Date post: | 09-Sep-2016 |
Category: |
Documents |
Upload: | david-banks |
View: | 215 times |
Download: | 0 times |
Language Sciences 26 (2004) 391–410
www.elsevier.com/locate/langsci
Anglophone systemicists and Frenchenunciativists: shall the twain never meet?
David Banks
Facult�e des Lettres et Sciences Sociales, Universit�e de Bretagne Occidentale, 20 rue Duquesne,29285 Brest, France
Accepted 25 June 2003
Abstract
Is a rapprochement between Systemic Functional Linguistics and the Th�eories de l’�enon-ciation possible within the French university system? These seem to represent respectively
inductive and deductive modes of thought. It is therefore of interest to look at possible points
of contact and dissimilarities. The Th�eories de l’�enonciation are metaphysical theories and as
such are not subject to the Principle of Falsification. The domains of the Th�eories de l’�enon-ciation seem conceptually close to the metafunctions of the systemic model. The Th�eories del’�enonciation might benefit from adopting the distinction between thematic and information
structure common in the systemic approach. Systemic Functional Linguistics treats texts as a
whole, whereas the Th�eories de l’�enonciation tend to analyse individual segments within a text.
They also tend to treat only those grammatical questions with which the theory deals well. The
Th�eories de l’�enonciation appear to have a link with formalist approaches in that they accept
basic forms which are subsequently modified in the cognitive process; Systemic Functional
Linguistics may have produced something similar in the concept of congruent forms. Al-
though starting from diametrically opposed points, it might be hoped that these two theories
are working towards each other, and might at some point meet.
� 2003 Elsevier Ltd. All rights reserved.
Keywords: Deduction; Grammar; Induction; Systemic Functional Linguistics; Text; Th�eories de
l’�enonciation
E-mail address: [email protected] (D. Banks).
0388-0001/$ - see front matter � 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.langsci.2003.06.002
392 D. Banks / Language Sciences 26 (2004) 391–410
1. Introduction
It seems opportune, at the beginning of a new century to speculate on the di-
rections in which Systemic Functional Linguistics might develop in the near future. I
should like to do this within the context of my own working environment, which is
the English Departments of French universities. 1
The teaching of language and linguistics in the English Departments of French
universities is dominated by cognitive theories which have been developed in France.These are generally known as Th�eories de l’�enonciation, of which there are two
variants. The first, based on the work of Antoine Culioli, is known as the Th�eorie desop�erations �enonciatives (Culioli, 1990, 1999a,b), while the other, originally an off-
shoot from the Th�eorie des op�erations �enonciatives, based on the work of Henri
Adamczewski, is known as La grammaire m�etaop�erationnelle (Adamczewski and
Delmas, 1982; Adamczewski, 1996). To some extent both of these can be traced back
to the work of Emile Benveniste (Benveniste, 1966, 1977), although Culioli himself
has recently disclaimed any significant influence of Benveniste on his own work(Culioli, 2002). An earlier cognitive theory known as Psychom�ecanique, based on the
work of Gustave Guillaume (Joly and O�Kelly, 1990) also has some currency, and
some try to find a rapprochement between this and the Th�eories de l’�enonciation. The
publications in this field have been almost exclusively in French (though this may
now be beginning to change), and published in French journals. As a result, these
theories are almost unknown outside of France. Indeed the work of Culioli himself is
scattered throughout a number of fairly obscure publications, not always easy for
anyone outside this school of thought to locate. Fortunately, the most important ofthese have recently been collected in three volumes (Culioli, 1990, 1999a,b).
On the other hand, Systemic Functional Linguistics has made little impact in
France. The name of Michael Halliday is well known, but most frequently associated
with Cohesion in English (Halliday and Hasan, 1976). It is rare to find more than
fleeting knowledge of later work, almost as though Systemic Functional Linguistics
had dropped dead in its tracks in the 1970s. There is a French functional school,
based on the work of Andr�e Martinet (e.g. Martinet, 1962, 1979, 1970), but this has
virtually no impact within the English Departments of the university system.This thumbnail sketch is, of course, not in any way an exhaustive survey of lin-
guistic activity in France. Significant work has been done in the area of Speech Acts
(e.g. Ducrot, 1984, 1972; Recanati, 1981) and Generative Grammar (e.g. Pollock,
1997), to mention only two other fields.
If Systemic Functional Linguistics is to develop in France, then it must, for the
foreseeable future, do so alongside the Th�eories de l’�enonciation. In this context, it is
1 An earlier version of this paper was read at the 12th Euro-International Systemic Functional
Linguistics Workshop, Glasgow, 2000. I should like to thank those who took part in the discussion
following the oral presentation of the paper, as well as Janet Ormrod, Pierre Cotte, and an anonymous LS
referee for their comments on earlier drafts. It goes without saying that none of these are responsible for
any shortcomings that remain.
D. Banks / Language Sciences 26 (2004) 391–410 393
useful to ask whether the development of these two very different theoretical ap-
proaches can be more than separate parallel developments. Are there any points of
contact and is there any room for mutual influence? Is there anything that they can
learn from one another? Is a rapprochement possible between these two apparently
separate worlds? It would be impossible to give a full summary of the Th�eories del’�enonciation here; such an undertaking would be far beyond the scope of this paper.
What I hope to do is to point out a few areas which indicate that although they are
theoretically very different, these two approaches do nevertheless have points ofcontact, and can in some ways complement each other. For those who want a fuller
picture, I can do no better than direct them to the relevant works in the reference list
at the end of this paper, notably Adamczewski, 1996, Bouscaren, 1991, Delmas and
Girard, 1993, Gilbert, 1993.
2. Two ways of thinking––a caricature
It seems to me that the Th�eories de l’�enonciation represent a typically ‘‘French’’
way of thinking, while Systemic Functional Linguistics is basically ‘‘Anglophone’’.
This is an old story which goes back at least as far as the disputes between Locke and
Descartes in the 17th century.
