and Phenomenon, as projected clauses, locution and idea, respectively.

[4] Here Martin mistakes an instance of substitution (so) for a reference item, and analyses the ideational function of the substitute, rather than what he believes to be the referent.

[5] This is misleading, since, as Halliday (1985: 234) points out, the distinction is between that and so is the distinction between reference and substitution, respectively.  He then goes on to explain why reference is typically used for quotes, and why substitution is used for reports, before concluding:

In verbal processes, therefore, he said that simply attests to his production of the wording, whereas he said so raises the issue of whether what he said is in fact the case.

[6] To be clear, what Halliday (1985: 234) actually wrote was as follows:

Confusing Ideational Construal With Textual Reference

Martin (1992: 139):

Alongside these grammatical resources for constructing participants, discourse semantics can also be used to turn non-participant meanings into things. It does this by using it, this and that to refer to text. This is discussed by Halliday and Hasan (1976: 52-3, 66-7) under the headings of extended reference (to text as act) and text reference (to text as projection). They point out that this is the main use made of demonstratives in most registers of English. Instead of nominalising non-participants to treat them as participants, it, this and that simply construct them by referring to them.

Blogger Comments:

[1] To be clear, here, as elsewhere, Martin confuses the ideational metafunction ('constructing participants', 'turning non-participants into things') with the textual metafunction ('referring to them').

[2] This is misleading.   For Halliday and Hasan (1976), as well as Halliday ± Matthiessen (1985, 1994, 2004, 2014), reference is a textual resource of the grammar.  Martin here leaves the naïve reader with the impression that Halliday and Hasan endorse Martin's relocation of their extended and text reference to his discourse semantics.

Misunderstanding Nominalisation And Reference

Martin (1992: 138-9):

The following text, villified [sic] as "superb balderdash" by Tarzie Vittachi (1989:11), creates numerous generic participants of this kind (nominalised participants in bold face).  None of the participants the grammar constructs are presumed in [3:76], but all have the potential to provide referents for subsequent phoric nominal groups.

[3:76]

The implementation of key targets as operational components of the new strategy — and hence also of the process of negotiation — may be conceived in the time frame of a decade but only in the form of a dynamic process, with different time frames for different components, and with an inbuilt and effective mechanism for review and reappraisal, leading to adjustments and correctives whenever the strategy is seen to deflect from the goals and objectives of development for which it was devised. It should be in the form not of a 'plan of action' but rather of a manifesto, which provides the framework of a sustained commitment to, and implementation of, developmental goals and their operational components, and embodies institutional mechanisms for continuous negotiation, monitoring, appraisal, criticism and modification.

Blogger Comments:

[1] As previously explained, Martin''s generic vs specific distinction is a confusion of composition (extending: part-whole) and instantiation (elaborating: token-type).  In terms of SFL theory, the relevant dimension here is countability;  Halliday & Matthiessen (2014: 385):

We could think in terms of a cline of countability, ranging from those nouns (and pronouns) which construe things as fully itemised, at one end, to those which treat them as totally unbounded at the other. Typically, living beings and concrete objects are itemised, abstract entities (and nominalised processes and qualities) are unbounded, with institutions and collectives falling in between.

[2] To be clear, Halliday & Matthiessen (2014: 729) describe nominalisation as follows:

Nominalising is the single most powerful resource for creating grammatical metaphor. By this device, processes (congruently worded as verbs) and properties (congruently worded as adjectives) are reworded metaphorically as nouns; instead of functioning in the clause, as Process or Attribute, they function as Thing in the nominal group.

On this basis, operational, developmental, strategy, mechanism and goal are not nominalisations because they are not processes or qualities (properties) reworded as nouns.

[3] This is manifestly untrue.  For example, of the four highlighted reference items:

• the makes anaphoric demonstrative co-reference to the first mention of strategy;
• it makes anaphoric personal co-reference to the second mention of strategy;
• It makes anaphoric personal co-reference to implementation; and
• their makes anaphoric personal co-reference to developmental goals.

