One participant elaborates another one along the dimensions of delicacy, realisation, or instantiation. In other words, the elaboration sets up a relationship either of generality (delicacy), of abstraction (realisation), or of token to type (instantiation). There is another variable whereby elaboration involves either identity or membership along the dimension in question.
Figures of being & having construe relations between participants. They construe the same overall range of relations as expanding sequences, and the basic subtypes also correspond to the subcategories of expansion, viz elaboration, extension, and enhancement. Once we realise that the semantic system construes phenomena according to trans-phenomenal (fractal) principles, it will seem natural that figures of being & having construe relations in such a way that they resonate with the semantic types manifested in expanding sequences. They do not construe an arbitrarily different theory of relations. In a figure of being & having, one participant may thus elaborate, extend, or enhance another one.
Cognition is arguably closer to perception than desideration is — there are certain cross-overs like see in the sense of ‘understand’ alongside its basic sense of visual perception, and both can be construed in an active mode as processes of behaviour.
Cognition and desideration are different from both [emotion and perception] in that they can project (ie bring the content of consciousness into existence), can stand for modalities, and are not in general like either behaviour or ascription; they may be interpreted as the most central classes of sensing.
Emotion seems to be closer to quality-ascription than to a prototypical process; it arises from but does not create projections. In contrast, perception is essentially closer to behavioural processes.
… when the different types of sensing are construed metaphorically as things, they are reified in different ways. Perception, cognition and desideration are reified as bounded, ie countable things, such as sight(s), thought(s), plan(s), whereas emotions are reified as unbounded things, ie masses, such as anger, fear, frustration. That is, emotion is construed as boundless — like physical resources such as water, air, iron and oil. … In being construed as unbounded mass, emotions are again more like qualities (cf unbounded strength, height, heaviness, redness).
Related to the possibility of construing emotion as an Attribute is the possibility of scaling or intensifying emotive processes: many qualities can be intensified. We find sets of processes differentiated essentially according to a degree of intensity — scare : terrify, horrify; and emotive processes can be intensified by means of adverbs of degree such as much, greatly, deeply. These options are also open to some cognitive and desiderative processes, although not to perceptive ones; but intensification is an essentially emotive characteristic.
With many processes of emotion, there is an alternative construal of the emotion as a quality that can be ascribed as an Attribute to a Carrier in a relational clause; and this alternative exists for both the ‘like’ type and the ‘please’ type. Thus I’m afraid of snakes is an ascriptive alternative to the mental I fear snakes; similarly, in the other direction, snakes are scary and snakes scare me. This relational type of alternative exists for some cognitive and desiderative processes, but is much more productive with emotive ones. Analogous attributes in the domain of perception seem always to involve potentiality (visible, audible).
Processes of perception are unique among the different types of sensing in that they are agnate to a set of relational processes of ascription, those that ascribe an Attribute in terms of the way in which it presents itself to our senses, as in Madam, you’ll look like a tulip.
The different types of sensing have somewhat different potentials for unfolding in time. With perception and cognition we have various categories of duration, inception, and the like: e.g. (perception) glimpse, sight, spot as well as see; (cognition) discover, realise, remember as well as know. But similar distinctions do not seem to obtain with desiderative and emotive processes.
There are some behavioural processes that are agnate to cognitive ones (pondering, puzzling, meditating) but none that are agnate to desiderative or emotive ones. (Behavioural processes of giggling, laughing, crying, smiling and the like are outward manifestations of emotions; but they are not active variants of inert emotive processing such as rejoicing, grieving, and fearing.
All the modes of perception may be construed either as behaviour or as sensing. One significant grammatical difference is that present behaviour would normally be reported as present-in-present (the present progressive) — What are you doing? I’m watching the last whales of August. — but present sensing would not — I (can) see the whales in the distance. Another one is that only sensing can involve a Phenomenon of the metaphenomenal kind. As long as the ‘phenomenon’ is of the same order of existence as ordinary things, there is no problem with either process type; we can both see and watch macro-phenomena: I saw/watched the last whales leave the bay. But while we can say I saw that he had already eaten we cannot say I watched that he had already eaten, which includes a metaphenomenon. This is the borderline between the mental and the material domains of experience.
