'Categorisation' In Ontogenesis

Halliday & Matthiessen (1999: 73):

Painter (1996) provides a key to the understanding of how linguistic resources are deployed in 'categorisation' in language development, drawing on her longitudinal case study of one child, Stephen, between 2 1/2 and 5 years.

The first stage in categorisation is naming individuals as members of classes; instances of the visual experience shared by the young child and his father or mother are ascribed to some general class of experience by means of a figure of being. At about 2 1/2, Stephen produced examples such as: Stephen (examining pattern on a rug): That's a square. What's that? — Mother: That's a circle. Here some perceived phenomenon of experience is brought "into intersubjective focus" by being referred to exophorically — pointing verbally, so to speak, to some feature of the material setting, sometimes accompanied by or replaced by a pointing as a gesture. This phenomenon is construed by Stephen as the Carrier of the figure of being, and is ascribed as a member of some general class of experience, construed as the Attribute of the figure. Stephen is 'importing' experience of instances into the semantic system by ascribing them to general classes in that system. This is an act of naming, and later this act itself gets named by call (see also Halliday, 1977, on calling as an early example of language being turned back on itself).

Children thus build up experience as meaning, in contexts such as the one exemplified above.

Ontogenesis: Generalising From Individual Names To Class Names

Halliday & Matthiessen (1999: 73):

The ideation base is thus a resource both for construing experience and for construing its own construal of experience. It has the potential for expanding itself precisely because it includes a theory of how meanings are construed When children begin to make the transition from proto-language into language (that is, when they begin to develop the system of the mother tongue, typically early in the second year of life), these resources for self-construal are not yet in place. The first things that are construed by naming are individuals; there is as yet no potential for taxonomies of general classes. But children soon take the critical step of generalising across individuals (see e.g. Halliday, 1993a, on generalisation). From a lexicogrammatical point of view, this means that naming has been generalised from individual names ("proper nouns") to class names ("common nouns").

Blogger Comments:

Viewed through the lens of SFL Theory, the shift from naming individuals to naming classes is the shift from assigning token-value relations (identity) to assigning token-type relations (ascription: instantiation); see Halliday & Matthiessen (1999: 145).

The Semantic System Includes A Theory Of How Meanings Are Construed

 Halliday & Matthiessen (1999: 72-3):

One central semantic motif that has emerged from these studies is that of elaboration: both in relational clauses (realising figures of being & having) and in complexes, especially nominal ones (the traditional notion of apposition, realising sequences of participants). Here elaboration is at work e.g. in 'distilling' new meanings over extended passages. That is, the semantic system includes a theory of how meanings are construed; and this theory is itself a resource for construing new meanings. Thus when a young child says cats also aren't people, he is using a type of figure of being & having to construe a taxonomic relationship in the ideation base, so that the membership of 'cats' (construed as Carrier) in the class of 'people' (construed as Attribute) can be probed, in this case to 'outclassify' cats from the class of people. Figures of being & having of the intensive and ascriptive kind thus construe, among other things, the taxonomic relation among classes or types in the semantic system.

The ideation base is thus a resource both for construing experience and for construing its own construal of experience. It has the potential for expanding itself precisely because it includes a theory of how meanings are construed

The Systemic-Functional Position On Categorisation

Halliday & Matthiessen (1999: 72):

From a systemic-functional point of view, the intellectual context was rather different. Systemic-functional theory was not oriented towards philosophy and logic, and the Aristotelian conception of a category did not figure as a traditional frame of reference that therefore had to be rejected. Already at the outset of the theoretical work that was to become systemic-functional theory, a major descriptive focus was on intonation (see e.g. Halliday, 1963a, b, 1967), a domain that clearly cannot be construed in Aristotelian terms; and the notion of cline was part of the theory already in the first major statement (see Halliday, 1961). Further, the system was conceived of as a probabilistic one even in proto-systemic work (see e.g. Halliday, 1956).

Systemic-functional work on 'categorisation' has thus not engaged with the philosophical tradition; nor has it tended to proceed by experimental methods. Rather, it has been concerned with how meaning is construed in naturally occurring text, e.g. in child language studies (Halliday, 1975 onwards; Painter, 1996) and in factual writing (e.g. Halliday & Martin, 1993; Harvey, in prep.). Here one of the questions has been what resources are available in the semantic system for construing new meanings — for 'category development', both in the ontogenetic time frame and in the logogenetic time frame.

The Cognitive Linguistic Position On Categorisation

Halliday & Matthiessen (1999: 72):

Categorisation has received much attention in recent work in linguistics, especially within the framework of cognitive science. The received tradition was the classical or Aristotelian conception of a category as a type definable in terms of necessary and sufficient conditions — its essential characteristics rather than its accidental ones. One central aim of cognitively oriented linguists has been to show that the Aristotelian conception does not apply to semantic categories, which have to be re-theorised in terms of prototype theory. The initial impetus came from a series of well-known experiments in categorisation by E. Rosch and by Labov (1973) and from Berlin & Kay's (1969) study of colour terms across languages; Wittgenstein's notion of family resemblances had provided an earlier insight within a philosophical frame of reference (cf. Ellis, 1993). Since there are now a number of summaries of these positions, and of the critique of the Aristotelian conception (e.g. Lakoff, 1987; Taylor, 1989), we will not review the discussion here.

The Relationship Between Typological And Topological Perspectives

Halliday & Matthiessen (1999: 70-1):

What is the relationship between these two perspectives on meaning? So far in work on semantics only the typological perspective has been able to be realised in a formal system; the power of the topological perspective still derives largely from the metaphor of space within theories of meaning. For present purposes, we can say that the topological organisation construes a semantic space, creating the correspondences shown in Table 2(4).

Both perspectives are valuable. The typological perspective allows us to gain insight into the organisation of meaning through the network, both as theoretical metaphor and as a system of formal representation. The topological perspective gives us complementary benefits — in the first instance, the general notion of a multidimensional elastic space. We have indicated in the rightmost column some implications that derive from adopting the topological perspective. The general motif here is that of indeterminacy. For example, we can show how regions of meaning overlap (e.g. doing & happening overlapping with sensing in an area of 'sensing as activity'). We will use informal topological diagrams at various points in our discussion to bring out this central feature of the ideational system.

The last row in the table above also represents indeterminacy, but this is indeterminacy of a particular kind, relating to the probabilistic nature of the semantic system. The types in the semantic system are instantiated according to probability values; these are manifested as relative frequencies in text. The equivalent in the spatial interpretation of meaning would be curvature or chreodisation. Chreodisation embodies time and represents the change of systemic probabilities over time (see e.g. Waddington, 1977, Sheldrake, 1988: Ch. 6, for discussion).

