… it has been the concern of systemic theory since the 1960s to model the “content plane” in terms of a stratal pairing within language, with “semantics” as the “upper model” that is construed [intellectually constructed] by the lexicogrammar as a whole.
… we stress that the categories and relations of experience are not “given” to us by nature, to be passively reflected in our language, but are actively constructed by language, with the lexicogrammar as the driving force. By virtue of its unique properties as a stratified semiotic system, language is able to transform experience into meaning.
The semantic perspective enables us to emphasise four aspects of human consciousness which have been rather less foregrounded in cognitive approaches. One is that of meaning as a potential, a systemic resource which is deployed in — and ongoingly modified by — individual acts of meaning in language. […] The second is that of meaning as growth, a semogenic resource which is constantly expanding in power by opening up new domains and refining those that are already within its compass. The third is that of meaning as a joint construction, a shared resource which is the public enterprise of a collective (whereas “thinking” is essentially a private phenomenon “located” within the individual). The fourth is that of meaning as a form of activity, a resource of energy which is powered by the grammar at the heart of every language.
… “understanding” something is transforming it into meaning, and to “know” is to have performed that transformation. There is a significant strand in the study of language […] whereby “knowledge” is modelled semiotically: that is, as system–&–process of meaning, in abstract terms which derive from the modelling of grammar.
… cognition “is” (that is, can most profitably be modelled as) not thinking but meaning: the “mental” map is in fact a semiotic map, and “cognition” is just a way of talking about language. … Instead of explaining language by reference to cognitive processes, we will explain cognition by reference to linguistic processes.
… what people do with language (whether this is modelled as social action, as cognitive process or as some form of abstract value system).
The system and the text are not two different phenomena: what we call the “system” of a language is equivalent to its “text potential”. Analysing discourse means, first and foremost, relating the text to the potential that lies behind it. There are perhaps three areas of discourse analysis that have figured prominently in systemic-functional research: literary-æsthetic, technical-scientific, and sociopolitical. In the first of these, the text carries value in its own right; when you analyse texts of this kind, you are aiming to explain not only why and how the text means what it does but also why it carries the value that it does. […] In analysing scientific and technical texts, the linguist is likely to be foregrounding the special properties that distinguish these texts from other varieties of written and spoken language, such that they are able to play a central part in the creation and transmission of knowledge. Scientific theories evolve from the conjunction of material and semiotic processes, and the advancement of science is powered by linguistic as well as technical resources. […] On the socio-political side, the researcher is investigating how discourse creates, maintains and transmits the social order (and hence may also be used to subvert it).
Semiotic phenomena — the patterns that evolve in a system that creates meaning — have this property: that they depend on the brain’s ability to adopt a multiple perspective. Meaning has to be at the same time both process and entity; both potential and actual; both general and particular. And since the only strategy we have for describing and explaining a system like this is that of refracting it through another cycle of meaning — a scientific theory — we find it difficult to transcend these awkward disjunctions.
The third complementarity is the outcome of the child’s learning to read and write; it means gaining access to the new dimension that was added to language by the evolution of writing systems. […] writing opened the way to a new and different mode of meaning, that we refer to as “written language”; that is where the complementarity resides, and full mastery of it not only takes a long time but also depends on maturation: it is simply not possible (or has not been up until now) for an eight- or nine-year-old to cope with the type of semiotic complexity that written language involves.
The next of the three complementarities to develop is that between grammar and lexis. This is achieved when the child moves out of protolanguage into mother tongue, typically during the first half of the second year of life. In the protolanguage there is no stratum of lexicogrammar — this has not yet developed; so obviously there can be no grammar and no lexis. The protolanguage is not grounded in reference, so there are no resources for the referential mode of meaning. Referential meaning develops with the mother tongue, beginning with the move into “common” as opposed to “proper” names (that is, naming classes rather than individuals), and then opening up this new semiotic dimension as the meaning potential expands to accommodate both particularised and generalised meanings. — meaning as lexis, and meaning as grammar.