Locke�s approach to knowledge is based on experience, even if he does allow for
‘‘internal’’ experience (Bennet, 1971). It is an empirical approach, and as such owes a
great deal to Bacon. Locke says:
Let us then suppose the mind to be, as we say, white paper void of all charac-
ters, without any ideas. How comes it to be furnished? Whence comes it by that
vast store which the busy and boundless fancy of man has painted on it with an
almost endless variety? Whence has it all the materials of reason and know-
ledge? To this I answer, in one word, from experience; in that all our knowledge
is founded, and from that it ultimately derives itself. Our observation, em-
ployed either about external sensible objects, or about the internal operations
of our minds perceived and reflected on by ourselves, is that which supplies our
understandings with all the materials of thinking. These two are the fountains
of knowledge, from whence all the ideas we have, or can naturally have, do
spring (Locke, 1688 [1965], 2.1.2).
Descartes on the other hand puts ideas first. Mathematics is the model on which
his vision of knowledge is based, and everything else is produced by a process of
deduction.
These long chains of reasonings, quite simple and easy, which geometers are
accustomed to using to teach their most difficult demonstrations, had given
me cause to imagine that everything which can be encompassed by man�sknowledge is linked in the same way, and that, provided only that one abstains
from accepting any for true which is not true, and that one always keeps the
394 D. Banks / Language Sciences 26 (2004) 391–410
right order for one thing to be deduced from that which precedes it, there can
be nothing so distant that one does not reach it eventually, or so hidden that
one cannot discover it. And I was in no great difficulty in seeking which to
begin with because I know already that it was with the simplest and easiest
to know; and considering that, among all those who have already sought truth
in the sciences, only the mathematicians have been able to arrive at any proofs,
that is to say, certain and evident reasons, I had no doubt that it was by the
same things which they had examined that I should begin, although I didnot expect any other usefulness from this than to accustom my mind to nourish
itself on truths and not to be content with false reasons (Descartes, 1637 [1968]
(trad. F.E. Sutcliffe)). 2
Thus Locke�s model is an empirical model, 3 while that of Descartes is mathe-
matical (O�Connor, 1967). These two opposing points of view, when applied tolanguage give two ways of dealing with linguistic phenomena. The inductive method,
derived from Locke, looks at data first, and then on the basis of that data attempts to
construct a theory. This is the bottom-up, empirical approach of the anglophone
tradition. The deductive method, derived from Descartes, is top-down. It develops a
theory, and then attempts to verify the theory by looking at data. This is the French
cartesian tradition. This is obviously a simplistic caricature. No academic enquiry
takes place in a vacuum, and so all studies are to some extent influenced (positively
or negatively) by previous work in the field. So even inductive enquiry has thebackground of existing theory and will have some idea about the theory that it will
(perhaps even wishes to) ultimately construct. In the same way deductive research
will have some knowledge of the data available while building its theoretical
framework. It is for this reason that those working within the Th�eories de l’�enonci-
2 The original French reads as follows:
Ces longues chaınes de raisons, toutes simples et faciles dont les g�eom�etres ont coutume de
se servir pour parvenir �a leurs plus difficiles d�emonstrations, m�avaient donn�e occasion de
m�imaginer que toutes les choses qui peuvent tomber sous la connaissance des hommes
s�entre-suivent en meme fac�on, et que pourvu seulement qu�on s�abstienne d�en recevoir au-
cune pour vraie qui ne le soit, et qu�on garde toujours l�ordre qu�il faut pour les d�eduire les
unes des autres, il n�y en peut avoir de si �eloign�ees auxquelles enfin on ne parvienne, ni de
si cach�ees qu�on ne d�ecouvre. Et je ne fus pas beaucoup en peine de chercher par lesquelles
il �etait besoin de commencer, car je savais d�ej�a que c��etait par les plus simple et les plus
ais�ees �a connaıtre, et consid�erant qu�entre tous ceux qui ont ci-devant recherch�e a v�erit�e
dans les sciences il n�y a eu que les seuls math�ematiciens qui ont pu trouver quelques
d�emonstrations, c�est-�a-dire quelques raisons certaines et �evidentes, je ne doutais point
que ce ne fut par les memes qu�ils ont examin�ees; bien que je n�en esp�erasse aucune autre
utilit�e, sinon qu�elles accoutumeraient mon esprit �a se repaıtre de v�erit�es et ne se contenter
point de fausses raisons (Descartes (1637 [1965]), 2e partie).
3 Some might claim that it is possible to espouse empirical method without necessarilly holding an
empiricist view of the nature of knowledge. Locke himself was of course both an empiricist, and a believer
in empirical method.
D. Banks / Language Sciences 26 (2004) 391–410 395
ation can claim that they also �start from data�. However, even if this view of things is
a caricature, all caricatures have an element of truth in them, and to that extent it can
serve as a useful starting point for this enquiry.
Although I am presenting Systemic Functional Linguistics as part of the anglo-
phone tradition derived from Locke, there are naturally exceptions to this trend.
Probably the most obvious case is that of Chomsky, and of generative grammar in
general, which is distinctly deductive in its approach and firmly attached to the
cartesian tradition, even to the extent of virtually enshrining that claim in the title ofone of his books, Cartesian Linguistics (Chomsky, 1966).
3. Systemic Functional Linguistics and the Th�eories de l’�enonciation
If the above simplistic caricature is now applied to Systemic Functional Lin-
guistics and the Th�eories de l’�enonciation, it might be said that Systemic Functional
Linguistics starts from language and moves towards the cognitive, while the Th�eoriesde l’�enonciation start from the cognitive with the intention of getting to language. If
this is true, it would seem that the two theories are on the same road, but are moving
towards each other from opposite ends. In this case they could be expected to meet
each other at some half-way point.
For many years the word �cognition� seemed to be taboo in systemic circles, but
more recently there has been a recognition that it does have a place. This is par-
ticularly noticeable in the work of Matthiessen (1998) and has recently been ex-
plicitly concretized in Halliday and Matthiessen, 1999.
