Misrepresenting Matthiessen As Endorsing Martin's Misunderstanding Of Register

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 207-8):

Modelling concurrence and resonance is more of a challenge. One response would be to quantify the semantic ‘weight’ of the contributions from alternative modalities in terms of degrees of commitment – where commitment refers, following Martin (2010), to the number of optional systems taken up and the degree of delicacy of selections from both optional and obligatory systems. Figueredo and Figueredo (2019) outline a quantitative model for measurements of this kind. This would offer us a gauge of how much meaning language and paralanguage were committing but not tell us much about the kinds of meaning involved. 

Another response, perhaps better suited to this shortcoming, would be to turn to a higher-order semiotic such as register (Matthiessen, 2007) or genre (Bateman, 2008) and assign it responsibility for the distribution of meaning across modalities. Models of this kind take advantage of work on the relation of hierarchically organised categories in language (i.e. system realised in structure, higher ranks realised by lower ones and more abstract strata realised through more concrete ones) to explore intermodality – in effect treating co-instantiation across modalities as if it were realisation within a modality.

Blogger Comments:

[1] As previously explained, Martin's notion of 'commitment' is invalidated by the fact that it is based on a fundamental misunderstanding of the system network, namely: that a speaker can choose the degree of delicacy to be instantiated during logogenesis. That is, Martin misconstrues what the linguist can do — decide on the degree of delicacy "pursued" in analysing a text — as what a speaker can do. But Martin also confuses 'delicacy' in the technical sense of a scale of decreasing generality in system networks with 'delicacy' in the sense of a scale of decreasing generality in a hyponymic taxonomy experiential meanings. See the earlier post Why Martin's Notion Of Commitment Is Invalid.

[2] This is very misleading indeed, because it knowingly misrepresents Matthiessen as supporting Martin's misunderstanding of register as a higher-order semiotic. Matthiessen, of course, follows Halliday in modelling register as a language variation (instantiation), not as a system more abstract than language (stratification).

[3] This confuses constituency (rank scale) with symbolic abstraction (realisation). All ranks are of the same level of symbolic abstraction, so a higher rank is not realised by a lower rank.

[4] To be clear, here the authors advocate making the same type of theoretical error that Martin made in his misunderstanding of register. With register, Martin modelled different types of language in terms of a more abstract stratum, and here he proposes modelling different types of semiotic system in terms of a more abstract stratum.

The Linguistic Coordination Of A Convergence Of Language With Language

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 207):

From the perspective of instantiation (the system to text relation in SFL), this raises the central challenge of intermodal studies, namely, the ineffable process whereby systems from different modalities end up seamlessly instantiated as coherent text. As in film (van Leeuwen, 1985, 2005), textual meaning has a critical role to play, as the ‘beat’ of feet and tone groups (TONALITY, TONICITY, SALIENCE and RHYTHM) coordinates the convergence of linguistic and paralinguistic resources (the focus of Chapter 6).

Blogger Comments:

[1] To be clear, instantiation is the relation between potential and instance at a given level of symbolic abstraction, of a given semiotic system. For example, instantiation is the relation between the system of phonology as potential, and an instance of that system in a text; or the relation between a system of paralinguistic content, and an instance of that system accompanying an instance of linguistic content.

[2] To be clear, the ineffability of grammatical categories means that they only mean themselves. Halliday (2002 [1984]: 303, 306):

The meaning of a typical grammatical category … has no counterpart in our conscious representation of things. … they do not correspond to any consciously accessible categorisation of our experience.

[3] To be clear, the expression plane systems that realise the grammatical system of the information unit are language, not paralanguage, whether vocal or gestural. So there is no coordination of a convergence of linguistic and paralinguistic resources.

[4] Trivially, SALIENCE is a lexicogrammatical distinction realised by distinctions in RHYTHM, its phonological counterpart. 

[5] To be clear, Chapter 6 presents a system of DEIXIS which classifies referents, and a system (though no network) that merely correlates speaker location with categories from writing pedagogy.