[This complementarity] is in fact the very beginning of semiosis — of meaning as a distinct mode of behaviour, distinct from material action. At the very beginning of life, each act (crying, smiling, waving limbs about) is recreated on each occasion; it is not dependent on memory of previous acts. There are instances, but there is no underlying system — no persisting resource such that each act comes as its realisation. Then the time comes when the child retains the act in the form of memory; it is not reinvented on each occasion, but as it were cloned from stock. […] When this moment has been reached [by the age of 5 months] the child has taken the first step towards his first semiotic system, the “protolanguage” …
As I described it, grammar and lexis form a complementarity of focus; system and text form a complementarity of angle; speaking and writing form a complementarity of state. We could gloss these further by identifying the dimension along which they are counterposed.
- Grammar and lexis are opposed in delicacy: meaning as generalities or as particulars.
- System and text are opposed in instantiation: meaning as (collections of) instances or as potential.
- Speaking and writing are opposed in (manner of) realisation: meaning as happening or as thing.
But they all have a critical property in common: they are all strategies for extending the semogenic power of language — its meaning potential as a semiotic system.
Again, like biological systems, semiotic systems evolve to withstand shock; a language may get perturbed, deflected, even thoroughly mixed with another one […]. This is where the complementarities come in: they provide the surplus energy, the flexibility that enables a language to flourish in its changing eco-social environment.
Language is a product of evolution, and like any evolved system it needs to be seen as a totality: to understand any part of it, we have to locate it in the context of the whole. But that “whole” is not some idealised, static construction; it is a conglomerate resulting from constant small adjustments each of which follows on from whatever had been there before. As with biological evolution, one is sometimes tempted to say that it would have been better to start from somewhere else! But however imperfect evolved systems are, they usually function better than designed ones. A language is a functioning whole, even though every single feature of it, taken on its own, could no doubt be improved. There always are […] leftover bits and pieces lying around from evolutionary processes […] but they get built into the workings of the language in one way or another.
But, if we see grammaticalisation as the dimensionalising of semantic space, are there other comparable processes? Lexicalisation is the complementary process, likewise in perspective “from above”, so not a shift from the grammatical to the lexical but a construal of meaning in lexical terms — that is, those regions in the semantics of a language that are realised as (sets of) lexical items.
More localised features are typically intersected and construed in bundles, usually being assigned some preferred rôle in the configuration of processes: as the process itself, as a participant in the process, as quality of a participant, or as a circumstance (i.e. being assigned to an open word class — in English and Chinese, verb, noun, adjective or adverb). These open word classes construe meanings that we call “lexicalised”; and their “vectors” may be just the presence or absence of some particular component feature.
What we usually call “grammatical” are those sets of features that are construed as single vectors, and which tend to permeate a large number of semantic domains; for example, all potential participants are countable or uncountable, and if countable, are singular or plural.
Thus the lexicogrammar — linguistic form — dimensionalises semantic space; and the grammatical system network theorises this process, just as the phonological system network theorises the dimensionalising of phonetic space.
polarity is as it were pre-metafunctional: it develops (and probably evolved) prior to the distinction between ideational and interpersonal meanings. (Rather in the same way that, in the physical world, positively and negatively charged particles seem to be prior to the distinction between matter and information!) It may be that the “energy levels” of modality (high / median / low) are also of this kind: both modality and polarity are regularly grammaticalised in combination with other features.
In terms of children’s early language development, the interpersonal metafunction provides the prototype of how meanings come to be grammaticalised. The two systems that were first grammaticalised by one small child (Nigel, aged 0;10) were:
polarity: positive / negative
modality: value: low / high
followed shortly by the two forms of “appreciation” in conjunction with the feature “positive”:
appreciation (positive): impact (“that’s interesting”) / quality (“that tastes nice”)
Note that these were not yet mother tongue; they were protolanguage, realised by sounds and gestures. But they were systemic, or at least proto-systemic; and they provided the model for the linguistic systems of appraisal, where each lexical item realises the intersection of an appraisal feature with polarity and/or modality.
This is a grammatical system that is realised by a selection of lexical items. Each such item is uniquely identified as a set of intersecting grammatical features; eg complicated isappraisal: attitude: appreciation: composition: ( complexity : complex / polarity: negative …)
as well as other general grammatical features (e.g. as distinct from confusing, which is “effective”, as in it confuses me, complicated is “descriptive”). Note that I am interpreting the feature "undesirable" (the "snarl" member of the "purr/snarl" opposition) as negative in the environment of the interpersonal.