Grammatical Metaphor Renders Many Of The Semantic Relationships Implicit

Halliday & Matthiessen (1999: 545):

We have suggested that the immediate context for this change was a discursive one: the evolution of a register of experimental science, in which certain forms of argumentation were highly valued. This is usually interpreted simply as the emergence of a particular genre, the scientific article; but that is only one side of the story — no such genre could have come into being without these changes in the grammar of the clause. At the same time, they have other significant consequences. 

We have already pointed out the fact that one effect of grammatical metaphor is to render many of the semantic relationships implicit: if the happening is construed as a clause, the semantic relations are spelt out in the configuration of grammatical elements, whereas if it is construed as a nominal group they are not, or only partially so (compare his energy balance approach to strength and fracture with he investigated how strong [glass] was, and how it fractured, using [the idea that] the energy [...] balanced out). 

On the whole, the greater the degree of metaphor in the grammar, the more the reader needs to know in order to understand the text.

The Metaphorical Use Of Verbs To Express Logical-Semantic Relations Between Nominalised Processes

Halliday & Matthiessen (1999: 544-5):

When the processes and properties turn into nouns, the verbs do not disappear from the scene. Scientific discourse is still written in clauses, and these clauses still have verbs in them. Let us return briefly to the examples from Newton's "Opticks":

those Colours argue a diverging and separation of the heterogeneous Rays from one another by means of their unequal Refractions

the variety of Colours depends upon the Composition of Light

the cause of Reflexion is not the impinging of Light on the solid, impervious parts of Bodies

We might suggest a more congruent form of the first and second examples here: 

colours vary because light is composed [in this way];  

because those colours (appear] we know the heterogeneous rays diverge and separate from one another.... 

The verbs depend upon and argue both express a logical-semantic relation between the two nominalised processes: either an external cause, 'a happens; so x happens', or an internal cause, 'b happens, so we know y happens'. This is another grammatical metaphor; the congruent form of representation of a logical relation is a conjunction. The two types of metaphor work together, to construe the two processes as one: 'happening a causes happening x', 'happening b proves happening y'. 

It is not the case, of course, that this type of construction had never occurred in English before; it had. But it was rare; whereas from the time of Newton onwards it gradually took over, becoming the most favoured clause type of scientific language — as indeed we find it today.

The Textual Motivation For Nominalisation

Halliday & Matthiessen (1999: 543-4):

There are similar contexts [in Newton's Opticks] for expressions such as an inequality of Refractions, the impinging of Light on ... Bodies, the thickness of the Body, reflecting power in the passages cited above. The nominalised forms inequality, impinging, thickness are not technical terms; or rather, they are so to speak technicalised for the given instance, but they do not lose their semantic status as property or event. Why then are they reconstrued as nouns? 

The reason is to be found in the grammar of the textual metafunction. In order to function with the requisite value in the message, which means either as Theme or as focus of information, they cannot remain as complete clauses; they have to be "packaged" into single elements of clause structure, and the only available constituent for this purpose is the nominal group. Instead of being a process in its own right, light impinges on a body, the phenomenon in question is construed as a participant, the impinging of light on a body. It can then take on a clearly defined status in the grammatical construction of the discourse.

What is beginning to emerge here is a grammar for experimental science: a way of construing experiential meaning so that it can be organised textually into a form of discourse for the advancement of learning.

A Function Of Grammatical Metaphor In Newton's Argument

Halliday & Matthiessen (1999: 543):

If we look at Newton's "Opticks", from which these are taken, we find that it consists of three simultaneous discourses interspersed. In one of these phases, Newton describes his experiments; in another he draws conclusions from the experiments; and in the third he provides mathematical explanations. … The examples just cited of grammatical metaphor are typical of the second phase. It is here that Newton is proceeding by logical steps through a reasoned argument; and he frequently needs to summarise the argument up to that point, or in anticipation of what is to come. A typical sequence would be the following:

… when Light goes out of Air through several contiguous refracting Mediums as through Water and Glass, … that Light … continues ever after to be white. … … the permanent whiteness argues, that …

The metaphorical nominalisation permanent whiteness summarises the earlier sequence of inductions.