Blogger Comments:

Waddington modelled biological development as unfolding in an abstract space in which there were canals (curvatures) that represented the necessary path of development ('chreode'), which, crucially, the process of development could return to after being perturbed. Here Halliday & Matthiessen are only concerned with using space curvature as a means of representing probability differences topologically.

Topology Example: Content Plane

Halliday & Matthiessen (1999: 69, 71):

We can now consider a comparable example from the content plane of language — from the ideation base. In our description of sequences, we recognised two simultaneous systems: the relative status of the figures ('equal/ unequal') and the kind of relationship between them ('projecting/ expanding'). This account constitutes the typological perspective on this region of the ideation base; it is mapped onto the topological perspective in Figure 2-9. Sequences are thus construed as a two-dimensional region within the overall semantic space. The correspondences are the same as those already noted for the vowel space. 

Topology Example: Expression Plane

Halliday & Matthiessen (1999: 69, 70):

We shall relate typology and topology to one another as complementary perspectives on meaning, and then we shall say a few words about the value of keeping the topological perspective in view. Let us start with our analogy from the expression plane: a vowel system can be construed both typologically as a set of systems, e.g. two systems 'front/ back' & 'open/ closed' defining four vowel values, and topologically as a two-dimensional space with four focal (core, cardinal vowel) locations. These two perspectives are related to one another in Figure 2-8. 

Each system in the typology corresponds to a dimension in the topology. We can thus say that the two simultaneous systems correspond to a two-dimensional space. The systemic terms, or values, correspond to regions within the vowel space along one of its dimensions; an intersection of two systemic terms such as 'front* & 'open' is a region located along two dimensions. If we add further systems in the typology, e.g. rounding ('rounded/ unrounded'), nasality ('nasal/ non-nasal') and tongue root position ('neutral position/ advanced position'), these will correspond to further dimensions in the vowel space. With tongue root position we are still maintaining a reasonably congruent relationship between our representation of the vowel space and the oral cavity in which vowels are articulated, since advanced tongue root position is simply a global shift of the whole space; but with nasality we are beginning to use our representation more metaphorically, since the control of airflow through the nose is not a feature of the oral cavity.

Typological And Topological Perspectives On The Organisation Of Meaning

Halliday & Matthiessen (1999: 68-9):

So far we have sketched the most general organisation of the ideation base as a typology of semantic types or classes. This is one theoretical perspective on the organisation — a perspective realised in certain conventional systems of representation, such as the system network. It allows us to bring out quite clearly the global organisation of the ideation base as a resource for construing experience; and it also allows us to show how the semantic types in this global organisation are interrelated, as in the case of participants filling Actor and Goal rôles in figures of doing. 

At the same time, there is another theoretical perspective on the organisation of meaning — the topological perspective. Here meaning is construed in terms of a spatial metaphor: we can view the ideation base as an elastic, multidimensional semantic space. This metaphor is already familiar in discussions of meaning; we find it in Trier's notion of semantic fields, in the distinction between core meanings and more peripheral meanings, in specifications of semantic distance, and so on.

The notion of a vowel space (with its 'cardinal vowels') provides a familiar analogy. As a material construction, it is limited to the three dimensions of physical space; but as a physiological space, it accommodates variation along a number of dimensions, and brings an elasticity to the expression plane that is in some extent analogous to the metaphorical elasticity that we are ascribing to the plane of the content.

Domains Of Categorisation

Halliday & Matthiessen (1999: 68):

Naturally our categorisation tends to be oriented towards phenomena on a particular scale, those that lie within the bandwidth of those phenomena which are most readily accessible to our senses and which we engage with in day to day existence. This is the realm which impinges most closely on our physical, biological and social being; the semantic system has evolved with this as its primary semogenic environment. Lying beyond this are the micro- and the macro-worlds which are accessible to us only by instrument and by inference. 

It is a powerful demonstration of the potential of the semantic system that it readily fashions new meanings that model these experientially remote domains. What is less obvious, perhaps, is that it has had to be equally resourceful in modelling the ongoing changes in our social environment, where both the overall social order and our local interpersonal networks are constantly being modified and realigned.

Categorising Is A Creative Act

Halliday & Matthiessen (1999: 68):

Categorisation is often thought of as a process of classifying together phenomena that are inherently alike, the classes being as it were given to us by the nature of the experience itself. But this is not what really happens. Categorising is a creative act: it transforms our experience into meaning, and this means imposing a categorical order rather than putting labels on an order that is already there. As Ellis has expressed it, categorisation consists in grouping together sets of phenomena that are essentially not alike (1993: Ch. 3). There would be indefinitely many ways of construing analogies among different elements in the total flux of experience; what our semantic resources enable us to do is to construe those analogies which yield categories resonating with what as a species, and as members of a particular culture, we have found to carry material and symbolic value.

Semantic Types Are Categories To Which Phenomenological Instances Are Ascribed

Halliday & Matthiessen (1999: 66):

We have presented an initial sketch of the ideation base as a resource for construing our experience of the world around us and inside us. The focus of this sketch was the most general system of semantic types such as 'figure', 'being-&-having', 'participant', 'conscious being'. These semantic types are categories to which phenomenological instances are ascribed; they thus embody the fundamental principle of generalising across individual phenomenological variation. And they are located somewhere in delicacy between the most general type, the all-inclusive class of 'phenomenon', and the most delicate types we can recognise as being codified lexically in English.

General Options For Sequence, Figure And Element

Halliday & Matthiessen (1999: 65, 67):

Plate 10 below shows the lattice as described up to this point, incorporating all the categories introduced above (but still omitting those embodying any grammatical metaphor).

Element: Process

Halliday & Matthiessen (1999: 64-5):

Processes serve in the most central or nuclear rôle in a figure; they embody the temporal properties of a figure unfolding in time: see Plate 9. Other than metaphorical processes, the process element is either polar (positive/negative) or modal (some intermediate degree between positive and negative); it may embody phase, or aspect; and it will refer to past, present or future time. Polarity and modality derive from the interpersonal perspective on the process.

Element: Circumstance

Halliday & Matthiessen (1999: 63-4):

Circumstances fill circumstantial rôles in figures. We can recognise two simultaneous distinctions; see Plate 8. One concerns the type of circumstantial relation construed; the primary contrast is between circumstances of projection and circumstances of expansion and within the latter we distinguish those of elaboration, extension and enhancement The other concerns the experiential complexity of the circumstance; circumstances are either 'simple' or 'macro', the former being more truly elemental while the latter are more like figures. …

Among simple circumstances, the most usual are those of time, place, manner-quality and intensity, all of which are circumstances of enhancement. Examples: [skies will be partly cloudy] today (locative: time), increasingly (manner: intensity), widespread (locative: place), easily, carefully (manner: quality).