Sensing is not construed in the grammar as activity. But certain types of conscious process may be construed not only as sensing but also alternatively as a kind of doing — as behaviour (as if active sensing). […] The difference is suggested quite clearly in the example — You can hear the sea sometimes if you listen very carefully.
Cognitive and perceptive processes may be bidirectional but favour the ‘like’ type — perception almost exclusively so; ‘please’ type perception such as the noise assailed my ears seems quite marginal.
Processes of desideration are not bi-directional; here there is no ‘please’ type, only the ‘like’ type. Here the grammar upholds the view that we are in control of our desires.
The talking snake tempted the naked woman to eat a piece of magic fruit.Cf
The talking snake made the naked woman want to eat a piece of magic fruit.
Processes of emotion are typically bidirectional. They can be construed either as the emotion ranging over the Phenomenon or as the Phenomenon causing the emotion — as in I like Mozart’s music (the ‘like’ type) : Mozart’s music pleases me (the ‘please’ type) Here the grammar of English construes a complementarity between two conflicting interpretations of emotional processes, with opposing angles on whether we are in control of our emotions, as if neither one by itself constitutes a rounded construction of experience.
Both cognition and desideration may be brought about through verbal action:I have told you that : you know that :: I have persuaded you to : you intend to.There are no related verbal types causing perception and emotion.
Emotion is related to interpersonal attitude — I rejoice that she’s returned : she has, happily, returned. Unlike modality, attitude is not an assessment of the validity of the clause (grammatically it is not a Mood Adjunct). Rather, it is a comment on the information presented in a clause.
Both cognition and desideration can come to serve as metaphors for the interpersonal system of modality — for modalisation and modulation respectively — alongside congruent realisations such as modal auxiliaries and adverbs. That is, a number of processes of cognition can stand for probabilities — I think : probably, I suppose : perhaps; and a number of processes of desideration can stand for inclinations and obligations — I want : should, I insist : must. For instance:I think that in a sense you’ve had to compromise, haven’t you?Neither perceptive nor emotive sensing can serve as metaphors for modalities.
A process of sensing may range over or be caused by metaphenomenon, i.e. by a pre-projected fact serving as Phenomenon, as in (the fact) that she is late worries me. The two types of sensing that can involve a Phenomenon of this metaphenomenal type are the ones that cannot project*, namely perception and emotion That is, while perception and emotion cannot create ideas, they can ‘react to’ facts. In this respect, they are like certain relational clauses such as (the fact) that she is late is a worry /worrying.
The fact [[that he had to diet]] made him wish —> he had bought a secret supply of chocolate.
The fact [[that the bailiff was coming on Monday]] made him think —> it might be time to emigrate.
While both cognition and desideration project ideas, they project ideas of different kinds. Cognition projects propositions — ideas about information that may or may not be valid: he believed/imagined/dreamt —> that the earth was flat. In contrast, desideration projects proposals — ideas about action that has not been actualised but whose actualisation is subject to desire: he wanted/intended/hoped for —> her to leave. Projection is the critical link between sensing and saying.
Cognitive and desiderative processing create ideas … but perceptive processing and emotive processing don’t … they are activated by [pre-projected] facts.
In contrast, perceptive and emotive types of sensing cannot project ideas into existence. That is, ideas do not arise as a result as a result of someone seeing, hearing, rejoicing, worrying, grieving or the like. However, these two types of sensing may accommodate pre-existing projections, i.e. facts, for instance:It assures me [[that I am as I think myself to be, that I am fixed, concrete]].
I was impressed, more or less at that point, by an intuition [[that he possessed a measure of sincerity the like of which I had never encountered]].