Grammatical Metaphor In The Writing Of Newton

Halliday & Matthiessen (1999: 542-3):

When we come to the writing of Newton, however, we find formulations such as the following:

… by these two Experiments it appears, that in equal Incidences there is a considerable inequality of Refractions. 

… the cause of Reflexion is not the impinging of Light on the solid impervious parts of bodies, but .. . 

… if the thickness of the body be much less than the Interval of the Fits of easy Reflexion and Transmission of the Rays, the Body loseth its reflecting power.

These contain a great deal of grammatical metaphor, contrast them with more congruent forms of expression such as 

light is refracted unequally (even) when it falls at the same angle; 

light is reflected not because it impinges on the solid, impervious parts of bodies'; 

if the body is much less thick than the interval between the points where the rays are easily reflected and (where they are easily) transmitted, the body is no longer able to reflect (light). 

Why has the mode of expression changed along just these lines?

The First Recorded Emergence Of Scientific Discourse In English

Halliday & Matthiessen (1999: 542):

This discourse first appears in English in the work of Chaucer, for example his "Treatise on the Astrolabe", written about 1391. Here we find the same linguistic resources brought into play: nouns as technical terms, and extended nominal groups. The former are partly technological (to do with the construction and operation of the astrolabe), generally Anglo-Saxon or Norman French, like plate, ring, turet (eye, or swivel), riet (from rete 'net', i.e. grid), moder ('mother', body of the instrument); and partly theoretical (from astronomy, mathematics or general methodology), mainly borrowings from Latin like altitude, ecliptik, clymat (climatic zone), degree, equation, conclusioun, evidence. The latter do not attain any spectacular length but involve the expected mixture of clauses and prepositional phrases, as in the same number of altitude on the west side of this line meridional as he was caught on the east side. This is clearly the discourse of organised knowledge; but it is not sharply set off from the language of everyday life.

It is with the "new learning" of the Renaissance that a distinct language of science begins to emerge, with a vastly greater dependence on grammatical metaphor. The earlier exercises in nominalisation had been abstract but only minimally metaphorical; there is a trace of grammatical metaphor in expressions like conclusion and the same number of altitude, but no more than is found in the language of daily life.

Semogenic Resources For Grammatical Metaphor: From Greek To Latin And Beyond

Halliday & Matthiessen (1999: 541-2):

The forms of scientific discourse developed by the Greeks were then taken over into Latin. Although Latin differed in certain significant respects (for example, it had no definite article, and it did not readily accept prepositional phrases as Qualifier in the nominal group), it was close enough to Greek, both linguistically and culturally, for this to present few problems. Most Greek derivational compounds could be calqued directly into Latin (e.g., peripherêia to circumferens ); Latin had its own stock of nominalising suffixes, like -atio(nem) and -mentum; and a reasonably similar potential for expanding nominal groups. 

As Latin took over as the language of learning throughout the greater portion of Europe, it had already developed an equivalent semogenic power. In the mediæval period Latin continued to serve; but by this time, although its morphology was largely unchanged, it had taken on the semantic patterning of the vernacular European languages. So when Latin itself was replaced by these languages, the transition was not unlike that which had taken place earlier into Latin from Greek: first the Greeks developed new meanings in Greek form, then these meanings were taken over into Latin forms, then new meanings were developed in Latin, then these new meanings were taken over into the modem European languages, then new meanings were developed within these languages. Thus there was a continuous evolution in the discourse of technology and science: in each transition, one component of the system was preserved.

A Second Semogenic Resource For The Emergence Of Grammatical Metaphor

Halliday & Matthiessen (1999: 540-1, 541n):

The second of the resources that was brought into play was a syntactic one, the structure of the nominal group. The nominal group of ancient Greek was very like that of modern English: it had a similar arrangement of elements around the Head noun, allowing both prepositional phrases and clauses in modifying function (with some difference of ordering), and included among its deictic elements one which was very close to the English the. Thus any noun could accumulate qualifying clauses and phrases which were explicitly signalled as defining, analogous to English the electrons in an atom, the angles which make up a triangle. 