Macro circumstances are those which are made up of a special type of figure having another participant inside it, for example (circumstances of enhancement): (locative: place [abstract]) in the low to mid 60s, (locative: place) from the northeast, (manner: quality) at 15 to 25 m.p.h., (locative: place) throughout the northern Rockies, (locative: place) in a casserole, (locative: place) in a hot oven, (extent: duration) for 10-15 minutes, (manner: means) with a clean absorbent cloth.

Element: Participant: Simple Quality

Halliday & Matthiessen (1999: 62-3):

Qualities characterise things along various parameters, as in green cabbage : red cabbage: see Plate 7.

Again, a number of simple qualities are metaphorical; of the remainder, the "ordinary" qualities, one subtype is qualities of projection and the other is qualities of expansion. (As we will show in subsequent chapters, the categories of projection and expansion are very prevalent in the organisation of the ideation base.)

Element: Participant: Simple Thing

Halliday & Matthiessen (1999: 60-1):

Some 'simple things' are metaphorical; the remainder are referred to as 'ordinary', and these are either conscious or non-conscious (this is the distinction that is actually made in the semantic system, not animate/inanimate or human/non-human): see Plate 6.

Non-conscious ordinary things are distinguished along more than one dimension, but the categorisation given here can be taken as primary, in the sense that it is the one that seems to have the clearest reactances in the grammar.

Blogger Comments:

When we construe any organism as being able to see, for example, we are construing it as conscious, since perception is a type of mental process.

Element: Participant

Halliday & Matthiessen (1999: 59-60):

Participant rôles in figures are filled by elements of the type 'participant'; they are phenomena capable of taking on a participant rôle in a process configuration, e.g. bringing it about or being affected by it. They are further differentiated according to two parameters: see Plate 5. Macro-participants are all metaphorical and will be left out of consideration for the time being. Simple participants may be things or qualities.

Types Of Element

Halliday & Matthiessen (1999: 58-9):

As we have seen, elements fill the roles of figures. Participant roles are filled by participants (things or qualities), circumstance roles by circumstances (times, places, causes, etc.), and the process role by a process. There are correlations here between the taxonomy of configurational phenomena and that of simple phenomena … . 

The elements of a figure are of three kinds:

(i) the process itself (action/event, process of consciousness, or relation),
(ii) a participant in that process, or
(iii) a circumstantial element or circumstance. …

As already noted above, processes are realised by verbal groups, participants by nominal groups, and circumstances by adverbial groups or prepositional phrases. In addition to the three types of element that serve in figures, there is one further type of element — the relator: see Plate 4. Relators serve to construe logico-semantic relations of expansion between figures in a sequence; they are realised by conjunction groups.

The Rôle Restrictions For Sensing, Saying And Doing Summarised

Halliday & Matthiessen (1999: 57-8):

The rôle restrictions for sensing, saying, and doing are summarised in Table 2(1) below.

These rôle restrictions represent a kind of metaphysics of English transitivity. For example, according to English, ideas and locutions cannot act on things, but there is no general restriction on what kinds of things may act on other things. Not only persons, but also inanimate things and abstractions may kill people (a figure of doing): the rifleman/the rock/his stupidity killed cousin Henry. 

In contrast, if we had built our model according to the demands of the transitivity grammar of Navajo, we would have had to rank things in terms of their capacity to act upon other things; e.g., an inanimate cannot act upon an animate thing (Witherspoon, 1977).

Subtypes Of Figures And Participant Role Value Restrictions

Halliday & Matthiessen (1999: 57-8):

As already mentioned, there is a typology of figures based largely on the nature of the particular types of the process, participant, and circumstance roles; the most general part of this is shown in Figure 2-6.

 

Each type of figure has its own set of more delicately specified roles with particular value restrictions. For example, figures of doing have an Actor, which in turn is a participant (rather than some other kind of phenomenon). (Notice that the specification of the filler of the Phenomenon role in a sensing figure is more general than 'participant' — any type of phenomenon can be sensed.)

Grammatical Realisations Of Process, Participant And Circumstance

Halliday & Matthiessen (1999: 54-5):

Grammatically, the nuclear process, its participants, and its circumstances are typically represented as constituents in the transitivity structure of a clause: see Figure 2-3.

The Principle Of Organisation Of A Figure (vs Sequence)

Halliday & Matthiessen (1999: 53-4):

The principle of organisation of a figure is different from that of a sequence. As we have seen, a sequence is constructed by interdependency relations of expansion and projection. In contrast, a figure is constructed as an organic configuration of parts. Each part stands in a specific relation to the figure as a whole. The parts of a given configuration are (i) a nuclear process, (ii) one to three participants of different kinds taking part in the process, and (iii) up to around seven circumstances of different kinds associated with it.

Participants are inherent in the process; they bring about its occurrence or mediate it. There are a number of specific ways in which a participant may take part in a process; it may act out the process, it may sense it, it may receive it, it may be affected by it, it may say it, and so on. The different configurations of participants are the bases for a typology of process types. The distinction between participants and circumstances is a cline rather than a sharp division, but it is semantically quite significant.

Circumstances are typically less closely associated with the process and are usually not inherent in it. They specify the spatial or temporal location of the process, its extent in space or time (distance or duration), its cause, the manner of its occurrence, and so on.

Types Of Figure

Halliday & Matthiessen (1999: 52-3):

A figure is a representation of experience in the form of a configuration, consisting of a process, participants taking pan in this process and associated circumstances. There are, of course, indefinitely many kinds of process in the non-semiotic world; but these are construed semiotically, according to the way in which they configure participants, into a small number of process types — being, doing, sensing, and saying. The first three of these have clearly defined subcategories: see Plate 3.

Then, each figure may be either projected (by another figure, or in some other way) or else not; and if projected, it may be an idea or locution or [fact] …

Types Of Sequence

Halliday & Matthiessen (1999: 50-1):

A sequence is a series of related figures. Consequently, sequences are differentiated according to the kinds of relations figures can enter into — temporal (x happened, then y happened, etc.), causal (x happened, so y happened, etc.), and so on; in the most general terms, the relations yield the following type of sequence: see Plate 2.

In any pair of figures related in sequence, one figure may (i) expand the other, by reiterating it, adding to it or qualifying it; or (ii) project (report, quote) the other by saying it or thinking it. In either case, the two may be either equal or unequal in status, or semantic weight. … Sequences are organised by interdependence relations and they are indefinitely expandable.