Misrepresenting Halliday On Theme

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 181-2):

As outlined by Halliday (1967, 1970a), English grammar and phonology structure textual meaning as waves of information. One peak of prominence is realised grammatically through Theme at the beginning of an English clause. It functions as the point of departure for the message by encoding an angle on the field. A complementary peak of prominence, termed New, is realised phonologically in the unmarked case through the major pitch movement on the final salient syllable of a tone group – its Tonic segment (Halliday, 1970a; Martin and Rose, [2003] 2007: 189–92). A secondary peak of informational prominence is realised through a salient syllable, which in SFL notation begins each foot. As noted in Chapter 3, Section 3.6, a salient syllable can be made super-salient where there is a significant jump in pitch, usually upwards, which does not involve a choice of tone. Super-salience is indicated via a vertical arrow, ‘↑’, before the syllable.

Blogger Comments:

[1] To be clear, the notion of the textual meaning of a clause as a wave of prominence is first set out in Halliday (1985: 169).

[2] To be clear, the Theme functions as the point of departure for the clause as message. Halliday & Matthiessen (2014: 89):

The Theme is the element that serves as the point of departure of the message; it is that which locates and orients the clause within its context. The speaker chooses the Theme as his or her point of departure to guide the addressee in developing an interpretation of the message; by making part of the message prominent as Theme, the speaker enables the addressee to process the message.

Field, on the other hand, is the ideational dimension of context — two strata above lexicogrammar — which Martin (1992) misunderstands as register.

[3] To be clear, the New is peak of prominence of the information unit, which may or may not be co-extensive with the clause.

[4] This is misleading, because it credits Martin and Rose with theorising that is entirely Halliday's.

[5] To be clear, a salient syllable is a peak of phonological (rhythmic), not informational prominence. Here the authors confuse expression with content. Each salient syllable that is non-tonic realises what was not selected as the Focus of New information.

Problems With The Authors' Scope And Demarcation Analysis

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 176-7):

In example (15) the PARALINGUISTIC DEIXIS identifies time as [virtual:location] and selects for both SCOPE and DEMARCATION. 

In image 1 in (15), the ‘pinch’ point of left thumb and index finger selects for SCOPE as [narrow], as does the left index finger point in image 2. Both these vectors contrast with the right-hand vector in image 2 where an open palm with spread fingers and thumb configures SCOPE as relatively [broad]. The narrow pinch point in image 1 syncs sonovergently with today and semovergently with the meaning of the narrowly defined time reference. The relatively broad righthand point in image 2 syncs sonovergently with future and semovergently with the relatively open time reference. 

The PARALINGUISTIC DEIXIS in (15) also selects for DEMARCATION as [delineation]. In the second image, the left index finger extends outwards from the body, sustaining its semovergence with today. The left index finger delineates a boundary line, a [virtual:location] from which time stretches into the future, the [virtual:location] identified to the right. Our data suggest that the selection of [virtual:semiotic], whether [prospective] or [retrospective], does not select for either relative SCOPE or DEMARCATION.

Blogger Comments:

[1] As previously explained, this is an example of using body language to make endophoric reference, with the left-right dimension of interpersonal space ideationally construing the past-future dimension of interpersonal time, and the pointing gesture signalling that the meanings 'present' and 'future' are recoverable from those construals by body language. Again, the vector is "resolved" and so the "deixis" is not virtual.

[2] To be be clear, selecting features from systems is the process of instantiation, and it is not a (more inclusive) system that does the selecting.

[3] To be clear, gestures don't instantiate ('select') content plane features, they realise them. That is, the authors here confuse interstratal realisation with system instantiation.

[4] To be clear, it is the expression (hand shape) that is broad or narrow, not the content. As previously explained, on the authors' model, this hand shape realises the ENGAGEMENT feature 'expansion'.

[5] To be clear, the timing of the gesture is linguistic and textual, because, like the tonic, it realises the focus of New information, in this case: today and future.

[6] To be clear, the gesture points to a spatial location that symbolises a temporal location. Any location can be construed as a boundary between other locations, but there is no evidence here that the gesture construes the location as a boundary. This is a case of making the data fit the theory instead of using the theory to account for the data.

Confusing Reference With Engagement

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 173-4):

In (9), from our cultural studies lecture, the lecturer is eliciting responses from students in relation to a projected orientalist image.

In the first image in (9), the lecturer verbally refers non-specifically to any student as a potential respondent (anyone). In paralanguage synchronous with underlined spoken language she extends both forearms with supine hands in front of her body – angling them outwards at roughly 45°. The deictic gestures select for relatively [broad] in SCOPE – the two diverging vectors effectively identify the whole class. 