One context which demanded elaborate nominal group structures of this kind was that of mathematics, as scholars conducted more and more sophisticated measurements, for example in their attempts to understand planetary motion. Here is the English translation of a nominal group from the work of Aristarchus of Samos, sometimes referred to as "the ancient Copernicus" because he was the first to propose that the earth revolved around the sun:

the straight line subtending the portion intercepted within the earth's shadow of the circumference of the circle in which the extremities of the diameter of the circle dividing the dark and the bright portions in the moon move

This has 32 words in the original Greek (fewer than the English because the equivalent of of is the genitive case of the noun); note that it is only the Subject of the clause, which continues ... is less than twice the diameter of the moon.¹

 

¹ The one minor piece of linguistic engineering that had to be undertaken was to ensure that the prepositional phrases in such constructions were placed after the Head word rather than before. It was also possible for a qualifying expression to be inserted between the Deictic and the Head, equivalent to English the in an atom electrons; this has only a limited potential for expansion, as can be shown by rewriting "The House that Jack Built" in this format:

This is the that Jack built house.

This is the that lay in the that Jack built house malt.

This is the that ate the that lay in the that Jack built house malt rat.

This is the that killed the that ate the that lay in the that Jack built house malt rat cat.

 With the qualifying phrase or clause at the end, the structure branches "to the right" (using the linear metaphor derived from European orthographies) and there is much less restraint on adding further elements. On the other hand, the fact that it is possible to put the qualifying element before the Head, which in English it is not, helps to avoid some ambiguity in the bracketing, as this example shows: here, both (which is) intercepted in the shadow of the earth and of the circle along which ... are qualifying the word circumference, and the Greek makes this clear with the intercepted-in-the-shadow-of-the-earth circumference of the circle along which ... . The English translation cannot follow this ordering. 

A First Semogenic Resource For The Emergence Of Grammatical Metaphor

Halliday & Matthiessen (1999: 539-40):

The first of these was their resource for creating technical terms. For systematic scholarship it is necessary to technicalise some of the words that are used, and this imposes two requirements: the words must be interpretable in an abstract sense, since they need to refer not to outward appearances but to the properties and principles that lie behind them; and they need to relate to one another in a regular and systematic way, so as to form stable taxonomies. 

Ancient Greek was a language of settlement, in which the potential for this kind of development lay predominantly in the nouns; and there existed already a number of noun-forming suffixes by which words of other classes — verbs and adjectives — were transcategorised. …

Using … nominalising suffixes the Greek scientists created hundreds of new technical terms; and by combining them with other derivational resources they developed extended series of semantically related forms …

In this way they established the foundations of the lexical component of a technical discourse, and the principles on which it could be indefinitely extended.

The Origins Of Grammatical Metaphor

 Halliday & Matthiessen (1999: 539):

Why did such a significant development take place? The most important single factor was undoubtedly the evolution of science and technology. It is possible to trace the emergence of this pattern of grammatical metaphor back to the origins of western science in ancient Greece, and to follow its development step by step; each stage in the evolution of the grammar realises a stage in the evolution of a world view.

The philosopher-scientists of the ancient Greek world, Thales, Pythagoras, Anaximander and their successors, inherited a language with a grammar of the kind outlined above, in which experiential meanings were construed in clausal patterns as a balanced interplay of happenings and things; nouns enjoyed no special privileged status. In the course of their writings (and no doubt first of all in the course of their sayings, only we have no access to these) they distilled this into a language of learning. We do not know how much they reflected on this process; it is unlikely they engaged in any very explicit language planning. What they did was to exploit the resources of everyday Greek, its fundamental semogenic potential. In particular, they exploited two of its grammatical powers: the power of forming new words, and the power of extending grammatical structures.

Why Metaphorical And Congruent Expressions Are Not Synonymous

Halliday & Matthiessen (1999: 538):

It might be maintained that a pair of expressions such as in times of engine failure and whenever an engine failed are simply synonymous, and do not imply any reconstruction of experience. But there are two problems with this view. 

One is that of history, referred to above. If neither had preceded the other, they could simply be free alternatives (though language is seldom so extravagant with its resources as this would imply!). But since one form of wording came first, it inevitably acquired a rich semantic loading. Since nouns evolved as names of classes of things, anything which is represented as a noun inevitably acquires the status of a thing, with the implication of a concrete object as the prototype. Thus in engine failure, the grammar has construed a thing called failure; and the nominal group then accommodates classes of failure (with another noun as Classifier), such as crop failure, heart failure and engine failure. Thus engine failure and engines fail are precisely not synonymous, because in engine failure the happening failure has acquired an additional semantic feature as the name of a class of things.