It seems to us that our dialogue is relevant to current debates in cognitive sci-
ence. In one sense, we are offering it as an alternative to mainstream currents in
this area, since we are saying that 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. In modelling
knowledge as meaning, we are treating it as linguistic construct: hence, as
something that is construed in the lexicogrammar. Instead of explaining lan-guage by reference to cognitive processes, we explain cognition by reference
to linguistic processes (Halliday and Matthiessen, 1999, p. x).
It is possible that this cognitive element has always been to some extent implicit in
the systemic model. The notion of a system network is one of the basic tenets of the
model, even if systems tend to get discussed rather less than they used to. The system
presents the set of options or choices which is available to the speaker, so it is ba-
sically to do with speaker choice. This is something which has remained fairlyconstant in systemic theory over at least a quarter of a century, as the following
quotes show.
A system is a set of options in a stated environment (Halliday (1970) in Kress,
1976, p. 26).
396 D. Banks / Language Sciences 26 (2004) 391–410
Systems are lists of choices which are available in the grammar of a language
(Berry, 1975, p. 142).
[Systemic grammar] interprets language not as a set of structures but as a net-
work of SYSTEMS, or interrelated sets of options for making meaning (Hall-
iday, 1994a, p. 15).
The notion of the system posits that the resources of a language are such that itsupplies the speaker with a series of choices. These choices form a network which
constitutes the system. If we conceive of this in terms of language production or
generation, then we must think of the system in terms of speaker choice, and if we
are concerned with speaker choice, we must in some sense be dealing with a cognitive
domain. Choosing is a mental, hence cognitive, activity, and even if it is claimed that
in this area choosing is subconscious, motivation and criteria must be present for the
choice to take place. It cannot be thought of as haphazard. Hence, the systemic
approach has always had a cognitive facet, albeit implicit and tacit.Turning now to the Th�eories de l’�enonciation, this approach places theory in a
primary position.
The linguist should. . . study the way language functions as a meaningful activ-
ity of representation, that is as an activity of production and recognition of ‘‘lin-
guistic forms’’. . ..Given the multiplicity and heterogeneity of the phenomena to be considered,
and if there is a wish to cast light on whatever regularity there is, it is not pos-sible to do so without a theoretical framework. . . It is first necessary to adopt a
theory of observable entities in order to determine what types of data are to be
subjected to analysis (Gilbert, 1993, 64–65). 4
Adamczewski and Gabilan (1996), writing for the secondary education sector, put
it like this.
The secret is to look in the grammar, understood as the set of construction pro-cedures proper to a given language. Creating even the simplest sentence brings
4 This, and all subsequent translations, are my own. they are not intended to be polished translations,
but to give something of the flavour of the original French. The original French of this quote is as follows.
Le linguiste doit. . . �etudier le foncionnement du langage en tant qu�activit�e signifiante derepr�esentation, c�est-�a-dire en tant qu�activit�e de production et de reconnaissance de formes ‘‘lin-guistiques’’. . ..
Etant donn�e la multiplicit�e et l�h�et�erog�en�eit�e des ph�enom�enes �a envisager, et si l�on d�esire
mettre �a jour quelque r�egularit�e que ce soit, il n�est pas possible de faire l��economie d�une con-
struction th�eorique,. . . Il faut tout d�abord se doter d�une th�eorie des observables pour d�eter-
miner quels sont les types de donn�ees que l�on va soumettre �a l�analyse (Gilbert, 1993, pp.
64–65).
5
D. Banks / Language Sciences 26 (2004) 391–410 397
into play a whole series of operations, of choices, carried out spontaneously, un-
consciously (automatisms) according to the rules of the game of the grammar. . ..How can the grammar (the rules of the grammatical game) be discovered? These
rules, these ordering operations and these choices are not visible. What is visible,
or rather audible in the case of spoken language, is the result of the subterranean
work carried out by the speaker (Adamczewski and Gabilan, 1996, p. 7). 5
The importance of speaker choice is particularly clear in the second of these quotes.
So both of these theories have a concern with the question of speaker choice, though
in rather different ways. In the case of systemic theory, it is a question of the set of
choices that the system makes available to the speaker, so systemic theory starts fromthe language end of the equation. In the case of Adamczewski�s version of the
Th�eories de l’�enonciation, it is the way in which the human mind operates in producing
language that interests him. He sets out to attempt to discover the mental operations
of the speaker in the course of the act of speaking. These constitute a set of choices.
Adamczewski is thus very much at the speaker�s end of things. However, although
these are different ways of approaching speaker choice, indeed, in terms of point of
view, they might seem diametrically opposed, nevertheless they are not totally dif-
ferent, and certainly not incompatible. They provide then a point of contact.There is of course a distinction between language as an abstract system, and
language as what is produced as speech or writing. It may be significant that French
distinguishes lexically between these, having langage for the former, and langue for
the latter, whereas in English language covers both these. Consequently the dis-
tinction, if it is to be made in English, must be made specifically, whereas in French
this distinction is in a sense automatic, being built into the lexis, and one might
hypothesize that French-speakers might as a result be more conscious of the move
from system to production.
4. Verification and falsification
I would now like to make a philosophical digression, but one that I hope will later
be seen to be significant. I would like to consider the nature of the knowledge that
Original French:
Le secret. . . est donc �a chercher dans la grammaire, comprise comme l’ensemble des proc�ed�es deconstruction propres �a une langue donn�e. Fabriquer la moindre phrase met en jeu toute une s�eried’op�erations, de choix effectu�es spontan�ement, non-consciemment (automatisme) selon les r�egles
du jeu de la grammaire. . ..
Comment d�ecouvrir la grammaire (les r�egles du jeu grammatical). . ., Ces r�egles, ces op�era-tions de mise en ordre et ces choix ne sont pas visibles. Ce qui est visible, ou plutot audible dans
le cas de la langue parl�ee, c�est le resultat de tout le travail souterrain effectu�ee par l’�enonciateur(Adamczewski and Gabilan, 1996, p. 7).
398 D. Banks / Language Sciences 26 (2004) 391–410
these two approaches provide. Ayer (1936 [1971], 1956) basing his point of view on
the work of the Vienna Circle, suggested that in order for knowledge to be accepted as
such it must satisfy the Verification Principle. According to this principle, we must
know what it would be necessary to do to verify a proposition in order for it to be
accepted as knowledge.