Martin's rebranding of (his misunderstanding of) Halliday & Hasan's (1976) demonstrative and personal co-reference is "reminding phoricity".

"Villified"

Martin (1992: 138-9):

The following text, villified [sic] as "superb balderdash" by Tarzie Vittachi (1989:11), creates numerous generic participants of this kind (nominalised participants in bold face).  None of the participants the grammar constructs are presumed in [3:76], but all have the potential to provide referents for subsequent phoric nominal groups.

[3:76]

The implementation of key targets as operational components of the new strategy — and hence also of the process of negotiation — may be conceived in the time frame of a decade but only in the form of a dynamic process, with different time frames for different components, and with an inbuilt and effective mechanism for review and reappraisal, leading to adjustments and correctives whenever the strategy is seen to deflect from the goals and objectives of development for which it was devised.  It should be in the form not of a 'plan of action' but rather of a manifesto, which provides the framework of a sustained commitment to, and implementation of, developmental goals and their operational components, and embodies institutional mechanisms for continuous negotiation, monitoring, appraisal, criticism and modification.

Blogger Comments:

Here is what Vittachi (1989) actually wrote (available here):

What on earth can that superb balderdash mean, especially to a villager in India or Tanzania even if it was translated into Hindi or Swahili? It was not intended for the man or woman at the receiving end but for the author's fellow elitists. The fact that none of them could decipher it matters not at all. The intention was to mystify everybody, including the writer. It was an exercise in total discommunication.

As Halliday & Matthiessen (1999: 272) point out:

So the more the extent of grammatical metaphor in a text, the more that text is loaded against the learner, and against anyone who is an outsider to the register in question. It becomes elitist discourse, in which the function of constructing knowledge goes together with the function of restricting access to that knowledge, making it impenetrable to all except those who have the means of admission to the inside, or the select group of those who are already there.

It is this other potential that grammatical metaphor has, for making meaning that is obscure, arcane and exclusive, that makes it ideal as a mode of discourse for establishing and maintaining status, prestige and hierarchy, and to establish the paternalistic authority of a technocratic elite whose message is 'this is all too hard for you to understand; so leave the decision-making to us.

To be clear, Tarzie Vittachi was an award winning journalist who took personal risks to fight for social justice:

Tarzie Vittachi (September 23, 1921 – September 17, 1993), a Ramon Magsaysay Award-winning journalist (1959), was born in Colombo, Ceylon. He authored two popular columns "Bouquets and Brickbats", and "Fly by Night" in the Ceylon Daily News. He later became the youngest editor (at 32) of the oldest newspaper in Asia, The Ceylon Observer which was founded in 1834. He wrote a book known as Emergency 58 about the Government's involvement in the country's race riots known as 1958 anti-Tamil pogrom that won him the Magasaysay Prize in 1959 and led to his being declared persona non grata. From 1960 to 1965 he was Asian director of the International Press Institute, an organisation of editors devoted to promoting the freedom of the press. He was, at the same time, a correspondent for The Economist, the BBC and The Sunday Times of London and wrote a column for Newsweek. A book about the role of the Children's Fund in arranging truces to protect children in time of conflict, called "Between the Guns," was published posthumously.

Two Bare Assertions Based On A Logical Confusion

Martin (1992: 138):

High levels of nominalisation characterise abstract written English, especially in the context of science, the humanities and administration.  The texts typically involve generic rather than specific reference and so few of the nominalised participants play a role in long reference chains; esphoric the is common in these participant creating nominal groups (e.g. the implementation of key targets as operational components of the new strategy — and hence also of the process of negotiation).

Blogger Comments:

[1] On the one hand, this is a bare assertion, unsupported by evidence, such as the analysis of data.  On the other hand, the distinction between generic and specific reference, which Martin (1992: 103) defines as: 

Generic reference is selected when the whole of some experiential class of participants is at stake rather than a specific manifestation of that class…

though labelled as a relation of delicacy (generic vs specific), confuses part-whole relations ('whole') with token-type relations ('manifestation of a class'), the former, the extending logical relation of composition, the latter, an elaborating logical relation of instantiation; see Halliday & Matthiessen (1999: 145-6).