The second problem is that of sheer scale. If only odd, more or less random instances of this kind of metaphor occurred, they could have little effect on the system as a whole. But given the massive scale of this shift in the grammar, affecting as it does entire registers of modern English, it cannot simply be dismissed as meaningless variation. As we saw in Chapter 6, the metaphoric processes themselves are highly systematic; moreover they occur in typical syndromes, so that it is not just one aspect of the construction that is affected. Rather, the entire perspective is shifted sideways, so that each element in the configuration is reconstructed as something else. When this pattern comes to predominate throughout a large proportion of the discourse of adult life, it amounts to a fairly major resemanticising of experience.

The Evolutionary Emergence Of Grammatical Metaphor

Halliday & Matthiessen (1999: 537-8):

Under certain historical conditions, such a theory [i.e. the congruent model] may come to be modified or reconstructed. No doubt there have been various more or less catastrophic changes in earlier human history which have brought about relatively rapid changes in language — relatively, that is, to the gradual evolution of the system that has taken place all the time. We have no means of knowing about these. But it seems likely that what we are here calling grammatical metaphor represents one such partial reconstruction, in which, in the context of science and technology, a rather different kind of "reality" is being construed.

Blogger Comments:

Note the much earlier emergence of lexical metaphor, and the rather different kind of "reality" it construed: the rich symbolism of the various mythic traditions.

The Congruent As A Common Coding Of Experience

Halliday & Matthiessen (1999: 537):

"Congruent" is of course a contingent term. What it is saying is that, at the present moment in human history we can recognise forms of language which seem to represent a common coding of experience: this is the configuration that we referred to as "process + participant + circumstance" which is construed in grammars through some version of the trichotomy of verb, noun and the rest. If we relate this to English, it is the form of English that is learnt as a mother tongue, in which phenomena are interpreted clausally, in a kind of dynamic equilibrium of happenings and things. 

The prototypical thing is a concrete object which can be related by similarity to certain other objects, such that taken together they form a class, like engines. The prototypical happening is a change in the environment that is perceptible to the senses, or a change in the senser's own consciousness. A process is a happening involving one or two such objects, or one object and a conscious being. 

When children move from their own constructed protolanguage into the mother tongue, this provides a theory which they can use to give a plausible construction to their own individual experience.

The More Congruent Form As Historically Prior

(i) in times of engine failure

(ii) whenever an engine failed

Halliday & Matthiessen (1999: 537):

From a purely descriptive point of view, each version is metaphorical from the standpoint of the other; there is no inherent priority accorded to either. Once we bring in considerations of history, however, a clear priority emerges; and it is the same priority whichever of the three diachronic dimensions we choose to invoke — the phylogenetic (history of the language), the ontogenetic (history of the individual) or the logogenetic (history of the text). In all these three histories, version (ii), the clausal comes first. 

This form of construal evolved first in the history of English; version (i) emerged only as the result of a long process of later evolution. It comes first in the life of a child; children master version (i) only after a long (in terms of their young lives) process of becoming literate and being educated And it comes first in the unfolding of a text; we are much more likely to be told first that engines fail and only then to hear about a phenomenon of engine failure. 

Once we take note of progression in time, then given a pair of such expressions we can identify one of the two as the more metaphorical. The process is one of movement away from what we referred to as a "congruent" form.

Grammatical Expressions In A Metaphoric Relation

Halliday & Matthiessen (1999: 536-7):

To recapitulate with a very simple example: Given a pair of expressions such as in times of engine failure and whenever an engine failed the two are related to each other by grammatical metaphor. A particular phenomenon has been construed (i) as a prepositional phrase with a nominal group as its Complement, (ii) as a hypotactic clause introduced by a conjunction; moreover, the lexical content has been construed in two quite different ways:

If we take just the question of which elements function as Thing, the two are exactly complementary: in (i) the Thing nouns are time and failure, while in (ii) the only Thing noun is engine. This relationship is analogous to that of metaphor in its usual, lexical sense; only here the transfer is not between words but between grammatical classes.