The Congruent Realisation Of Ideational Semantics In Lexicogrammar

Halliday & Matthiessen (1999: 48, 49):

While figures are said to consist of elements and sequences are said to consist of figures, the 'consist-of' relation is not the same: elements are constituent parts of figures, functioning in different roles; but figures form sequences through interdependency relations. … The typical representation of sequences, figures and elements in the grammar is as in Figure 2-1.

Ideational Semantics: The Phenomena Of Human Experience As Three Orders Of Complexity

Halliday & Matthiessen (1999: 48, 49):

A phenomenon is the most general experiential category — anything that can be construed as part of human experience. The phenomena of experience are of three orders of complexity: elementary (a single element), configurational (configuration of elements, i.e. a figure) and complex (a complex of figures, i.e. a sequence) — see Plate 1.

Four Representational Challenges For SFL Metalanguage

Halliday & Matthiessen (1999: 45):

We shall mention four such challenges briefly here.

(i) The first representational challenge is the need to handle the dimension of instantiation.  

As a process, instantiation can be represented as involving traversal of the system network and activation of realisation statements. The instance is thus a set of features (semantic types) selected, with associated realisational specifications — an instantial pattern over the semantic potential. 

However, instantiation also defines a scale between the potential and the instance, with intermediate patterns of instantiation. We will introduce these later as register-specific domain models. Such patterns of instantiation will need to be represented: perhaps as systemic probabilities, or as domain-partitions within the overall ideation base.

(ii) The second representational challenge is the need to model how the overall ideation base is expanded by grammatical metaphor. It must be shown how metaphor adds junctional types to the ordinary types …

(iii) The third representational challenge comes from outside the account of the ideation base itself: the ideation base has to be related to the other metafunctional modes of meaning, the interaction base and the text base. The interaction base will include alternative 'projections' of the ideation base to account for the relationship between speaker and addressee. …

(iv) The fourth representational challenge concerns the non-discreteness of the various systems that construe semantic space. The semantic types represented by the features of systems in the system network do not constitute discrete Aristotelian categories; they are values on semantic clines — core regions, to use the metaphor of semantic space. We can bring this out by adopting a topological view on meaning; we can also explore the possibility of interpreting features as names of fuzzy sets.

Representation Stratum Of SFL Metalanguage Exemplified

Halliday & Matthiessen (1999: 44):

Figure 1-12 shows an example of a systemic representation, with the graphic conventions for the system network. Preselection is shown by means of a pointer leading from the function (role) being restricted to the feature that it is restricted to.

Representation Stratum Of SFL Metalanguage: Types Of Realisation Statement

Halliday & Matthiessen (1999: 43-4):

The general form of a realisation statement is 'realisation operator + one or more realisation operands', as in 'insert Senser' and 'conflate Medium and Senser'. The operators used to specify functional structuring are insert, conflate, and preselect. The first operand is always a semantic function (role); additional operands may be functions or features. The realisation statements we will make use of in our representation of ideation base information are as follows:

(1) Presence of functions in the structure: the presence of a function in a function structure is specified by inserting the function into the structure; the operation of insertion is symbolised by '+'; e.g. +Actor, +Senser, etc.

(2) Conflation of one function with another: one function from one perspective is conflated with a function from another perspective — they are identified with one another. Conflation is symbolised by '/'; for example, Medium/Senser means that Medium (ergative perspective) and Senser (transitive perspective) apply to the same element of a figure.

(3) Restriction on the type of phenomenon that can serve a particular function: this is stated by preselecting one or more features from the unit serving that function; preselection is symbolised by ':', e.g. Senser: conscious being, which means that the participant serving as Senser is restricted to the type 'conscious being'. (Since 'conscious being' is more delicate than 'participant', the fact that a participant serves as Senser can be inferred from the statement 'Senser conscious being'. In general, it is only necessary to state the most delicate or specific restriction along any subsumption path in the system network.)

Representation Stratum Of SFL Metalanguage: The Realisation Statement

Halliday & Matthiessen (1999: 42-3):

If we use systemic representation to encode information in the ideation base, we factor out the logic of subsumption and alternatives from the syntagmatic realisation: this logic is represented in the system network. If each feature is interpreted as a type, this gives us a lattice of types. In European structuralist terms, the network is a representation of the paradigmatic organisation of a linguistic system. However, it is also necessary to show how feature choices are realised structurally; i.e., what are the structural properties of the types. This is the province of the realisation statement.

A realisation statement is a minimal specification of a piece of structure or configuration of roles presented in a paradigmatic context; it is always associated with a particular systemic feature. For instance, the realisation statement '+ Senser' occurs in the context of 'sensing' in the system of figures; it is the syntagmatic realisation of that feature: a figure of sensing is a configuration of roles one of which is Senser.

Representation Stratum Of SFL Metalanguage: The System Network

Halliday & Matthiessen (1999: 41-2):

Systemic representations include (i) system networks and (ii) realisation statements.

(i) A system network is an acyclic directed graph, consisting of systems partially ordered in delicacy. Each system constitutes a choice (alternation, opposition) between two or more terms. These terms are represented by features, and a system as a whole is a Boolean combination of features:

(1) It has an entry condition, the condition under which the systemic choice is available. The entry condition may be a single feature or a complex of features, conjunct and/ or disjunct. These features must serve as terms in other systems.

(2) It has a set of terms, the options that are available given the entry condition. The terms are represented by features, which are related by exclusive disjunction. Collectively, a set of related systems form a system network (since features in the entry conditions to systems must be terms in other systems).

Metalanguage: Representation Stratum

Halliday & Matthiessen (1999: 33-4):

The theoretical model makes considerable demands on the representational system or systems in which it is realised. Such a system must be able to handle the axial differentiation of the meaning base so that it represents not only structural configurations but also variation in delicacy, the indeterminacy between options in meaning, the move in instantiation from potential to instance, and other theoretical specifications. …

As we have already noted, the representational semiotic may also be language itself — the theory may be represented discursively in the register(s) of linguistics. In this case, the relationship between theory and language is similar to … a relationship where theory might be construed as a connotative semiotic (in Hjelmslev's, 1943, conception: a semiotic system whose expression plane is a semiotic system) …

Finally, the representational semiotic may also be a diagrammatic one — e.g. network graphs, tree diagrams, our circle diagrams. With such graphic representation, it is important to ask (i) if the information represented graphically can be restated in some other form of representation and (ii) if its realisational relationship with respect to theory is clear. There is a certain danger that the graphic representation is simply assumed to be natural and transparent even though it depends as much on semiotic conventions as any other system. Our own experience with circle diagrams is that, unless the representational convention has been made explicit, readers favour certain types of meaning: the circles are interpreted in terms of 'extension' rather than in terms of 'elaboration' — which is to say, they are read in constituency terms rather than in realisational terms. In any case, diagrams will only serve us as 'visualisations' as long as they construe a metaphor of abstract space at the theoretical stratum.