In the second image, synchronous with the lexical construal of a location in up the back, the lecturer points with an index finger, narrowing the SCOPE of identification to a specific student.

Blogger Comments:

[1] To be clear, anyone has no reference function because it does not signal that a specific identity is recoverable elsewhere. Non-specific determiners like any do not function as reference items (Halliday & Matthiessen (2014: 365).

[2] To be clear, in the first image, these are not pointing gestures, which is consistent with the absence of reference in the language it accompanies. Instead, on the authors' own model, the supine hands realise the engagement feature 'expansion', acknowledging other voices, which is consistent with the instantiation of the engagement feature 'expansion' in the language it accompanies.

[3] To be clear, if the gesture is interpreted as pointing to the whole class, then the feature 'broad' describes the referent, the class.

[4] To be clear, in the second image, the gesture simply makes exophoric reference to the environment of the paralanguage: to a student remote from the speaker.

Problems With The Authors' Analysis Of 'Actual' Paralinguistic Deixis

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 166, 167): 

Example (2) illustrates contrasting instances accompanying the verbal text you’re wearing a stripy shirt – explain this image to me. In each of the three frames a resolved vector is expressed with a hand or index finger point. …  In the first two frames the point is directed outwards selecting [other], first to a student and then to a projected image. In the third it is directed back to the lecturer, selecting [self].

Each of the entities identified through deictic paralanguage in (2) is also tracked exophorically in the spoken text – to a student (you), to a thing (this image) and to the lecturer herself (me). However, as revealed in the first two images in (2), the resolution of the paralinguistic vector does not sync sonovergently with the verbal expressions of identification (i.e. you and this) but rather with the underlined lexis realising relevant entities – specifically the stripy quality of a student’s clothing and the thing entity image. 

In the third image the PARALINGUISTIC DEIXIS is synchronous with the presuming pronoun me which identifies the entity (lecturer). In this instance me is not salient as might be expected. 

This is accounted for in that the synchronous deictic gesture in image 3 is part of a gestural flow that begins on ‘explain’ and culminates with the completion of the tone group – in this case a tail that follows the tonic. The gestural movement maps the flow of information from ‘about what’ to ‘to whom’.

Blogger Comments:

[1] To be clear, the selection of the features 'other' and 'self' is instantiation. The relation between a pointing gesture and these features is realisation.

[2] To be clear, these three pointing gestures are exophoric to the environment of paralanguage. The features 'other' and 'self', on the other hand, distinguish the referents in the environment of paralanguage (in terms of deixis), not the paralinguistic means of referring to them.

[3] To be clear, this confuses two distinct textual functions of body language: reference and salience. The function of the pointing gesture is reference, and this is a feature of epilinguistic body language, as demonstrated by the fact that members of other species, such as rainbow lorikeets, do not respond to them as meaningful. The function of the timing of the beat of a gesture with speech is salience, and it is a feature of linguistic body language because it serves the same function as the beats of speech rhythm.

The direction of the pointing gestures identifies the referents (you, image), whereas the timing of the beat of the gestures highlights elements as salient (stripy, image), both of which, despite the authors' phonological analysis, are likely to be tonic, with each realising a focus of New information.

[4] To be clear, if the pointing gesture is timed to beat with me, it highlights me as salient. This suggests that the phonological analysis mistakes a salient syllable for a non-salient one. A more congruent rhythm would be:

// 1 ‸ex/plain this / image to / me // 

[5] To be clear, this gallant attempt does not account for the supposed lack of salience, since a lack of salience has to be explained in terms the function of salience — to highlight a potential focus of information — which the authors' account does not do. 

Problems With The Authors' Analysis Of Paralinguistic Focus

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 146-7):

In (24), PARALINGUISTIC FOCUS as [sharpen] is expressed in the narrowly targeted index-finger point that zooms in towards the Other Mother. This expression of sharpened FOCUS functions to identify the target (Other Mother) in a highly specifying manner (see Chapter 6). This together with the expression of negative judgement in the spoken language in stole serves to amplify the expression of [disdain] in FACIAL AFFECT.