Mode And Metaphor

Halliday & Matthiessen (1999: 536):

Many "spoken" languages may be realised in two different media: in speech, and in writing. At first this presents itself just as two modes of expression; but when we look more closely at discourse in spoken and written language we find regularly associated differences in grammatical construction. We find written language constructed in nominal groups, whereas spoken language is typically constructed in clauses. And since it is in the grammar that our experience is construed into meaning, what we are seeing are different forms of the construction of experience, one couched primarily in terms of figures, the other in terms of elements that make up such figures, mainly those that function as participating entities.

Our basic approach to this is embodied in the term "metaphor", as used in the context of metaphor in the grammar. We used the expression "grammatical metaphor" to refer to a complex set of interrelated effects whereby, in English and many other languages, there have evolved what seem to be alternative representations of processes and properties, in terms of word classes, meanings prototypically construed as verbs or adjectives come to be construed as nouns instead. But, as we saw, this is simply the superficial manifestation of a wider and deeper phenomenon affecting the entire construal of experiential meanings in the grammar.

The Ideational Semantics Of Sign

Halliday & Matthiessen (1999: 535):

… the semantic system that is construed in (say) Australian or British Sign is not the semantic system of English, despite its being constantly permeated by English in the ways referred to above. It is a system needing to be described in its own terms, based on detailed studies of its grammar and discourse such as now being carried out for Auslan by Trevor Johnston. Such a description is designed first and foremost for the needs of the deaf community; but it will have general significance as a source of further insight into the semantics of spoken languages, making it possible to view them comparatively in the light of an alternative construction of reality.

From Sign, we can get further insight into the construal of experience as ideational meaning because of its greater potential for iconicity in the expression. Semantic space can be construed iconically in signing within a continuous space-time, constituted as bodily experience for the signer and as part of shared visual experience for signer and addressee.

From a comparative standpoint, spoken languages and deaf sign languages stand to each other in a metaphoric relationship, as alternative construals of a (largely, though not totally) shared experience. What appear at first simply as different modes of expression turn out to have, associated with them, somewhat different constructions of meaning.

 

Influences Of Spoken On Sign Language

Halliday & Matthiessen (1999: 534-5):

Signers are also members of another language community, that of the (predominantly hearing) speakers of English, or whatever language is spoken around them; the two groups interact, and there is obviously no insulation between the two language systems. This gives rise to contact phenomena of two kinds: on the one hand, intermediate forms whereby English is realised in sign expressions (signed English, and finger-spelling), including a large number of new, "contrived" signs; and on the other hand, constant intrusion of English forms of expression, and therefore of English modes of meaning, into the sign language itself.

The Greater Potential Of Gestural Resources For Iconicity

Halliday & Matthiessen (1999: 534):

Gestural systems, by contrast, have a far greater potential for construing experience iconically. Thus in Johnston (1989: 16) "signs are roughly graded into four classes of transparent, translucent, obscure and opaque signs, depending on how iconic a sign is"; and while most signs fall in between the two extremes — Johnston grades them as obscure or translucent, rather than opaque or transparent — many of those he labels "obscure" have a popular explanation in iconic terms (e.g. 2388: n. CAMERA, v. TAKE A PICTURE, PHOTOGRAPH. Obscure action. Popular explanation: 'holding a camera and depressing the shutter button' [p. 301]). This suggests that even if particular explanations are "nothing more than deaf folklore" (p. 16), the system as a whole is perceived as prototypically iconic; and this feature is borne out in two important respects. 

One is that many of the signs construing basic categories of experience that would be learnt very early in childhood, in the transition from protolanguage to mother tongue — examples are 1291 GET; 1479 HOLD; 1824 RUN; 2473 BIRD; 2759 DRINK, CUP; 36 BED; 163 UP — are clearly iconic, and so would tend to establish iconicity as the norm. 

The other is that individual signs may be modified in a distinctively iconic fashion; e.g. 1471 "v LARGE, BIG, (with amplification) great, (with amplification and stress) enormous, huge, immense"; see in particular the section on "sign modification" in Johnston (1989: 494-9). As Johnston comments (p. 513), "A language which is itself visual and spatial has far more opportunities than an auditory one to map onto itself those very visual and spatial qualities of the world it wishes to represent".