We say that a sentence is factually significant to any given person, if, and only
if, he knows how to verify the proposition which it purports to express. . .[otherwise] it is, if not a tautology, a mere pseudo-proposition. The sentence
expressing it may be emotionally significant to him; but it is not literally signif-
icant Ayer (1936 [1971], p. 48).
If we have no way of verifying or testing, even in principle, the proposition in
question, then, it is claimed, we cannot know what it means. It is literally mean-
ingless, and hence cannot constitute knowledge.
A statement. . . has meaning if and only if it�s testable in some way. If you pre-
sent me with a statement and can think of no way it could be tested, if no ob-
servation you can think of would have any bearing on its truth or falsity, would
you know what it means? And if two statements could be verified by exactly thesame series of observations, wouldn�t they have the same meaning? (Hospers,
1967, 262).
According to this point of view, it was not necessary that it be possible to actuallycarry out the verification. At that time, in the 1930s, it was technically impossible to
verify statements about the dark side of the moon. Nevertheless, such statements
were considered meaningful, since what had to be done to verify them was known,
even though technically impossible at that time. Where, however, verification is not
possible, not even in principle, the proposition is a �metaphysical pseudo-proposi-
tion�; it belongs to the realm of metaphysics, which for Ayer means that it has no
literal significance, and hence is, strictly speaking, nonsense.
This position runs into trouble in the case of universal statements (Caws, 1965). Forexample, the propositionAll crows are black can never be totally verified, for nomatter
how many black crows have been observed, the possibility of observing a non-black
crow always remains. Thus verification is impossible, and if one sticks to the strict
Verification Principle one would have to claim that universal statements are mean-
ingless, which is obviously counter-intuitive. In order to accommodate this objection,
Popper (1959, 1994) suggested replacing the Verification Principle with a Falsification
Principle. According to this principle, for a statement to be meaningful, it is necessary
to know what it would be necessary to do to show it to be false. In Popper�s sense�meaningful� is to be taken in the context of distinguishing science from non-science,
rather than distinguishing sense from nonsense; and it formed part of a larger strategy
by which Popper attempted to distance himself from the Vienna Circle (Edmonds and
Eidinow, 2001). In the case ofAll crows are black, it would be necessary to observe one
non-black crow to show this statement to be false; hence it is meaningful.
D. Banks / Language Sciences 26 (2004) 391–410 399
. . . only if I can say how my theory might be refuted, or falsified, can I claim
that my theory has the character of an empirical theory.
It does not imply that irrefutable theories are false. Nor does it imply that they
are meaningless. But it does imply that as long as we cannot describe what a pos-
sible refutation of a certain theory would be like, that theory may be regarded as
lying outside the field of empirical science Popper (1963 [1994], p. 88).
This is the version which I prefer, and I believe it constitutes a difficulty, thoughnot an insuperable one, for the Th�eories de l’�enonciation.
The findings of the Th�eories de l’�enonciation are essentially about the workings of
the human mind, in other words about non-observable phenomena. As such they are
non-falsifiable, they do not fulfil the Principle of Falsifiability. These findings then are
non-physical in the sense that they do not relate to the physical world; they fall within
the metaphysical domain. For example, one of the main findings of this approach in
relation to English is that the use of progressive aspect in a proposition, or predicate
relationship (relation pr�edicative in enunciative terminology) is that this indicates thatthe proposition has been preconceived or preconstructed by the speaker. The evidence
for this in language is that there is frequently a trace of the alleged phenomenon
in the form of some sort of anaphoric reference. This is an interesting point because
as long as anaphora is taken in a fairly wide sense, these anaphoric references can
be found in most cases where progressive aspect is used. There remain, however,
a few cases where it is impossible to find anything that can be described as anaphoric,
and this is where the difficulty arises, since it is claimed that even here mental
preconstruction has taken place. Thus in this case it is claimed that absence of theexpected trace does not invalidate the theory. Hence it is impossible to falsify, and
is not subject to the Principle of Falsification. A further example is that it is claimed
that in using the English word ‘‘any’’, the speaker scans all items of a class
without selecting any particular individual item. Once again it is difficult to see what
sort of information or observations would be necessary to verify or falsify this
claim.
It seems to me then that the Th�eories de l’�enonciation provide hypotheses, but that
they cannot go beyond this hypothetical state. The only person I know of workingwith this framework who openly accepts the purely hypothetical nature of their
results is Cotte (1996, 1997). It is more usually stated, and taught, in axiomatic
fashion, as though it were established incontravertible truth. I do not think that the
hypothetical nature of the findings of the Th�eories de l’�enonciation in any way in-
validates them. Hypotheses and statements of a metaphysical nature can be highly
interesting in their own right. Theology is the obvious area where many of the claims
are of a metaphysical nature, and whether one agrees with these claims or not, it
would be difficult to say that they are of no interest simply because of their meta-physical nature. Even without going into an area which some might think of as
marginal, many fields within the Social Sciences (psychology, sociology, etc.) make
claims of a metaphysical nature. Again one would not want to deny the interest
of these claims because they are metaphysical. On the other hand it is important
to recognise the true nature of the results of academic enquiry, and to distinguish
400 D. Banks / Language Sciences 26 (2004) 391–410
between metaphysical results which of their very nature must remain hypothetical,
and empirical results, which, if true, can be taken as verifiable (or falsifiable) fact.
5. Domains and metafunctions
I would now like to look at a detail of theory, where it seems to me that what is
said within the Th�eories de l’�enonciation parallels fairly closely the position in Sys-temic Functional Linguistics. Delmas, working in the Adamczewskian variant of this
approach, has developed the notion of �domain� (Delmas et al., 1993; Delmas and
Girard, 1993). He suggests that there are three domains.
The first of these is the Infraverbal Domain (D1), which concerns the speaker�spersonal analysis of the real. It is
. . . the stage which precedes encoding. . .The operations brought into play in D1 (visual and auditive perception, cate-gorization, taking into account spatial and localizing relationships, the location
of objects, gestural and graphic links). . . give structure to an extralinguistic rep-
resentation (Delmas and Girard, 1993, p. 5). 6
The second domain is the Symbolic Domain (D2), which concerns the analysis of
the texture of the utterance.