Blogger Comments:

[1] To be clear, GRADUATION is the scaling of an interpersonal APPRAISAL (Martin & White 2005: 135). The evaluation here is made through the process stole, so any upscaling of the evaluation must be an upscaling of stole not of the evaluated you. So there is no sharpening of the FOCUS of evaluation here (nor a quantifying of the FORCE of evaluation expressed by the extended arm). The interpersonal function of the orientation of the index-finger of the clay puppet here is simply deictic: it points to the addressee you from the speaker I .

[2] To be clear, this is an instance of 'pointing the finger' which is to accuse or blame (someone) — a 'j'accuse' — a gesture which the authors might easily have classified, in their own terms, as an emblem, since they write (p38):

Gestures treated as playing a speech functional role in dialogue in other models are treated as emblems in our framework.

In terms of Cléirigh's original model, this is a genuine example of epilinguistic body language — not to be found in protolinguistic species — in which the gesture expresses an appraisal of judgement, an accusation of wrongdoing, which the authors claimed (pp118, 121) that body language can not do.

[3] To be clear, the judgement instantiated as body language is consistent with the judgement instantiated as language. Whether or not the face of the clay puppet here specifically represents disdain, a feeling of contempt for someone or something regarded as unworthy or inferior is arguable, at the very least.

Why The Argument For Ideational Convergence (Concurrence) Is Invalid

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 94-5):

Here we draw on the concept of commitment ‘which refers to the amount of meaning instantiated as a text unfolds’ (Martin, 2011b: 255) as developed in Martinec (2008) and Martin (2010). 

Language and paralanguage can vary in terms of the amount of meaning that is specified by each semiotic mode. For instance, returning again to the example from the ‘Visit to the Dermatologist’ phase, and as noted in Chapter 1, some entities were committed in the language alone (e.g. the occurrence film in I didn’t film it) and not in the paralanguage. 

There can also be differences in how delicately meaning is committed in language and paralanguage. For example, the needle and the foot bump were more delicately committed in the paralanguage than in language, as far as qualities such as size and shape are concerned. 

So rather than separating gestures into a catalogue of types based on their purported resemblance to things in the world, the approach adopted in this chapter considers how gestures function as a resource which supports ideational meaning-making – focusing on how they concur with ideational discourse semantic selections.

Blogger Comments:

To be clear, this is the authors' argument for modelling paralanguage as convergent with ideational discourse semantics. It can be characterised as:

Premiss 1: Some meanings are made in language, but not in paralanguage.

Premiss 2: Paralanguage and language vary in the degree to which meanings are specified.

Conclusion: Paralanguage will be modelled as realising the ideational meaning of language.

There are two basic reasons why this argument is fallacious. The first is formal: the conclusion does not logically follow from the premisses, since the variation across modes is distinct from the question of whether one realises ("supports") the other. The second is informal: in the premisses and the conclusion, the authors misunderstand paralanguage as an expression-only semiotic system ('gestures'). See further below.

[1] As previously explained (here), Martin's notion of commitment is based on a fundamental misunderstanding of the system network, namely: that a speaker can choose the degree of delicacy to be instantiated during logogenesis. That is, it confuses systemic delicacy, in this case, with the degree to which a Thing (needle, bump) is expanded by Qualities (size, shape).

[2] As previously explained, it was not the needle that was gestured, but how a needle is held, and it was not simply the bump that was gestured, but the bubbling up of the bump (granuloma) after an injection.

[3] To be clear, both of these alternatives misunderstand paralanguage as an expression-only semiotic system ('gestures'), and the second preferred alternative proposes that this expression-only system realises ('supports') the ideational meanings of language.

The first rejected alternative, the only other possibility recognised by the authors, proposes that this expression-only system be categorised in terms of the material order phenomena that the gestures visually resemble.

In Cléirigh's model, the gestures of body language simply realise the meanings of body language, whether used paralinguistically or on their own.

The Contradictions Of 'Different Degrees Of Concurrence'

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 93):

Each phase involves figures and entities that have different degrees of concurrence between language and paralanguage: some are realised only in language, some co-realised in language and paralanguage and some are realised only in paralanguage (see the discussion of mime in Chapter 7). We begin by exploring how entities are multimodally realised in the vlog and then move on to consider figures.

Blogger Comments:

[1] To be clear, by 'concurrence' the authors mean the realisation of the ideational meaning of language in the expression-only system of paralinguistic body language.