[2] On the one hand, no reason is given why generic reference should affect the length of reference chains, and on the other hand, this is a bare assertion, unsupported by evidence, such as the analysis of data.

[3] To be clear, 'esphoric' reference is Martin's rebranding of Halliday & Hasan's structural (rather than cohesive) cataphoric reference.

Oversimplifying Nominalisation

Martin (1992: 138):

Certainly the major resource as far as the grammar creating participants is concerned is nominalisation. Through this resource the grammar is able to turn a range of non-participant meanings into participant-like ones.

Blogger Comments:

[1] It is the transitivity system of the clause that is the grammatical resource for construing experience as participants in processes.

[2] This construal of nominalisation is potentially misleading.  Through nominalisation, meanings that are not congruently realised as participants can be metaphorically realised as wordings that construe participants.

The metaphorical form then also embodies semantic features deriving from its own incongruent lexicogrammatical properties, such that the two meanings, "non-participant" and participant, are themselves in an elaborating token-value relation within the semantic stratum, with the metaphorical Token (participant) realising the congruent Value ("non-participant").

Another Reference Chain Of Non-Participant And Participant

Martin (1992: 137-8):

Similarly, the nominal groups realising Circumstances of Extent, or filling Pre-Deictic, Pre-Numerative, Pre-Epithet or Pre-Classifier positions may be taken up as participants in a discourse, although the textual motivation for selecting a nominal group realisation may not have been to introduce a new participant in the first place.  This is illustrated with respect to Pre-Numeratives in [3:75].

[3:75] I got two jars of mulberry jam.
           — Let's see them / I hate it.

Blogger Comments:

Here again Martin violates his own model in its own terms by proposing a reference chain between a non-participant (pre-Numerative: two jars) and a participant (them).

To be clear, in SFL theory, the reference item them makes anaphoric personal co-reference to the referent two jars of mulberry jam (just as the reference item it makes anaphoric personal co-reference to the referent mulberry jam).

A Reference Chain Of Non-Participant And Participant

Martin (1992: 137):

Finally, it has to be noted … that the grammar may be actively involved in creating participants in discourse.  Once the decision for example has been made to realise a process as a Process + Range structure (e.g. have a bath), then the fact that the process has been nominalised potentialises it as a participant; and a text can pick it up and develop it along these lines:

[3.74] Where have you been?

           — Oh I had a bath;

           in fact it was quite a splendid one.

           There was plenty of hot water;

           so I just luxuriated in it for hours.

Blogger Comments:

[1] To be clear, if the grammar isn't "actively involved in creating participants in discourse", then there are no "participants in discourse".

[2] To be clear, in SFL theory, a Range is a participant, whereas in Martin's discourse semantics, it is not, since he has defined participants in terms of nominal groups functioning as the Agent or Medium of a clause (p129).

[3] To be clear, in Martin's model, identification involves reference chains that link presuming and presumed participants (p140), the participant being the unit of identification (p325).  By proposing a  reference link between a non-participant and a participant, Martin violates his own model in its own terms.

[4] To be clear, this it is not a participant, but the Range of a minor Process of a prepositional  phrase realising a circumstance of Location.

As a consequence, the reference chain of "participants" in this example begins with a non-participant (in Martin's terms) and ends with a circumstance (in SFL terms).

Confusing Semogenesis With Stratification

Martin (1992: 137):

Such examples demonstrate one of the respects in which textual meaning provides only a partial account of text structure and the importance of taking the contribution a meaning making grammar makes to discourse structure into account.

Blogger Comment:

This confuses semogenesis, meaning making, with the stratification of the content plane into two levels of symbolic abstraction: higher level meaning (semantics) and the lower level wording (lexicogrammar).  The wording of the grammatical stratum (Token) realises the meaning of the semantic stratum (Value).  As an identifying relation, each can be used to identify the other:

• The identity encodes the meaning of the semantic stratum (Value/Identified) by reference to the wording of the grammatical stratum (Token/Identifier).