The Limited Potential Of Vocal Resources For Iconicity

Halliday & Matthiessen (1999: 533-4):

We have referred all along to the primarily visual nature of human experience: how much of it is constituted as location, and especially movement, in space. Now, both gestural and vocal resources involve positioning and moving the organs of articulation in space; but the position and movement of the vocal organs, besides being largely out of sight of the listener, is very much mote constrained; and while this permits a limited degree of iconicity (association of close vowels with 'small', open with 'large', for example) this can never be more than a marginal feature of the system as a whole. Thus even allowing for the additional iconic potential of loudness and length, the expression systems of spoken languages must remain prototypically conventional. This is the familiar principle of the "arbitrariness" (i.e. conventionality) of the linguistic sign.

The Potential Of Vocal And Gestural Expressions For Construing Wording

Halliday & Matthiessen (1999: 533):

In the most immediate sense, as regards their potential for construing signifiers — elements of wording and their arrangement in combination — both these forms of expression, vocal and gestural, are open-ended. Neither of them imposes a limit on the inventory of morphemes or their configuration in grammatical structures. Nevertheless they are significantly different in the kinds of resource they offer for making meaning. Perhaps the major difference between the two is their potential for iconicity.

Deaf Sign Expression

Halliday & Matthiessen (1999: 533):

From one point of view, as we introduced it in the earlier discussion, Sign will appear as a realisational variant: that is to say, the meanings of language — its semantic system — may be construed either in sound or in gesture. There are, in fact, sign systems that are constructed along these lines: in principle this is what is meant by "signed English". But if we consider the nature of the two forms of expression, vocal on the one hand and gestural on the other, it is clear that they have very different properties. Gesture operates in a 3-dimensional "signing space" defined by reference to the signer's body and its parts, and movement within that space is entirely accessible to the receiver, thus in addition to succession in time (which is common to both), the gestural medium can exploit a number of parameters of spatial variation: the "articulatory organs" (fingers, hands, arms, other body parts), their location, orientation, thrust (direction and speed of movement) and so on.

The Interdependence Of The Metafunctions

Halliday & Matthiessen (1999: 532-3):

These three "metafunctions" are interdependent; no one could be developed except in the context of the other two. When we talk of the clause as a mapping of these three dimensions of meaning into a single complex grammatical structure, we seem to imply that each somehow "exists" independently; but they do not. There are — or could be — semiotics that are monofunctional in this way; but only very partial ones, dedicated to specific tasks. A general, all-purpose semiotic system could not evolve except in the interplay of action and reflection, a mode of understanding and a mode of doing — with itself included within its operational domain. Such a semiotic system is called a language.

The Evolution Of The Clause

Halliday & Matthiessen (1999: 532):

For a fuller understanding of the clause, we have to recognise that it evolved simultaneously as reflection, as action and as information: that is, not only as a representation of the phenomena of our experience but also as a means of social action, of moving around in interpersonal space (and so defining that space and those who occupy it); and as a semiotic construct, whereby language itself becomes a part of, and a metaphor for, the reality it has evolved to construe and to construct.

The Enabling Role Of The Textual Metafunction

Halliday & Matthiessen (1999: 532):

As far as the textual metafunction is concerned, therefore, any one clause will typically embody two sets of semantic choices. One will be its organisation as a message, a piece of information flowing from speaker to hearer, its limits defined by the speaker's point of departure and the focus of attention projected by the speaker on to the listener. The other will be the cohesion it sets up with the preceding moments of the discourse, as well as with other discourses and with the total semiotic environment. These enable the clause to function effectively as reflection and as action. 

But in the course of serving this enabling role, the textual component opens up a new dimension of meaning potential, in that it construes a further plane of "reality" that is as it were made of language — meaning not as action or as reflection but as information. In the modern world, when we increasingly live by exchanging information, rather than by exchanging goods-&-services as hitherto, this aspect of meaning potential is coming more and more to be foregrounded But it has always been there; and this is not the first time in history that it has proved to be an indispensable resource.