[2] To be clear, if ideational meaning is only realised in the expression system of language, there is no concurrence and no paralanguage.

[3] To be clear, if meanings are realised only in body language, then the meanings are not those of the text as an instance of language, so there is no concurrence, since the meanings of the text are not realised in the expression-only system of paralinguistic body language. That is, here the authors incongruously propose a realisation relation between the content potential of language and an expression instance of paralinguistic body language.

[4] To be clear, the discussion of mime in Chapter 7 is recycled from Martin & Zappavigna (2019: 22-6) and discusses the miming of protolinguistic body language. See the comments from the review of Martin & Zappavigna (2019): Mime As Paralanguage.

The Argument That 'Emblems' Are Part Of Language

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 42):

These gestures differ from the semovergent ones illustrated thus far in critical ways (cf. McNeill, 2012: 7–10). For one thing they commit very specific meanings and can be readily recognised without accompanying co-text (linguistic or paralinguistic). As part of this specificity they can enact moves in exchange structure on their own, for example, statements and requests, alongside greetings and leave-takings (hand waving), calls (beckoning gestures), agreement (nodding head), disagreement (shaking head), challenges (upright palm facing forward for ‘stop’) and so on. For another they are the first thing that comes to mind when someone mentions gesture. And in this regard they are often commented on as culturally specific (e.g. the difference between an Anglo supine hand beckoning gesture and its Filipino prone hand equivalent). In both respects emblems contrast with common-sense dismissals of the paralanguage as idiosyncratic (although none of us has any trouble successfully interpreting another speaker’s sonovergent and semovergent systems). From the perspective of the sign languages of the deaf, emblems most strongly resemble signs; they are expression form gestures explicitly encoding meaning. Similarly, from the perspective of character-based writing systems (such as those of Chinese), emblems most strongly resemble characters (but gestured rather than scribed). 

This indicates that from an SFL perspective emblems are better treated as part of language than as a dimension of paralanguage.

 Blogger Comments:

This is recycled almost verbatim from Martin & Zappavigna (2019). Here are the comments from the review of Martin & Zappavigna (2019): The Argument That 'Emblems' Are Part Of Language.

To be clear, here the authors outline their argument for classifying what Kendon terms 'emblems' as language rather than semovergent paralanguage (Cléirigh's epilinguistic body language).

[1] Incidentally, here the authors exemplify the use the word one as a constituent of a conjunctive Adjunct; see the preceding post on the vlogger gesturing the meaning 'one'.

[2] To be clear, in SFL theory, unknown to the authors, the conventionalisation of the meaning of specific gestures in a community corresponds to the move of the sign (content/expression pair) from the instance pole to the system pole of the cline of instantiation.  However, since this can occur in the development of semiotic systems in general — e.g. protolanguage, emoji, pictorial signage — it does not support the authors' argument that emblems are part of language.

[3] To be clear, gestures don't "commit" meanings, they realise them, since realisation is the relation between expression and content.  'Commitment', on the other hand, in Martin's own terms, is concerned with  instantiation, the relation between potential and instance, though, as previously explained here, the notion derives from Martin's misunderstanding of systemic delicacy.

[4] To be clear, here the authors have switched attention from tone groups to exchange structures in an attempt to fudge their argument.  In their own terms, these moves would constitute examples of interpersonal semovergent paralanguage, since the meaning of these gestures "resonates" or "converges" with the meanings of Martin's interpersonal discourse semantic system of NEGOTIATION.  Accordingly, this does not support the authors' argument that emblems are part of language.

[5] The authors' "argument" here is that because these gestures are regarded as prototypical gestures, they are therefore part of language.

[6] To be clear, on the one hand, some emblems are culturally-specific and some are not.  So culture specificity cannot be used as an argument about emblems as a type.  On the other hand, in any case, the culture-specificity of semiotic systems is not confined to language, as demonstrated, for example, by differences in the protolanguages of separated populations of the same species.

[7] To be clear, Halliday (1989: 30-1) distinguishes paralanguage from indexical features, the latter being those that are peculiar to the individual ("idiosyncratic").  So the authors' argument here is that  emblems are language because they are not indexical features.

[8] As this blog demonstrates, the authors do have trouble in interpreting both the meaning of the vlogger gestures and the type of body language involved.