We heard [[that you kindly let rooms for gentlemen]].Thus ‘that I am fixed, concrete’ is construed as something already projected (hence we could add assures me of the fact that) and this fact brings about the emotion of assurance.
Sensing projects ideas into existence; the projection may take place either through cognition or through desideration, for example (from Pinter, The Birthday Party):I just thought —> I’d tell you that I’d appreciate it.I think —> I’ll give it up.They want —> me to crawl down on my bended knees.Thus the idea ‘I’ll give it up’ is created by the process of thinking; it does not exist prior to the beginning of that process. Similarly, the idea ‘me to crawl on my bended knees’ is brought into hypothetical existence by the process of wanting.
These can be treated as a subtype of material processes or as a borderline category between material and mental. They include conscious processing construed as active behaviour (watching listening, pondering, meditating) rather than just passive sensing (seeing, hearing, believing). Like the Senser in a mental clause, the ‘Behaver’ in a behavioural one is endowed with consciousness; whereas in other respects behavioural clauses are more like material ones. Like material clauses (but unlike mental ones), behavioural clauses can be probed with do: What are you doing? — I’m meditating but not I’m believing. Furthermore, behavioural clauses normally do not project, or project only in highly restricted ways (contrast mental: cognitive David believed —> the moon was a balloon with behavioural: David was meditating —> the moon was a balloon; nor can they accept a ‘fact’ serving as Phenomenon (mental: David saw that the others had already left but not behavioural: David watched that the others had already left). In these respects, behavioural processes are essentially part of the material world rather than the mental one. Many of them are in fact further removed from mental processes, being physiological rather than psychological in orientation.
As with all systems in language, any given instance will be more or less prototypical; and there may be subtypes lying immediately at the borderline of the primary types. The grammar construes the non-discreteness of our experience by creating borderline cases and blends. One such area is that of behavioural processes (Halliday 1985: 128-9): “processes of physiological and psychological behaviour, like breathing, dreaming, smiling, coughing”. […] Such borderline cases, in which the pattern of reactances does not conform exactly to that of a major type, are typical of grammatical systems in general.
Mental and verbal clauses are distinct from material and relational clauses in that the former can project ideas and locutions (quote or report). These represent the ‘content’ of sensing and saying, as in David thought —> the moon was a balloon, where the relation of projection is represented by an arrow. Verbal clauses are distinct from mental clauses in that the Sayer is not necessarily an entity endowed with consciousness; and in verbal clauses there may be a further participant, the Receiver, which is not found in a mental clause.** Except, of course, where sensing is construed as interior saying: he thought to himself.
Material clauses have a special pro-verb, do (to/with), as in what he did to the lawn was mow it. This does not occur in mental clauses: what he did to the story was believe it; nor in relational ones: what he did to the lawn-mower was have it.
In material clauses, the unmarked present tense is present-in-present (he is mowing the lawn; I’m doing the job), whereas with the other process types it is the simple present (mental: she believes he’s mowing the lawn; relational: he has a lawnmower; I’m busy).
In a mental clause, the Senser is endowed with consciousness … . This constraint does not apply to any of the participants in material and relational clauses. While the Senser is heavily restricted in this way, the other mental participant, the Phenomenon, is entirely unrestricted: it can be not only phenomenal (she remembered the old house) but also macro-phenomenal (act: she remembered him coming down the stairs) or meta-phenomenal (fact: she remembered that they had been happy in the old house). Participants in material clauses cannot be meta-phenomenal. For instance, while it is possible to demolish not only concrete things such as buildings but also abstract things such as ideas and arguments (she demolished the house / their ideas / his argument), it is not possible to demolish “meta-things” (we do not find she demolished that the earth was flat).
Many mental processes are typically bidirectional, appearing in two opposite configurations (I like it / it pleases me; cf. detest/revolt; fear/frighten; remember/remind, notice/strike). It is thus possible to construe conscious processing either as the Phenomenon impinging on the Senser’s consciousness (the music pleases him) or as the Senser’s consciousness having the Phenomenon as its domain (he likes the music). neither material nor relational* clauses display this dual directionality.