Metalanguage: Theoretical Construal Stratum

Halliday & Matthiessen (1999: 32-3):

Thus any account of the ideation base has to be metalinguistically stratified It has to be constructed as a theoretical model out of the resources the theory provides and according to the constraints imposed by these resources. From a systemic-functional point of view, this means that the ideation base is construed as a multidimensional, elastic semantic space. This space is organised as a meaning potential, with an extensive system of semantic alternatives; these alternatives are ordered in delicacy. Each set of alternatives is a cline in semantic space rather than a set of discrete categories, and any alternative may be constituted structurally as a configuration of semantic roles. The meaning potential is thus differentiated axially into (i) systems of options in meaning and (ii) structural configurations of roles by which these options are constituted.

The meaning potential itself is one pole on the dimension of instantiation: it is instantiated in the unfolding of text, with patterns of typical instantiation (specific domains of meaning) lying somewhere in between the potential and the instance. At the same time, this overall ideation base can be expanded by various semogenic strategies, among which we are foregrounding that of grammatical metaphor.

The Stratification Of Metalanguage

Halliday & Matthiessen (1999: 30):

Metalanguage has the same basic properties as any semiotic system. This means that it is stratified. It construes language in abstract theoretical terms; but this construal is in turn realised as some form of representation — either language itself, in discursive constructions of theory, or some form of designed semiotic (system networks, constituency rules, conceptual networks, logical formulae, and so on).

Metalanguage

Halliday & Matthiessen (1999: 30):

Johnston's insight into the power of representation embodied in Sign is central also to the general challenge of representation in metalanguage. We noted above that the semantics/ lexicogrammar of natural language is itself a 'realisation' (an abstract construction) of daily experience. Likewise, the system we use to explore the semantics/ lexicogrammar — our theory of semantics and our grammatics — is a 'realisation' of that part of daily experience that is constituted by semantics and lexicogrammar; that is, it is an abstract construction of language. This system is itself a semiotic one — a metalanguage; in Firth's more everyday terms, it is language turned back on itself. So whereas a language is (from an ideational point of view) a resource for construing our experience of the world, a metalanguage is a resource for construing our experience of language.

Blogger Comments:

Here Matthiessen confuses 'a realisation' with 'a construal' (an abstract construction). To be clear, a realisation is the Token of an identifying relation; that is, it is less abstract than the Value it signifies. On the other hand, a construal is an intellectual construction that is the Value of an identifying relation; that is, it is more abstract than the Token that signifies it. The first two sentences above can be made consistent with the final sentence above by re-expressing them as:

We noted above that the semantics/ lexicogrammar of natural language is itself a 'construal' (an abstract construction) of daily experience. Likewise, the system we use to explore the semantics/ lexicogrammar — our theory of semantics and our grammatics — is a 'construal' of that part of daily experience that is constituted by semantics and lexicogrammar; that is, it is an abstract construction of language.

Language is a construal of daily experience, not a realisation of it.

The Parallelism Of The Semantics And Spatial Expression Of Sign Languages

 Halliday & Matthiessen (1999: 30):

This parallelism is even more foregrounded where the modality of expression is spatial, as in the Sign Languages of deaf communities (see Johnston, 1989; 1992 on AUSLAN). Here the domain of expression is a 4-dimensional signing space-time in a field of perception shared by signer and addressee (though clearly perceived from different angles). The spatial orientation and the shared perception increase the potential for iconicity in the expression; Johnston (1992) points out: "Despite an oral-aural language being suited to iconically encode sounds, the fact that our experience as a whole is visual, temporal and spatial means that a language which has itself visual and temporal resources for representation has greater means than an auditory one to map onto itself those very visual and spatial qualities of the world it wishes to represent."

The Parallelism Of Semantics And Phonology

Halliday & Matthiessen (1999: 30):

The phonological representations are still abstract; they have to be manifested in bodily movements — in the ongoing movement of the parameters of the articulatory system. The sound system thus categorises bodily processes; and in this respect, it is similar to the semantic system: both are ways of construing human experience. Meaning is thus represented by modes of organisation that are similar to its own.

The Relation Between Content And Expression

Halliday & Matthiessen (1999: 29-30):

Grammatical representations are in turn represented in linguistic expressions — prototypically, in sounding. Here the relationship is more complex than it is between semantics and grammar, in that it is both natural and conventional. 

In the interpersonal and textual domains of content, it is often natural: thus interpersonal content tends to be represented prosodically by movements or variant levels in pitch, and textual content tends to be represented by prominence achieved phonologically (e.g. by the major pitch movement in an intonation contour) or sequentially (e.g. by using distance from initial position in the clause as a scale of prominence). 

In the ideational domain, the representation is usually conventional; but, even here there is a relationship of analogy, where we find in the sounding modes of organisation similar to those of wording (and therefore of meaning). 

Systemically, we find that the system construes a phonetic space — notably the vowel space; and that this provides a model for semantic space. 

Structurally, we find that sound is structured both as chains of segments (e.g. rhythmic units interpreted as syllable complexes) and as configurations of segmental constituents (e.g. syllables interpreted as configurations of phonemes).

Blogger Comments:

This is clearly Matthiessen, rather than Halliday. For Halliday, the natural relation between semantics and grammar is between meaning (e.g. participant) and grammatical form (e.g. nominal group). To be consistent with Halliday's formulation, the relation between grammar and phonology is the conventional relation between grammatical form (clause, group, etc.) and phonology (tone group, foot, etc.), where relations can not be said to be natural. (If the relations were natural, the phonological expressions of all languages would considerably more similar than they are.)

However, Matthiessen switches track in this exposition from the relation between the grammatical stratum and the phonological stratum to the relation between the content plane and the expression plane, where relations between interpersonal and textual meanings and prosodic phonology can be said to be natural.