In this domain the words are presented in a string. . . Two types of problem. . .are found:
• the handling of the linear ordering of the words. . .• the handling of other words, which are structural, and thus create hierarchies,
and which are linked to taking into account the texture of linguistic linearity
(Delmas and Girard, 1993, p. 5). 7
6 Original French:
. . .l��etape qui pr�ec�ede le codage verbalis�e. . .
Les op�erations mises eu jeu en D1 (la perception visuelle, auditive, la cat�egorisation, la prise en
compte des relations spatiales, localisantes, des rep�erages d�un objet, les relais gestuels ou graph-
iques). . . structurent une repr�esentation extralinguistique (Delmas and Girard, 1993, pp. 4–5).
7 Original French:
Dans ce domaine, les mots se pr�esentent en une ligne. . . On trouve. . . deux ordres de probl�emes:
• la gestion dans le lin�eaire des mots. . .
• la gestion d�autres mots, structuraux, donc hierarchisants, li�es �a la prise en compte de la tex-
ture du lin�eaire linguistique. . . (Delmas and Girard, 1993, p. 5).
D. Banks / Language Sciences 26 (2004) 391–410 401
The third domain is the Pragmatic Domain which is intersubjective in nature.
This domain concerns everything that the speaker makes or attempts to make
his addressee do.
• The speaker appeals to his addressee.
• The speaker can indicate that he and his addressee have already collaborated in
a task of communication and construction. . . (Delmas and Girard, 1993, p. 5). 8
A tripartite system is present, though in less obvious fashion, in the work of Cotte
(1996, 1997).
If Delmas� domains are compared with the metafunctions of the systemic model a
series of relations is apparent. According to Halliday (1994a, 179) the ideational
metafunction is the clause as representation; it is �what the clause is about�. The
interpersonal metafunction, the clause as exchange, is �what the clause is doing, as a
verbal exchange between speaker–writer and audience�. And the textual metafunc-
tion, the clause as message, is �how the clause relates to the surrounding discourse,and to the context of situation in which it is being produced�. Or in Thompson�s(1996, 28) version, the ideational metafunction is used
. . . to talk about our experience of the world, including the worlds in our own
minds, to describe events and states and the entities involved in them.
The interpersonal metafunction is used
. . . to interact with other people, to establish and maintain relations with them,to influence their behaviour, to express our own viewpoint to things in the
world, and to elicit or change theirs.
And the textual metafunction is where
. . . we organise our messages in ways which indicate how they fit in with the
other messages around them and with the wider context in which we are talking
or writing.
Although the metafunctions and domains are obviously not identical there seemsto be a significant series of relationships. The infraverbal domain and the ideational
8 Original French:
Dans ce domaine, les mots se pr�esentent en une ligne. . . On trouve. . . deux ordres de probl�emes:
• la gestion dans le lin�eaire des mots. . .
• la gestion d�autres mots, structuraux, donc hierarchisants, li�es �a la prise en compte de la tex-
ture du lin�eaire linguistique. . . (Delmas and Girard, 1993, p. 5).
402 D. Banks / Language Sciences 26 (2004) 391–410
metafunction both deal with representations of the world, even if the former deals
with a pre-encoding situation and the latter with the semantics of the message as
encoded. This is perhaps the weakest of the three relationships, for although both are
related to content, the infraverbal domain, dealing with a pre-language level, is in the
cognitive or mental field, where the ideational metafunction deals with the clause, a
language phenomenon. The pragmatic domain and the interpersonal metafunction
both deal with the relationships between the speaker and his addressees. The prag-
matic domain is concerned with collaboration in a task of communication and caninclude the speaker�s attempts to influence his addressee�s actions. The interpersonal
metafunction is involved with the verbal exchange, and is about the way people
interact, which includes influencing their behaviour. And the symbolic domain and
the textual metafunction both deal with the construction and cohesion of the mes-
sage itself. The symbolic domain concerns the linear ordering of words and the
texture of linguistic linearity, while the textual metafunction deals with the clause as
message and the way it relates to surrounding discourse and fits in with other
messages.It seems reasonable then to see a fairly direct connection, or at least a certain
parallelism, between Delmas� domains and systemic metafunctions:
In the work of Culioli, the concept of domain is rather different. Only one domain
is posited, the notional domain, and this seems closest to the infraverbal domain of
Delmas� system. For Culioli, this notional domain is a theoretical space where ideas
or notions, and the relations between them, are developed at a prelinguistic stage,
before they are in any way realised in language (Culioli, 1990, 1999b).
Infraverbal domain Ideational metafunction
Symbolic domain Textual metafunction
Pragmatic domain Interpersonal metafunction
6. Thematic and information structure
The distinction between theme and rheme, and that between given and new are
less theoretically marked than many other theoretical concepts, and are common to
many theoretical approaches. However, the distinction between thematic structure
and information structure is perhaps more clearly brought out in Systemic Func-
tional Linguistics than in some of the other treatments of these notions. Many lin-
guists in the French tradition find themselves with difficult problems because they do
not distinguish between the two types of structure. This is not universally the case,for the distinction does exist in French (e.g. Peeters, 1999) between th�ematisation,which corresponds to thematic structure, and focalisation, which corresponds to
information structure. This distinction could quite easily be incorporated into the
enunciative framework and would solve many of the problems which cause difficulty
at the moment, such as the difference between the kind of �emphasis� implied by
D. Banks / Language Sciences 26 (2004) 391–410 403
thematic position and that implied by new position. This can be illustrated by
the following definition which appears in a glossary printed at the end of two
otherwise excellent books aimed at university students (Lapaire and Rotg�e, 1992,
1993).
Thematic
• For the majority of linguists it is a case of the information status of a given
segment. Any unit which refers to something known (given information) is
thematic. On the other hand any unit which carries new information is de-
scribed as rhematic (Lapaire and Rotg�e, 1992, p. 234; 1993, p. 301). 9
It is clear that since this can only apply to unmarked cases any attempt to apply it
in marked cases is going to lead to trouble.