Medium
|
Process
|
Range
|
|
emanating:
|
Senser
|
Process: mental
|
Phenomenon
|
decoding:
|
Identified
|
Process: identifying
|
Identifier
|
Agent
|
Process
|
Medium
|
|
impinging:
|
Phenomenon
|
Process: mental
|
Senser
|
encoding:
|
Identifier
|
Process: identifying
|
Identified
|
… doing & happening are realised as material clauses, sensing as mental ones, saying as verbal ones, and being (at, etc.) & having as relational clauses. The different process types are not signalled overtly in the grammar; they are covert or cryptotypic categories and emerge only when we consider their reactances.
The complementarity can also be seen in the different kinds of deixis (relation to the here & now) associated with processes on the one hand and participants on the other. A process is made finite — it is pinned down in time, with point of reference in the act of speaking. A participant is made determinate, being held in a location within a referential space. The same distinction also appears in the temporal unfolding of a text, where participants have the potential to persist as discourse referents, but processes are excluded, unless they are turned into honorary participants through the use of grammatical metaphor.
Whatever the mode of occurrence of any figure, it will always unfold in time. This temporal unfolding is construed as an inherent property of the process itself, realised grammatically as tense and aspect; it thus serves to validate the distinction between process and participant. Whereas on the one hand in its manifestation as a process, the figure unfolds in time, in its manifestation as participant, on the other hand, it persists through time — whether or not the participant undergoes any change of state. The limiting case is a creative or destructive process, such as writing or erasing a symbol, through which a participant comes into being or ceases to exist.
Figures of doing and being can be interpreted as complementary perspectives on a ‘quantum of change’. Construed as doing, a change appears as a change in the thusness of a participant. Construed as being, the change appears as an achieved or attainable result. Consider a causal consequence such as [doing:] ‘he washed it’ —<so>—> [being:] ‘it was clean’. This quantum of change may be construed as two figures, as it is here (He washed it, so it was clean.). Alternatively, it may be construed as one figure, in which case it may adopt either point of view. If construed as doing, he washed it clean, the figure is elaborated with a result. If construed as being, he made it (be) clean, the figure is enhanced with an agentive Attributor. The wording he cleaned it embodies both perspectives in a single process.
Doing requires some input to occur. This will typically come from one of the participants involved, the doer (as with voluntary motion); but the source of energy may also be outside the figure (as with falling). No input of energy is required with a figure of being.
Doing involves a change over time of occurrence (including maintaining a state in spite of force for change). The change may take place along any one of a number of dimensions:
- (a) circumstantial: spatial (motion or disposition, concrete or abstract);
- (b) intensive: qualitative (colour, size, shape, solidity, etc.), quantitative (increase, decrease);
- (c) possessive (transfer of ownership, loss or accretion of parts);
- (d) existential (creation or destruction).
In contrast, being does not depend on change over time. As a figure of being unfolds over time, the only change is that embodied in the temporal unfolding of the process itself. The nature of of the actualisation will be the same at any point of time.
Doing and being do not preclude the involvement of a conscious participant; but they do not require it — and hence do not have the effect of endowing a participant with human-like consciousness. They can be differentiated in terms of two parameters, time and energy, both of which are involved in their actualisation.
There is of course a great deal of indeterminacy here, including such borderline cases as those where sensing and saying are construed as forms of action (and therefore cannot project), e.g. watching, listening, chatting, speaking. These properties reflect the borderline location of such processes in the overall semantic space.
The domain outside this conscious-semiotic [sensing, saying] centre of the ideational universe is then quintessentially either active (doing) or inert (being).
Saying is construed as proceeding from Sayer to Receiver (she asked/told/commanded him — ‘she addressed him’). In contrast, sensing embodies two complementary perspectives: either the Senser’s involvement in the sensing ranges over the Phenomenon or the Phenomenon is construed as impacting on the Senser’s consciousness (she likes the design : the design pleases her).