The Natural (Non-Arbitrary) Relation Of Semantics To Grammatical Form

Halliday & Matthiessen (1999: 29):

… the realisation in lexicogrammar, is natural, in the sense of being nonarbitrary: for example, the grammatical constituency structure of a clause provides a natural representation of the semantic configuration of a process, participants and circumstances. By attending to grammatical representations, we can thus learn a good deal about the more abstract organisation of meaning at the higher stratum of semantics. We can learn about the different modes of meaning — logical, experiential, interpersonal, and textual — by exploring their different modes of representation in the grammar — chaining, constituency, prosody, and wave. Grammar is thus a hybrid system for representing meaning in the sense of embodying different modes of representation; but it is this that allows it to maintain a natural relationship with respect to semantics, with each mode of representation realising a different mode of meaning.

The Paradigmatic Orientation Of SFL Theory

Halliday & Matthiessen (1999: 28):

A functional grammatics thus allows us to approach semantics from a deeper and more wide-angled perspective. To this general property, systemic functional grammar adds another characteristic— its paradigmatic orientation. For instance, while more formally oriented accounts may approach transitivity patterns essentially in terms of sequences of grammatical classes such as 'nominal group + verb (+ nominal group)' and speak of classes of verb followed by one nominal group ('mono-transitive') or two nominal groups ('di-transitive'), a systemic grammar interprets such sequences in terms of systems of distinct and contrasting process types.

The Gateway To Semantics Is The Clause Rather Than The Word

Halliday & Matthiessen (1999: 28):

The gateway to semantics is the clause rather than the word. Consequently, grammatical categories will typically be interpreted 'from above', within their context in the clause or the group, rather than 'from below' within their context in the word. This has rather far-reaching consequences for the understanding of the semantic systems realised by the grammar. Systems that are approached 'from above' in this way include:

projection — clause complex: traditionally a form of 'subordination' within clause; reinterpreted as distinction between hypotaxis in clause complex vs. rankshift in clause, laying the foundation for a semantic distinction between reports and facts.

transitivity — clause: traditionally a word category, transitive = verb taking object/ intransitive = verb not taking object; reinterpreted as (i) process types (material/ mental/ verbal/ relational) and (ii) an ergative system (middle/ effective) in the clause.

tense — group: traditionally a mixture, because the model was taken over from Latin with richer word-rank realisations than English, but more recently in this century often a word category, past/ non-past; reinterpreted (relative to this) as (i) past/ present/ future and (ii) recursive, with secondary tense.

Taking the clause as starting point facilitates the exploration of cryptotypes: the chain of realisation often starts cryptotypically in the clause, whereas the final stages of realisation at word and morpheme rank are more oven — although, as noted in connection with tense and number, the oven marking is seldom the only factor involved.

Reactance

Halliday & Matthiessen (1999: 27-8):

The concept of reactance is particularly significant for our purposes where it involves a relationship between an ideational category and features of other metafunctions, interpersonal or textual. For instance, the interpersonal grammar provides for participants, within the ideational dimension of the clause, to function as Subjects; but this potential is not in general open to circumstances, and this is a principal reason for distinguishing these two classes within the ideational metafunction. Among reactances from the interpersonal and textual components of the grammar, we could mention the following:

interpersonal:

can/ cannot serve as Subject

can/ cannot serve as 'focus' of alternative question

can/ cannot serve as Wh element

textual:

can/ cannot serve as Theme

can/ cannot serve as 'focus' of theme predication (it is... that...)

can/ cannot be presumed by substitution/ ellipsis

Examples Of Cryptotypes (Cryptoclasses And Cryptosystems)

Halliday & Matthiessen (1999: 27):

There are many examples of cryptotypes in this sense, both as classes and as systems (i.e., cryptoclasses and cryptosystems), in our ideational semantics. For example:

process types: doing & happening/ sensing/ saying/ being & having
transitivity model: ergative/ transitive
projections: locutions/ ideas
expansions: elaboration/ extension/ enhancement
number: plural/ non-plural; singular/ non-singular

Overt Categories (Phenotypes) vs Covert Categories (Cryptotypes)

Halliday & Matthiessen (1999: 26-7):

The understanding of covert categories in the grammar is due to Whorf (1956: 88 ff), who made the distinction between overt categories or phenotypes and covert categories or cryptotypes; he is worth quoting at some length:

An overt category is a category having a formal mark which is present (with only infrequent exceptions) in every sentence containing a member of the category. The mark need not be part of the same word to which the category may be said to be attached in a paradigmatic sense; i.e. it need not be a suffix, prefix, vowel change, or other 'inflection', but may be a detached word or a certain patterning of the whole sentence. ...

A covert category is marked, whether morphemically or by sentence pattern, only in certain types of sentence and not in every sentence in which a word or element belonging to the category occurs. The class membership of the word is not apparent until there is a question of using it or referring to it in one of these special types of sentence, and then we find that this word belongs to a class requiring some sort of distinctive treatment, which may even be the negative treatment of excluding that type of sentence. This distinctive treatment we may call the reactance of the category. ... 

A covert category may also be termed a cryptotype, a name which calls attention to the rather hidden, cryptic nature of such word-groups, especially when they are not strongly contrasted in idea, nor marked by frequently occurring reactances such as pronouns. They easily escape notice and may be hard to define, and yet may have profound influence on linguistic behaviour. ... Names of countries and cities in English form a cryptotype with the reactance that they are not referred to by personal pronouns as objects of the prepositions 'in, at, to, from'. We can say 'I live in Boston' but not 'That's Boston — I live in it'.

Traditional Grammar vs Functional Grammar

Halliday & Matthiessen (1999: 26 , 28):

In traditional grammar, only certain grammatical categories were taken into consideration; these categories were (i) overt and (ii) word-based. In particular, inflectional categories of the word such as tense, case, and number were described and then interpreted semantically. In a functional grammar, while such categories are not ignored, they tend to play a less significant role, appearing at the end point of realisational chains. 

For instance, it is not possible to base a functional interpretation of number in English simply on the presence or absence of 'plural' as a nominal suffix (as in grammar+s); the category of number is rather more complex, involving two complementary systems (see Halliday, 1985: 161-2). 

Similarly, the general properties of the construal of time embodied in the English tense system are not revealed by only looking at the overt suffixal past tense marker- (as in laugh+ed); again the scope of the semantics of tense in English is far greater than this oven word category would suggest (see Halliday, 1985: 182-4; Matthiessen, 1996). In general our move into semantics from grammar differs from the traditional one along the following lines.

(i) We consider not only overt categories but also covert ones. …

(ii) The gateway to semantics is the clause rather than the word.