Systemic Functional Linguistics, on the other hand, has a terminological problemin the term �new�. The new can be subdivided into �new and fresh� and �new but
contrastive�, and indeed probably into other subcategories as well (Banks, 1999b).
However, �new but contrastive� is confusing since it is not �new� in any ordinary sense
of the word. In this ordinary sense �new but contrastive� is a contradiction. Perhaps
systemicists could take a leaf from the French terminological book, with the term
focalisation providing the key. I would like to suggest that the �new� be rebaptised the
�focalized�.In this area, the waters have been further muddied by Adamczewski (1996),
(Adamczewski and Delmas, 1982) who uses rh�ematique and th�ematique in senses
which are rather different to the use of theme and rheme in terms of the textual
metafunction. For Adamczewski, rh�ematique refers to those mental operations which
are primary, elementary or non-presupposing, and he refers to this as Phase 1. The
term th�ematique is the inverse of this referring to mental operations which are sec-
ondary, complementary and presupposing; this is Phase 2. These terms then refer to
different levels of mental work. The rh�ematique, Phase 1, is an elementary level of
mental operations, while the th�ematique, Phase 2, is a more advanced stage, after thebasic mental work has been carried out. Thus he claims that simple verb forms
belong to Phase 1 and are rhematic, while verbs with progressive aspect belong to
Phase 2 and are thematic; similarly, the indefinite article beings to Phase 1, and the
definite article to Phase 2.
9 Original French:
Th�ematique
• chez la plupart des linguistes, il s�agit du statut informationnel [angl. information status] d�unsegment donn�e. Est th�ematique, toute unit�e qui renvoie �a du connu [angl. given information].
Est, en revanche, qualifi�e de rh�ematique [angl. rhematic, information providing] toute unit�e qui
v�ehicule une information nouvelle [angl. new information] (Lapaire and Rotg�e, 1992, p. 234;
1993, p. 301).
404 D. Banks / Language Sciences 26 (2004) 391–410
Although this seems fairly distinct from the concept of thematic structure, there
may be some, albeit fairly tenuous, relationship between them. Lapaire and Rotg�epoint out the possible links between the two terminological usages.
. . . being elementary and factual [they] often go together, in other words what
is rhematic in the domain of operations is frequently so in the domain of infor-
mation. . .. what has already been dealt with by the mind is often known. . .(Lapaire and Rotg�e, 1992, p. 234; 1993, p. 301). 10
7. Text
Both Systemic Functional Linguistics and the Th�eories de l’�enonciation emphasize
the importance of text, but they treat it in different ways. The systemic approach tends
to treat a text as a whole, attempting to bring out the meaning of the text as a unit in
itself. In the enunciative approach what is usually analysed is an individual segment in
the text, but using the co-text to bring out the meaning of that particular segment, aswell as its contribution to the text. This is not an essential element of the approach,
more a question of usual practice, but exceptions (e.g. Souesme, 1998) are extremely
rare. This strongly affects the type of exercise students are required to do, particularly
in competition examinations (concours) such as the CAPES and Agr�egation, which
are competition examinations for the recruitment of secondary school teachers. In the
linguistic component of these examinations, the candidate is presented with a text in
which a number of items are underlined (e.g a determiner, an auxiliary verb, etc); the
underlined items are those which are to be analysed. In principle this exercise isneutral and theory-free. The candidate can use the theoretical framework of his
choice. In practice, the conception of the exercise itself and the items to be analysed
are heavily influenced by the Th�eories de l’�enonciation, concentrating on those fea-
tures where this theory works well, and has significant things to say in relation to
language or the interpretation of the text.
8. Grammar
There has so far been little attempt to provide a complete university level grammar
of English on �enonciative principles, of the type and level that Downing and Locke
(1992) or Morley (2000) attempt for Systemic Functional Linguistics. Souesme, 1992,
while it may not be totally successful, is the only attempt to date to provide a
10 Original French:
. . . �el�ementarit�e et factualit�e vont souvent de pair, autrement dit que ce qui est rh�ematique dans
le domaine des op�erations l�est fr�equemment dans le domaine de l�information. . . ce qui a �et�e
trait�e par l�esprit est souvent connu. . . (Lapaire and Rotg�e, 1992, p. 234; 1993, p. 301).
D. Banks / Language Sciences 26 (2004) 391–410 405
grammar of this type. Other books dealing with grammatical questions (e.g.
Lapaire and Rotg�e, 1991; Bouscaren and Chuquet, 1987; Bouscaren, 1991; Ad-
amczewski and Delmas, 1982) treat a number of isolated problems, such as
BE+ ING (i.e. progressive aspect), the articles, the demonstratives ‘‘this’’ and
‘‘that’’, the distinction between the genitive (�s) and related constructions with
‘‘of’’, etc. For example Bouscaren�s (1991) slim volume (130 pp.) has a short
introduction followed by 20 sections dealing with the following subjects:
1. The simple present
2. The BE+ ING present
3. The preterite
4. The BE+ ING preterite
5. The present perfect: HAVE -EN
6. The present perfect +BE+ ING7. The past perfect: HAD -EN
8. Modality
9. Shall, should, will, would
10. Be going to
11. Is to
12. Have
13. A note on the passive construction
14. Nominal determination15. Quite
16. Some notes on the infinitive and the gerundive
17. Some note on relatives. The difference between the use of that and of who or
which
18. ; and that after declarative verbs
19. Some notes on resultative constructions
20. Levels of utterance
The much longer book by Lapaire and Rotg�e (1991), which weighs in at a hefty734 pages, has, in addition to a theoretical introduction, three parts, the first of
which deals with various markers (determiners, proforms, connectors, prepositons
and adverbial units), and has sections on the following:
This and that
The articles ;, a and the
Any
Some
From ever to every
All
It
Then and There
As and so
And, or and but
406 D. Banks / Language Sciences 26 (2004) 391–410
The second part, which deals with the verbal domain, has sections on:
Simple assertion and the subject–predicate relationshipBE+V-ING
HAVE+V-EN
Modals
Lexical verb to auxiliary do
And the final part, dealing with the complex sentence, has sections on:
A syntactic and functional study of the complex sentence
Elements in WH-/TH-
;/ThatWhich/that
Which/what
Where/when
I hope that these two examples give some idea of the range of subject matter
covered in books of this type. As in the case of the exams mentioned above, it wouldseem that these books treat those areas of the grammar about which the theory has
significant things to say, but this means that fairly large areas are not treated at all.