Figures of saying construe the addressee of exterior symbolic processing in the form of a participant, the Receiver, as in She told/asked/commanded him…; She said to him / asked of him…. In contrast, interior symbolic processing cannot be addressed; figures of sensing cannot be configured with a Receiver.
Sensing and saying construe the "Symboliser" along different lines. The interior Symboliser of sensing is construed as a participant engaged in conscious processing; hence it is endowed with consciousness by virtue of serving in a figure of sensing. … The “Symboliser” of a figure of saying often is a conscious speaker. However, since saying is exterior rather than interior symbolic processing, the Symboliser of saying, unlike that of sensing, is not restricted to human consciousness; it may also be any kind of symbol source, a ‘semiotic thing’ such as institutions, documents and instruments of measurement. … We recognise the difference between a Symboliser of sensing and a Symboliser of saying by calling them Senser and Sayer, respectively.
“Symbolic processing” is a generalisation across sensing and saying that foregrounds the fact that they can both project. But sensing and saying differ in the level of projection: sensing projects interior content, ideas; saying projects exterior content, locutions. The level of the projected content determines the typical status of the projected content: locutions may be either quoted or reported, with quoting being favoured in many types of discourse; in contrast, ideas are typically reported and only rarely quoted. That is, ideas are construed as being further removed [than locutions] from experience that is shared. Projection thus construes a distinction between interior symbolic processing (sensing) and exterior symbolic processing (saying).
… the projecting figure represents symbolic processing, processing that brings another figure into symbolic existence. Figures of symbolic processing involve the symbolic process itself (thinking, saying etc) and a participant engaged in the symbolic processing, as in ‘Symboliser’: she + ‘Process’: said/thought —> that he had left. The projected symbolic content is either a proposition (she said/thought —> he had left) or a proposal (she asked him —> to leave; she wanted —> him to leave).
Projecting sequences differentiate rather sharply between figures of sensing and saying, on the one hand, and figures of doing and being, on the other, by selecting figures of sensing and saying as the ones that have the special power of setting up other figures as second-order, semiotic reality. That is, projecting sequences construe figures of sensing and saying on two levels, the level of sensing/saying itself, and the level of the content of sensing/saying.
The system of figures construes experience as falling into four broadly conceived domains of goings-on: doing (including happening), sensing, saying and being (including having). Each type of figure has its own set of participant rôles.
There are two interlocking aspects of the configuration of a figure:
(i) the domain of experience to which the figure belongs; and
(ii) the nature of the interaction among its participants.
To put this another way: as a theory of experience, the semantic system of figures embodies two subtheories: one concerning different domains of experience and one concerning the ways in which participating phenomena can interact.
Compositionally, figures are phenomenal units that are formed by configurations of other phenomena (elements). Being “units” means that they are constituted as organic wholes with functionally distinct parts. In this respect they differ from sequences, which are not compositional units but loci of serial expansion and projection.
… a figure is a basic fragment of experience that embodies one quantum of change. As such, it is like a little drama — it is a constellation of actors and props; and it unfolds through time.
Sequences at lower ranks than that of figures retain their logical mode of realisation in the grammar. Expansion and projection thus flow throughout the system, forming sequences. This is in fact an instance of a general principle: expansion and projection are trans–phenomenal categories in the sense that they are manifested over the system as a whole — not merely in different logical environments across ranks, but also experientially. […] This feature is particularly exploited when the system is expanded through grammatical metaphor.
One question that arises here is the extent to which the semantics and the grammar are in phase with one another. It is conceivable that a semantic sequence of figures could be realised in the grammar by a single clause with complexes at group/word rank — i.e., the sequencing is downranked in the course of realisation. One argument in favour of exploring this possibility is that it would be possible to sort out the potential ambiguity of examples such as Henry and Anne went to the movies — grammatically, we have a group complex (Henry and Anne), but semantically, it could be either a sequence of figures […] or a sequence of elements […].