Why Semantics Is Not A Relabelling Of Lexicogrammar

Halliday & Matthiessen (1999: 26):

Thus when we move from the lexicogrammar into the semantics, as we are doing here, we are not simply relabelling everything in a new terminological guise. We shall stress the fundamental relationship between (say) clause complex in the grammar and sequence in the semantics, precisely because the two originate as one: a theory of logical relationships between processes. But, as we have shown, what makes such a theory (i.e. an ideation base as the construal of experience) possible is that it is a stratal construction that can also be deconstructed, every such occasion being a gateway to the creation of further meanings which reconstrue in new and divergent ways. Thus a sequence is not 'the same thing as' a clause complex; if it was, language would not be a dynamic open system of the kind that it is. This issue will be foregrounded particularly in our discussion of grammatical metaphor.

Every Scientific Theory Is A Stratal-Semiotic System

Halliday & Matthiessen (1999: 25):

Every scientific theory is itself a stratal-semiotic system, in which the relation among the different levels of abstraction is one of realisation. This is to be expected, since all such theories are modelled on natural language in the first place; and, as we have seen, the semantics of natural language is itself a theory of daily experience.

Realisation

Halliday & Matthiessen (1999: 25):

We have retained the term 'realisation' to refer to the interstratal relationship between the semantics and the lexicogrammar: the lexicogrammar 'realises' the semantics, the semantics 'is realised by' the grammar. … In any strata! system (i.e. any system where there are two strata such that one is the realisation of the other) there is no temporal or causal ordering between the strata. It makes no sense to ask which comes first or which causes which. That would be like taking an expression such as x = 2 and asking which existed first, the x or the 2, or which caused the other to come into being (it is not like the chicken and the egg, which are temporally ordered even though in a cycle). There is a sense in which realisation is the analogue, in semiotic systems, of cause-&-effect in physical systems; but it is a relationship among levels of meaning and not among sequences of events, … the relationship is an intensive one, not a causal circumstantial one.

Blogger Comments:

This unexplained analogy is potentially misleading, given that the identifying relation between strata is intensive (elaborating), not circumstantial (enhancing: causal).

One possible explanation is that, as an identifying relation, if semantic values are encoded by reference to lexicogrammatical tokens, the lexicogrammar serves as the agent (causal participant) of the identifying process, with semantics as the medium through which the process unfolds. And this is essentially Halliday & Matthiessen's theoretical method: construing a semantics by reference to the lexicogrammar.

However, whether this is what Halliday & Matthiessen mean by this unexplained analogy is anyone's guess.

Semantics Unfolds By Semogenic Principles In Three Semohistories

Halliday & Matthiessen (1999: 25):

Thus an interpretation of semantics must account not merely for the system at some particular point in its evolution but also for the processes by which it got there and the changes that will shape it in the future. As far as text is concerned, the changes in semantic styling that take place in the course of a text cannot be dismissed as simply ad hoc devices for making the text shorter (or longer!), more interesting or whatever, they should be seen as the operation of general semogenic principles in the specific context which is engendering and being engendered by that text.

The Interaction Between Recycling Meanings And Constructing New Ones

Halliday & Matthiessen (1999: 24):

Of course, not every instance of the use of language involves the creation of new meanings. The greater part of most discourse consists of wordings which have been constructed on countless previous occasions — in the language, in the individual, and even in the course of the text. When we come across the sentence Rain is expected in the northern part of the region, falling as snow over high ground we recognise probably all of it as something that is ready to hand: not only has it occurred in the English language many times before, but the same writer has probably written it many times before, and many of these instances could be seen as forming part of the same discourse (that is, day-by-day weather reports in a sense constitute one continuous text). The storing of meanings for repetitive use and reuse is just as important as the potential for creating new ones. 

The production of discourse by an individual speaker or writer can be seen as a dialectic between these two semiotic activities: between 

(i) recycling elements, figures and sequences that that individual has used many times before, and so for him or her are already fully codified, and  

(ii) constructing new ones that are being codified for the first time (and some of which may remain codified for future use — especially with a child who is learning the system).

Codifying As Generalisation: Syntagmatic Compacting vs Paradigmatic Condensing

Halliday & Matthiessen (1999: 24):

In general the process of creating meaning involves constructing some kind of lexicogrammatical generalisation — some form of wording that is in some respect unique. It is not possible to quantify the degree to which any semantic feature or domain has been codified at any one moment in semohistory; but meanings that are more highly codified are those that have been to a greater extent condensed and/or compacted, where 'compacting' is generalising on the syntagmatic axis (e.g. animal that has four legs > quadruped), while 'condensing' is generalising on the paradigmatic axis (forming into a system at some point along the scale of delicacy). The evolution of language (i.e. of specific languages in their various registers), the learning of language by children, and the production of language in the form of discourse constitute the historical contexts in which meanings are continuously being created along these lines.

Codifying: From Grammaticalisation To Lexicalisation

Halliday & Matthiessen (1999: 23-4):

This process of codifying may take place at any point along the cline from grammar to lexis, from grammaticalisation at one end (cf. Hopper & Traugott, 1993) to lexicalisation at the other. Perhaps the most highly coded meanings are those which are fully grammaticalised: that is, organised into grammatical systems, such as tense and polarity in English. This does not mean that they must be overtly signalled in syntax or morphology; some of them are, but others are uncovered only through systematic analysis, such as the different types of process configurations in English.

Lexicalisation may take the form of the instantaneous creation of new lexicalised meanings; like sputnik in 1958 or gazumping sometime in the seventies. But more often it is the end point of a process of lexical compacting, as in the example of quadruped above. Since lexicalised meanings do not form clearly defined and bounded systems in the way that grammaticalised ones do, we might consider meanings of this kind less highly codified, although the process of codification is the same in both cases.

Somewhere between the two extremes of grammar and lexis we may recognise the emergence of distinct grammatical structures and lexical classes. In the course of the history of English the meaning 'it is precipitating' became highly codified, in that types of precipitation came to be lexicalised as verbs (rain, hail snow, sleet, thunder, lighten) in a unique class having no participants associated with it e.g., it's raining, where the it functions as Subject but has no role in transitivity. (Note humorous back-formation on model of Actor-Process: What's it doing ?—Raining.)

Codifying Example

Halliday & Matthiessen (1999: 23):

Consider a series of examples such as the following:

What happens here is that a meaning has gradually crystallised, as it were, out of the total meaning potential of the system so that it can be deployed in codified form instead of being constructed afresh each time. In an animal that has four legs, each of the component elements animal four & legs is codified separately, as are the various grammatical relations involved; but the complex is not codified as a whole. When we come to quadruped, it is. Again, this codifying progression takes place in all the three dimensions of history: quadruped evolves later in the system, is learnt later by a child, and typically at least appears later in the text (cf. an animal that has four legs is called a quadruped).