Lexical Cohesion

Halliday & Matthiessen (1999: 531):

Lexical cohesion refers to cohesion that is brought about by lexical means: choosing a word that is related in a systematic way to one that has occurred before. The range of semantic relations that can create cohesion in this way is very wide; but there are five principal conditions under which it occurs. These are: repetition, where the speaker simply repeats the same word; synonymy/antonymy, where a word is chosen that is similar or opposed in meaning; hyponymy/meronymy, where a word is chosen that is related by 'kind of' or 'part of — either vertically, like melon ... fruit or car ... wheel, or horizontally, like melon ... plum, or wheel ... mudguard; and collocation, which does not necessarily imply any particular semantic relationship but means simply that a word is chosen which is regularly associated with a previous one, like aim coming shortly after target, such that a resonance is felt between the two.

Cohesive Conjunction

Halliday & Matthiessen (1999: 531):

In conjunction, the various logical-semantic relations of expansion that construe clause-complex structures (discussed above under the "logical" metafunction) are deployed instead as a source of cohesion. There are a large number of such conjunctive expressions, ranging from single words like however, moreover, otherwise (many of them originally composite forms) to prepositional phrases like in that case, in other words, at the same time (often containing a reference word inside them). They cover more or less the same range of meanings that we referred to as "elaborating", "extending" and "enhancing"; but they do not establish any structural relationship in the grammar, and this is recognised in written English, where they regularly occur following a full stop.

Ellipsis And Substitution

Halliday & Matthiessen (1999: 531):

In ellipsis, some features which are present in the semantic construction of the clause (or other unit) are not realised explicitly in the wording, which cannot then be interpreted unless these features are retrieved from elsewhere. Here it is not the meaning that is being referred to; it is the wording that is being retrieved, usually from the immediately preceding clause (whereas reference can span considerable distances in the text). Ellipsis is particularly characteristic of dialogue, especially adjacency pairs such as question and answer. Sometimes in English a substitute element is inserted as a placeholder; e.g. ones in Which lanes are closed? — The northbound ones.

Referential Cohesion

Halliday & Matthiessen (1999: 530-1):

Reference is a way of referring to things that are already semiotically accessible: either actually, in the text, or potentially, in the context of situation. The English reference systems are the personals, especially the third person pronouns and determiners he/him/his she/her/hers it/its they/them/their/theirs, and the demonstratives this/these that/those and the maverick the (which emerged as a weakened form of that). Such systems evolved in a deictic function; when used anaphorically or cataphorically (that is, in deictic relation to the text), they create cohesion. There is also a third source of referential cohesion, through the use of comparison, with words such as same, other, different, less, smaller.

The Resources For Creating Cohesion

Halliday & Matthiessen (1999: 530):

Over and above its contribution to the grammar of the clause, what we are calling the "textual" metafunctional component comprises a further set of resources, which construe clauses and clause complexes into longer stretches of discourse without the formality of further grammatical structure. These are the resources for creating "cohesion". These are of four kinds: reference (sometimes called "phora", to distinguish it from reference as defined in the philosophy of language), ellipsis, conjunction, and lexical cohesion.

Theme And Information As Prototypical?

Halliday & Matthiessen (1999: 530):

All languages display some form of textual organisation of the clause. How far this kind of speaker - listener complementarity, with a quantum of information being construed out of the tension between the two, is a general or prototypical feature of this aspect of the grammar is not at all clear. 

In Austronesian languages, for example, there is typically a much more complex pattern of interrelationship between the textual and the ideational structures of the clause (mapping of Theme on to different transitivity roles) than is found in Indo-European. Even with regard to English, where it is well established how the flow of information is engendered in the grammar, opinions differ as to how far this should be seen as one continuous movement and how far as the intersection of two different periodicities (as we are inclined to interpret it). 

It may be a general principle that thematic status is more closely tied to the clause (as the locus of experiential and interpersonal choices) than is the listener-oriented pattern of given and new; in English the "quantum" of information that is defined by this latter construction is not, in fact, identical with the clause and may be smaller or larger. But all discourse is organised around these two motifs, which between them "add value" to the clause, enabling it to 'mean' effectively in the context in which it occurs.