• The identity decodes the wording of the grammatical stratum (Token/Identified) by reference to the meaning of the semantic stratum (Value/Identifier)

Confusing Participant Identity With The Systemic Means Of Referring To Referents

Martin (1992: 136-7):

Grammar also plays a rôle in transforming the identity of a participant.  The Attributes referred to above as not themselves realising participants are commonly used to add qualities to and/or classify a participant in a new way.  The Attribute in [3:72] serves both functions:

[3:72] The woman went off to visit her step-sister, who was an old potter who lived in the woods.

Attributes extend the ways in which a participant can be identified; the grandmother can be referred to as the potter, the old lady, the older woman and so on as well as her step-sister in the text following [3:72].

Material processes are also commonly involved in transforming participants.  Renovation of this kind is exemplified in Halliday and Hasan's (1976: 2) example, reproduced as [3:73] below.  Obviously the process of washing and coring apples transforms their identity, and the participant presumed by them is not experientially identical to that referred to with six cooking apples; this would be even more radically the case had the recipe suggested wash and chop instead of wash and core.

[3:73] Wash and core six cooking apples.

           Put them into a fireproof dish.

Blogger Comments:

[1] This confuses the instantial identity of a participant with the cohesive resource for referring to elements in discourse —the system of reference.  This confusion undermines the model in this chapter because it makes it inconsistent with its purported aim: 'reference as semantic choice'.

[2] In the realm of the ideational metafunction, the outcome of 'transformative' material processes is an elaboration, extension or enhancement of the Actor (‘intransitive’) or Goal (‘transitive’) (Halliday & Matthiessen 2004: 186).

[3] To be clear, what the Halliday and Hasan example demonstrates is the anaphoric personal co-reference made by the pronoun them.

Confusing Identity With Identifiability

Martin (1992: 136-7):

To this point participant identification has been approached from the point of view of discourse semantics, with a focus on the nature of the phoric items and the work interlocutors must do to track down presumed information.  But it is important not to overlook the rôle that grammar plays in participant tracking.  For one thing, identifying structures (Halliday 1985: 112-128) may grammaticalise the relation between presuming and presumed.  This happened in [3:31] above, where the Token that (referring to John) was explicitly assigned the Value my husband.

[3:31] …Friends of my parents had a beagle and John hadn't seen a beagle (that's my husband) so he went and looked at one…

Apposition is also used in this way; [3:31] might have been spoken as follows, with the nominal group complex John, my husband doing the work of the Token^Value structure above:

[3:71] …Friends of my parents had a beagle and John, my husband, hadn't seen a beagle so he went and looked at one…

This kind of identification is referred to by Hasan (1984, 1985) as one kind of instantial reference. 

Blogger Comments:

[1] The rôle of the grammar, with respect to semantics, is to realise the meanings under consideration as wordings.

[2] The inclusion of identifying clauses here from the experiential grammar highlights the fundamental confusion that pervades this chapter: misconstruing the system concerned with identifiability — cohesive reference — as a system concerned with the instantial persistence of participants (participant identification).

[3] This is not quite right.  The apposition expresses the elaboration that also obtains in intensive identifying clauses; it does not do 'the work of the Token^Value structure', which is an identifying relation between two levels of abstraction.

[4] This is misleading, since it misrepresents Hasan (1984, 1985).  Hasan's term is 'instantial semblance', and it is a type of lexical cohesion, not reference.  Hasan (1985: 81):

All lexical cohesive devices discussed above are general in nature.  For example, the relation of synonymy between lady and woman is a general fact of English.  There are cohesive devices that are entirely specific to a single text, for example, those of INSTANTIAL SEMBLANCE as in all my pleasures are like yesterdays (Hasan 1984).

To be clear, all co-reference, personal and demonstrative, is necessarily instantial.

Martin's example confuses the instantial lexical cohesion of John and husband with the demonstrative reference of that to John (anaphorically) and to my husband (cataphorically).