As a result these books, while being interesting on the points they deal with, tend to
have a piecemeal feel about them.
9. Form(alism)
Linguistic theories fall into three main groups, the functional, the cognitive, and
the formal. However, whichever group one works within, language is encoded in
forms. Consequently, form must have some place in all linguistic theories. The dif-
ficulty with formalist approaches such as Generative Grammar (from a nonformalist
point of view) is in the notion that language consists of nothing but form, or at least
that form is the only thing worth considering.
It is, however, perhaps possible to see some sort of link between Generative
Grammar and the Th�eories de l’�enonciation. Generative Grammar, even from itsearly stages, suggested that language was built up by operating transformations on a
small number of kernel sentences (Chomsky, 1957, 1965). In the Th�eories del’�enonciation, particularly in its Adamczewskian form, basic forms are worked on by
means of mental operations to produce language. More recently, Cotte (1998, 1999)
has developed a th�eorie de la r�e�elaboration.
The reelaboration hypothesis posits that meaning is constructed over time and
that its recent states retain a memory of its former states. It predicts that its
D. Banks / Language Sciences 26 (2004) 391–410 407
various age constituents sometimes follow each other in the string, and it in-
vites us to turn away from a superficial view of syntax (Cotte, 1998, p. 438). 11
Although transfered to a cognitive context, this notion of basic forms which are in
some way modified to produce actual language does appear to bear some debt to the
original transformational concept. This however poses the question of which forms
are to be considered primary. Are there criteria by which these primary forms can be
recognized? Or is it simply a question of intuition; are there some sort of linguisticprimes, rather like mathematical primes, which we simply recognize as such?
Systemic Functional Linguistics seems to take a different tack in that the concept
of system treats alternative encodings as being forms which are simultaneously
available to the speaker. It is not a question of one being derived from the other. One
may be selected more frequently than another, but they exist in parallel.
There may, however, be an area where the systemic approach is not so different.
In some recent work, notably in discussions of grammatical metaphor (Halliday,
1988, 1994b, 1998; Halliday and Martin, 1993; Banks, 1996, 1999a), great play hasbeen made of the notion of �congruent forms�. Halliday (1994a) describes a con-
gruent form as �the less metaphorical variant�, while Thompson (1996) says that �theterm congruent can be informally glossed as ‘‘closer to the state of affairs in the
external world�� �, which seems to have faint echoes of Adamczewski�s Phase 1.Consequently there is a tendancy to think of congruent forms as being more basic or
primary. However, Halliday and Matthiessen admit that
From a purely descriptive point of view, each version is metaphorical from thestandpoint of the other; there is no inherent priority accorded to either (Hall-
iday and Matthiessen, 1999, p. 537).
This admission notwithstanding, they claim that congruent forms are prior from
three points of view: phylogenetically, in the history of the language; ontogenetically,
in the development of the individual; and logogenetically, in the development of text.
I do not see how the phylogenetic priority squares with Halliday�s claim that totally
non-metaphoric texts occur only in the language of young children, which is pre-
sumably reflected in the ontogenetic priority. And even if logogenetic priority is thegeneral rule, surely it is a tendency to which there are exceptions. Nevertheless, it
would seem that there is general agreement on which forms are congruent, despite
the difficulty of establishing criteria for congruency.
I therefore feel that there are theoretical difficulties related to the concept of
congruency as it is used in Systemic Functional Linguistics that have not yet been
11 Original French:
L�hypoth�ese de la r�e�elaboration suppose que le sense se construit dans le temps et que ses �etatsr�ecents gardent la m�emoire des anciens. Elle pr�evoit donc que des constituants d�age diff�erent se
succ�edent parfois dans la chaıne et elle invite �a se d�etourner d�une vision superficielle de la synt-
axe (Cotte, 1998, p. 438).
408 D. Banks / Language Sciences 26 (2004) 391–410
worked out. On the other hand I feel, intuitively, that this notion is basically correct.
What is interesting from the present point of view, is that insofar as the notion of
congruency suggests that there are basic forms and derived forms, this might supply
us with a further point of contact with the Th�eories de l’�enonciation, where the notion
is common that some language forms are produced by mental operations carried out
on more basic linguistic forms.
10. Closing remarks
In this short paper I have tried to show that Systemic Functional Linguistics and
the Th�eories de l’�enonciation are travelling towards each other from opposite direc-
tions, but whereas Systemic Functional Linguistics is an inductive method of en-
quiry, the Th�eories de l’�enonciation are deductive. This places Systemic Functional
Linguistics in a Lockean empirical tradition, and the Th�eories de l’�enonciation in the
cartesian tradition.The Th�eories de l’�enonciation do not conform to the Principle of Falsification and
consequently their findings are of a hypothetical, not axiomatic nature. This in no
way obviates their interest, but it is important to recognize their true nature. The fact
that the findings of the Th�eories de l’�enonciation are hypotheses that cannot be fal-
sified places them in the metaphysical rather than a physical field.
The concept of domains, as developed by Delmas, parallels fairly closely that of
semantic metafunctions in systemic theory. This constitutes a strong point of contact
between these theories. Two of the three domains are virtually the same as the cor-responding metafunctions, while the third deals from a cognitive viewpoint matters
which Systemic Functional Linguistics treats from a functional point of view.
The systemic version of thematic and information structure is one that could be
beneficially integrated into the Th�eories de l’�enonciation without too much difficulty.