The limiting case is that where what is being expanded is not a figure but an element of a figure, in which case instead of a sequence of figures we get a sequence of elements, realised by group or phrase complexes. […] Putting this in terms of the grammar, the sequential relationships remain constant, but the rank at which the complex occurs depends on the domain being related.
To say that these relations are between figures means in principle that they hold between figures as a whole; that is, given a pair of related figures the domain of expansion or projection relation between them is each of the figures in its entirety. But in some instances, some subdomain may be particularly implicated. This is perhaps especially true with elaborating sequences — the grammar of hypotactic elaborating clause complexes tells us as much: the elaborating dependent clause includes a relative reference expression (if the clause is finite) and the clause is placed immediately after the domain that is being elaborated, whether that is the whole clause […] or some element within the clause […].
The grammar forms complexes at ranks below the clause, of course […]. It is always possible to interpret certain instances of these as sequences that have been ‘shrunk’ by the grammar because they share one or more elements. Thus Henry and Anne went to the store might be interpreted as a simple figure if they went to the store together but as a sequence if they went there separately.
In addition, one or more of the figures in the [sequence] may be realised by something ‘less than’ a clause. One major source of this is ideational grammatical metaphor […]. But there are also other cases such as circumstances of Rôle with temporal implications — for example: as a child, he was very shy ‘when he was a child, he was very shy’.
On the other hand, a clause complex may in principle correspond to a figure rather than a sequence. This happens through grammatical metaphor of an interpersonal kind: an interpersonal modality that would be realised congruently as a modal auxiliary (can, may, will &c.) or a modal adjunct (perhaps, probably &c.) is ‘upgraded’ to the status of projecting clause in a clause complex; for example:I don’t suppose || there’s very much ‘there is probably not’
On the one hand, a sequence may extend beyond a single clause complex […]. That is, the general potential is simultaneously semantic and grammatical; but in the creation of […] text, this potential may be taken up semantically to create a sequence that is more extensive than the clause complexes realising it. […]
In other words, once sequence and clause complex have become partly dissociated so that one sequence does not automatically imply one clause complex, the decision how to associate them in realisation becomes a meaningful, significant choice. (This significance may vary from one register to another, but the principle that the choice is meaningful is quite general.)
… semantics and grammar evolve together (in all three senses of semohistory …). In the present context that means that sequences and clause complexes evolve together. The basic principle is that a sequence is realised by a clause complex. But the two may become dissociated from one another.
We can generalise as follows: the logical resources for forming sequences have evolved in the environment of ideational meaning as sequences of figures. But these highly generalised resources can then also be applied in an interpersonal environment to form interactional sequences. (We should note here, however, that evidence from language development studies suggests that the logico-semantic relations are first construed in interpersonal contexts). In producing a text, we may use either or both, depending on the nature of the text.
As already noted, text organisation also draws on the interaction base; and there are certain parallels with the ideation base with respect to sequences. The interaction base certainly has text forming resources that are uniquely interpersonal. In particular, it has the resources for the collaborative exchanges that are embodied in the notion of interaction — for producing dialogue jointly by means of coordinated moves alternating between the interlocutors. But these interpersonal moves may also form sequences of moves in a way that is similar to the formation of sequences in the ideation base. Typical examples involve motivating condition (‘I invite you to accept x, if you want x’; for example: If you’re thirsty, there’s a beer in the fridge) and evidence (‘I think, infer / you should believe x because y’; for example: John’s in Germany because I just talked to him).
Many text types are heavily influenced by patterns in the meaning base — they can be seen as ‘macro-figures’, i.e. as expansions of figures by means of logico-semantic relations. This is not to say that that the relationship between organisation in the meaning base and the discoursal organisation is always one-to-one even when a text is organised according to an ideational sequence. In particular, a text may leave to be inferred certain steps that would be specified in the sequence in the meaning base (e.g. to make explicit the inferential processes involved).