Codifying

Halliday & Matthiessen (1999: 22-3):

Semogenic processes of the kinds just described take place in all three dimensions of semohistory: as the system of language evolves, as children develop their language, and as the language of a text unfolds. Hence language embodies the potential for its own ongoing expansion; and since the system at any moment is the repository of its own history, we can sometimes recognise disjunctions or interstices that offer a likely context for new meanings to appear. For example, the 'double -ing ' form of the English verb, which has recently been establishing itself (e.g. being raining, as in it seemed better to stay at home with it being raining), could have been predicted from a knowledge of the present state and recent history of the tense system. A change of this kind will propagate steadily throughout the system: sometimes very rapidly, but more often in an irregular and rather uneven flow.

Let us refer to this process as that of codifying, noting that as always it is at once both semantic and lexicogrammatical: there is no implication that meanings are already there and waiting to be codified.

Semogenic Processes And Markedness

Halliday & Matthiessen (1999: 22):

Typically processes of this type leave their traces in the form of marking, as marked/unmarked oppositions. The original member of the set remains the unmarked one. (One could say 'the unmarked term in the system'; but this formulation assumes that the offspring combine with the parent to form a system. Sometimes they do, but not always.) In these first set of examples, the unmarked mapping is that of participant ^ noun; when the noun realises some other element in the figure, it is a marked variant (grammatical metaphor). Presumably many 'unmarked' variants originate in this way, although in most instances we no longer have the evidence which would enable us to judge.

Deconstructing The Two Components Of The Sign: The Dissociation Of Associated Features In Wording

Halliday & Matthiessen (1999: 21-2):

A common variant of this process is that of the dissociation of associated features in the wording. We could represent this as:

Here again one meaning has been replaced by three: we now have (say) question₁ ↘ interrogative x rising tone, question₂ ↘ interrogative x falling tone, and question₃ ↘ declarative x rising tone, e.g. is she cóming? is she còming? she's cóming?

Expanding Meaning Potential: Deconstructing The Two Components Of The Sign

Halliday & Matthiessen (1999: 20-1):

(iii) There is also a third kind of semogenic process which arises from the nature of the sign itself. Our "sign" is not the Saussurean sign: we are not talking about the relationship between a word and its phonological representation (between content and expression, in Hjelmsiev's terms). The relationship is within the content plane, between a meaning and a wording — the non-arbitrary relationship between the system of semantics and the system of lexicogrammar.

This process, then, takes the form of deconstructing the two components of the sign. How is this possible? This can happen because, once a 'pair' of this kind has come into being, each component takes on an existence of its own. To pursue the example of the complex 'participant ^ noun' above: the category of 'participant becomes detached from that of noun, so that we can have participants realised by other things than nouns, and nouns realising other things than participants.

We now have three meanings instead of one: participants constructed by nouns, as hitherto, but now contrasting (a) with participants constructed by something other than nouns, and (b) with nouns constructing something other than participants. (Of course, neither of these two entails the other; there might be just many to one, not many to many. To take this actual example, there are many cases of nouns realising things other than participants; but relatively few cases of participants being realised otherwise than by a noun.)

Expanding Meaning Potential: Elaborating A Semiotic Domain In Delicacy

Halliday & Matthiessen (1999: 20):

(ii) Secondly, we may expand the meaning potential by increasing the semantic delicacy; for example

Here the semiotic domain has not been extended but rather has been brought into sharper focus, so that further shades of meaning are differentiated. A finer grid has been applied to the given semantic space.

Expanding Meaning Potential: Extending The System By Constructing New Semiotic Domains

Halliday & Matthiessen (1999: 19-20):

Suppose we now consider the semogenic processes whereby this potential expands. (i) We may, for example, construe new participants by creating new thing/name complexes: thus

Of course, the 'thing' may have been 'there' all along but it is only newly observed and semanticised:

Here we are extending the system by constructing new semiotic domains.

Wordings And Meanings Emerge Together

Halliday & Matthiessen (1999: 18-9):

In all these histories, the wordings and the meanings emerge together. The relationship is that of the two sides of the Stoic-Saussurean sign — best represented, perhaps, in the familiar Chinese figure yin & yang (which is in fact just that, a representation of the sign):

Thus, to return to our earlier illustration of the noun: what evolved, in the history of the system, was an entity on the content plane which had a structure as follows:

The relationship between the two sides of the sign is that of realisation: thus the meaning 'participant in a process: conscious or non-conscious being' is realised as the wording (class of wording) 'noun'.

Blogger Comments:

The dualities associated with yin & yang are of the same level of symbolic abstraction, whereas the Stoic-Saussurean sign involves two different levels of symbolic abstraction.

The Three Semohistories Related

Halliday & Matthiessen (1999: 18):

These are the three major processes of semohistory, by which meanings at continually created, transmitted, recreated, extended and changed Each one provides the environment within which the 'next' takes place, in the order in which we have presented them; and, conversely, each one provides the material out of which the previous one is constructed: see Figure 1-6.

As the upward pointing arrow suggests, the individual's (transfinite) meaning potential is constructed out of (finite) instances of text; the (transfinite) meaning potential of the species is constructed out of (finite) instances of individual 'meaners'. 

Following the downward arrow, the system of the language (the meaning potential of the species) provides the environment in which the individual's meaning emerges; the meaning potential of the individual provides the environment within which the meaning of the text emerges.

The sense in which grammar is said to construe experience will be somewhat different in each of these three time frames.

The Three Time Frames Of Semogenic Processes

Halliday & Matthiessen (1999: 17-8):

We need, therefore, a further guiding principle in the form of some model of the processes by which meaning, and particular meanings, are created; let us call these semogenic processes. Since these processes take place through time, we need to identify the time frames, of which there are (at least) three.

(i) First, there is the evolution of human language (and of particular languages as manifestations of this). Known histories represent a small fraction of the total time scale of this evolution, perhaps 0.1%; they become relevant only where particular aspects of this evolutionary change have taken place very recently, e.g. the evolution of scientific discourse. This is the phylogenetic time frame.

(ii) Secondly, there is the development of the individual speaker (speaking subject). A speaker's history may — like that of the biological individual — recapitulate some of the evolutionary progression along epigenetic lines. But the individual experience is one of growth, not evolution, and follows the typical cycle of growth, maturation and decay. This is the ontogenetic time frame.

(iii) Thirdly, there is the unfolding of the act of meaning itself: the instantial construction of meaning in the form of a text. This is a stochastic process in which the potential for creating meaning is continually modified in the light of what has gone before; certain options are restricted or disfavoured, while others are emprobabled or opened up. We refer to this as the logogenetic time frame, using logo(s) in its original sense of 'discourse'.