[9] To be clear, the authors' argument here is that emblems are part of language because their expressions resemble the expressions of language (Sign and Chinese), and that, in the case of one of these, at least, the expressions "explicitly encode" meaning.

On the one hand, if this is true, it applies to all languages, not just Sign and Chinese.  On the other hand, the reason it is not true is that the expressions of Sign and Chinese, encode the wording that encodes meaning, whereas the expressions of emblems only encode meaning.  That is, Sign and Chinese, being languages, are tri-stratal, whereas emblems, not being language, are bi-stratal.  Once again, the authors' argument does not support their claim that emblems are part of language.

[10] As the above clarifications demonstrate, not one of the arguments offered by the authors supports their hypothesis that emblems are part of language.

Why Martin's Notion Of Commitment Is Invalid

Ngo, Hood, Martin, Painter, Smith & Zappavigna (2022: 30):

In terms of commitment (i.e. the amount of meaning specified across semiotic modes; Martin, 2010; Painter et al., 2013), the ‘dermatologist’ and ‘steroid’ are committed in the language but not the paralanguage; but the ‘needle’ is more delicately committed in the paralanguage as a tiny pointed entity and then as a syringe. And the paralinguistic commitment of the ‘bump’ convergent with (69) in fact takes place two tone groups after it is committed verbally in (67).

 Blogger Comments:

This is recycled verbatim from Martin & Zappavigna (2019). Here are the comments from the review of Martin & Zappavigna (2019): Martin's Notion Of Commitment.

[1] To be clear, Martin's notion of 'commitment' is invalidated by the fact that it is based on a fundamental misunderstanding of the system network, namely: that a speaker can choose the degree of delicacy to be instantiated during logogenesis.  As Martin (2011: 255-6) explains:

Instantiation also opens up theoretical and descriptive space for considering commitment (Martin 2008, 2010), which refers to the amount of meaning instantiated as the text unfolds.  This depends on the number of optional systems taken up and the degree of delicacy pursued in those that are, so that the more systems entered, and the more options chosen, the greater the semantic weight of a text (Hood 2008).

To be clear, a system network is not a type of flowchart, such that instantiation involves a movement through more and more delicate systems.  A system network is a network of relations.  In the case of lexicogrammar, the system specifies how all the features are related to each other, such that the instantiation of each lexical item in a text is the instantiation of all the features that specify it, from the most general all the way to the most delicate.

In short, Martin misconstrues what the linguist can do — decide on the degree of delicacy "pursued" in analysing a text — as what a speaker can do; but see also [3] below.

[2] Translating into SFL theory, the claim here is that the meanings 'dermatologist' and 'steroid' are instantiated in the language but not in paralanguage.  However, this is manifestly untrue.  As previously explained, the body of the speaker herself represented the dermatologist in two figures ('taking the needle' and 'injecting the steroid').

The instantiation of the meaning 'steroid' is more subtle.  Because, as Halliday & Matthiessen (1999: 156) point out, a Process requires a Medium for its actualisation, the gesture representing the 'injecting' Process implicates a Medium, and so the meaning 'fluid' is at least implicated in the gesture, even if the meaning 'steroid' is not precisely specified.  (Try gesturing the meaning 'steroid'.)

To be clear, the authors' false claim derives from two procedural errors:

• assuming that handshape is the only bodily expression of ideational meaning here, and
• analysing at the level of element ("entity") instead of figure (while claiming the latter).

[3] As explained in [1], Martin's notion of "more delicate commitment" is nonsensical, based as it is on his misunderstanding of what system networks represent.

However, here it can also be seen that Martin confuses 'delicacy' as a scale of decreasing generality in system networks with 'delicacy' as a scale of decreasing generality in construing experience as meaning, as in 'needle' vs 'tiny pointed entity'.   It can also be seen that, even in these terms, the authors have the relation backwards, since 'tiny pointed entity' is a more general construal than 'needle', not more delicate, since, as a class, it includes a broader range of potential members.

[4] To be clear, the word 'syringe' is not instantiated in the data.  It appears only in the authors' gloss of the body language accompanying the wording and injected this like steroid.

[5] As explained in the previous post, the reason why this gesture is made with the final figure, and not the second, is that it realises the nucleus of the final figure, it all bubbled up, rather than the meaning of the word bump in the second figure.  The authors' confusion again arises from analysing isolated elements instead of their functions in figures.