Texts and sequences are of the same order of abstraction; both are semantic phenomena. A text is a piece of language that is functional in context. It draws on the ideational meaning base but it involves the full metafunctional spectrum; i.e. there are interpersonal and textual contributions as well. Since text draws on the ideational meaning base, sequences are one principle for organising text.
Sequences impose a certain order on our experience in terms of the relations that connect [one] happening with another. Hence sequences can be used to store information about the world in the form of organised text — ‘this is how to change tyres on your car’, ‘this is how to make cauliflower surprise’, etc.. Such texts often fall into a clearly recognisable text type, such as procedures, proofs, explanations, and episodic narratives. Not all texts are as highly regulated as these; but it is usually possible to make some prediction about the kinds of sequence, and the complexity to which sequences extend, in most our culturally recognised modes of discourse.
Thus sequences are formed by binary logico-semantic relations which may relate either single figures or sequences of figures [i.e. internal nesting]. Certain relations may impose constraints on the phenomena being related; for example, a projecting relation can only obtain when the first figure is one of sensing or saying. But the range of different logico-semantic relations is highly varied, so that constraints tend to be specific to particular subtypes.
In contrast to figures, sequences are not constructional units. We can specify the range of projecting and expanding relations available for further developing a sequence, but we cannot specify where a sequence has to come to an end — that is, we cannot specify a sequence as a unit whole with a conventional configuration of parts. Thus if we have expanded one figure we can always repeat the operation … . In contrast, a figure is a unit with a finite number of elements … . Hence a sequence can be indefinitely complex, whereas a figure cannot.
… the distinction between expansion and projection is less determinate than we have suggested. The logico-semantic relation of condition, which is prototypically construed as a form of enhancement, could also be construed as a kind of projection; and this is also brought out in the grammar. Conditions specify a potential and actualisable but non-actual situation. This potential situation can also be set up through projection:If the power supply fails, what’s the best thing to do?
Supposing the power supply fails, what’s the best thing to do?
Say the power supply fails, what’s the best thing to do?
Words such as supposing and assuming are verbs of projection which have come to function as conjunctions in conditional figures; while other words such as imagine and say retain more of their projecting force. Sometimes even variants of the same word have come to differ a little in their locations on this cline: for example, suppose and assume seem closer to projection than their corresponding participial variants. This is an uncertain region in which a figure hangs in the air, so to speak, suspended between the hypothetical material plane and the semiotic one.
Relations of expansion are typically realised in the grammar by conjunctions or conjunctive expressions linking a pair of clauses, either paratactically or hypotactically (e.g. that is, in other words; and, but, or, also, besides; so, yet, then, when, if, because, unless). Some of these may realise more than one category; for example, but may be adversative ‘and yet’ (extending), or concessive ‘and in spite of this’ (enhancing); while may be additive ‘and in addition’ (extending), or temporal ‘and at the same time’ (enhancing); or may be alternative ‘or else’ (extending: alternatives in the external world, like take it or leave it), or restating ‘in other words’ (elaborating: alternatives in the world ‘internal’ to the discourse, like they are reduced to the smallest size, or micro-miniaturised.). Overlaps of this kind show that the primary categories we have set up do in fact shade into one another; in particular, extending in some sense occupies a space intermediate between elaborating and enhancing, and shares a fuzzy borderline with each.
There is no sharp line between a figure and a sequence of figures: a quantum of experience is not defined before it is construed, and the grammar rather sets up a cline from sequence to simple figure. […] For example:
1 sequence of 2 figures: Tarzan pulled the rope; so it broke.
Tarzan made the rope break (by pulling it).
Tarzan broke the rope.
1 figure: The rope broke.
And there is a third type of accumulation where the logical relation is that of ‘equals’: x is the same figure as a. Here at this end the limiting case is a simple repetition; this may be further elaborated in such a way that one figure is reworded as another, or else further clarified or brought out by an example.
But figures may also be added to one another, making them part of the same story without assigning any kind of logical priority to either: x as well as / instead of / in contradistinction to a’. We have referred to these as extensions.