Distinguishing between thematic structure and information structure would avoid
many of the problems caused by the fused approach which is commonly used, but
not essential to, the Th�eories de l’�enonciation.Both approaches analyse texts extensively, but in different ways. Systemic Func-
tional Lingusitics analyses whole, or at least large stretches of text, whereas the
Th�eories de l’�enonciation usually analyse single constituents within a co-text.
The Th�eories de l’�enonciation tend to concentrate on specific grammatical ques-
tions rather than the grammar as a whole. Once again this is a question of current
general practice, not an essential element of the theory, so it is not too difficult to see
future developments extending towards a more complete grammar of English.
Both approaches face a potential theoretical problem concerning the criteria for
distinguishing a basic form from which other forms are derived. In the case ofSystemic Functional Linguistics the problem relates to the criteria for distinguishing
congruent forms. The problems faced by each of these theories in this area seem to be
of the same type.
It is my contention that these very different approaches have a number of points
of contact and meaningful differences which could be investigated with mutual
D. Banks / Language Sciences 26 (2004) 391–410 409
benefit. I earlier used the image of two parties on the same road but travelling to-
wards each other from opposite ends, and suggested that if this were the case they
should meet somewhere in the middle. I would like to think that this paper moves us
a little way towards that meeting point.
References
Adamczewski, H., 1996. Gen�ese et d�eveloppement d�une th�eorie linguistique, suivi de les dix composantes
de la Grammaire M�etaop�erationnelle de l�anglais. La TILV, Paris.
Adamczewski, H., Delmas, C., 1982. Grammaire linguistique de l�anglais. Armand Colin, Paris.
Adamczewski, H., Gabilan, J.-P., 1996. D�echiffrer la grammaire anglaise. Didier, Paris.
Ayer, A.J., 1936 [1971]. Language, Truth and Logic. Penguin, Harmondsworth.
Ayer, A.J., 1956. The Problem of Knowledge. Penguin, Harmondsworth.
Banks, D., 1996. The passive and metaphor in scientific writing. Cuadernos de Filolog�ıa Inglesa 5 (2), 13–22.
Banks, D., 1999a. Aspects of the development of grammatical metaphor in scientific writing. Les Cahiers
de l�APLIUT 19 (1), 5–25.
Banks, D., 1999b. Decoding the information structure of journalistic clefts. Interface, Journal of Applied
Linguistics 14 (1), 3–24.
Bennet, J., 1971. Locke, Berkeley, Hume, Central Themes. Oxford University Press, Oxford.
Benveniste, E., 1966. Probl�emes de linguistique g�en�erale, Tome 1. Gallimard, Paris.
Benveniste, E., 1977. Probl�emes de linguistique g�en�erale, Tome 2. Gallimard, Paris.
Berry, M., 1975. Introduction to Systemic Linguistics. In: Vol. 1 Structures and Systems. B.T. Batsford,
London.
Bouscaren, J., 1991. Linguistique anglaise, Initiation �a une grammaire de l��enonciation. Ophrys, Gap.
Bouscaren, J., Chuquet, J., 1987. Grammaire et textes anglais, Guide pour l�analyse linguistique. Ophrys,
Gap.
Caws, P., 1965. The Philosophy of Science, A Systematic Account. D. Van Nostrand, Princeton, NJ.
Chomsky, N., 1957. Syntactic Structures. Mouton, The Hague.
Chomsky, N., 1965. Aspects of the Theory of Syntax. MIT Press, Cambridge, MA.
Chomsky, N., 1966. Cartesian Linguistics. Harper & Row, New York.
Cotte, P., 1996. L�explication grammaticale de textes anglais. Presses Universitaires de France, Paris.
Cotte, P., 1997. Grammaire linguistique. CNED/Didier Erudition, Paris.
Cotte, P., 1998. Have n�est pas un verbe d�action: L�hypoth�ese de la r�e�elaboration. In: Rousseau, A. (Ed.),
La transitivit�e. Presses Universitaires du Septentrion, Lille, pp. 415–439.
Cotte, P., 1999. R�e�elaboration et structure: l�h�eritage dans la langue et en linguistique. ASp, la Revue du
GERAS 23/26, 7–27.
Culioli, A., 1990. Pour une linguistique de l��enonciation, Tome 1, Op�erations et repr�esentations. Ophrys,
Gap.
Culioli, A., 1999a. Pour une linguistique de l��enonciation, Tome 2, Formalisation et op�erations de
rep�erage. Ophrys, Gap.
Culioli, A., 1999b. Pour une linguistique de l��enonciation, Tome 3, Domaine notionnel. Ophrys, Gap.
Culioli, A., 2002. Variations sur la linguistique. Entretiens avec Fr�ed�eric Fau. Klincksiek, Paris.
Descartes, R., 1637 [1965]. Discours de la m�ethode. Bordas, Paris.
Descartes, R., 1637 [1968]. Discourse on Method and the Meditations (trad. Sutcliffe, F.E.) Penguin,
Harmondsworth.
Delmas, C., Adams, P., Del�echelle, G., Girard, G., Lancri, A., Naud�e, G., 1993. Faits de langue en anglais,
M�ethode et pratique de l�explication grammaticale. Dunod, Paris.
Delmas, C., Girard, G., 1993. Grammaire m�etaop�erationnelle et th�eorie des phases. In: Cotte, P., Joly, A.,
O�Kelly, D., Gilbert, E., Delmas, C., Girard, G., Gu�eron, J. (Eds.), Les th�eories de la grammaire
anglaise en France. Hachette, Paris, pp. 97–124.
Downing, A., Locke, P., 1992. A University Course in English Grammar. Prentice Hall International,
Hemel Hempstead.
410 D. Banks / Language Sciences 26 (2004) 391–410
Ducrot, O., 1984. Le dire et le dit. Minuit, Paris.
Ducrot, O., 1972. Dire et ne pas dire, Principes de s�emantique linguistique. Hermann, Paris.
Edmonds, D., Eidinow, J., 2001. Wittgenstein�s Poker, The story of a ten-minute argument between two
great philosophers. Faber & Faber, London.
Gilbert, E., 1993. La th�eorie des Op�erations Enonciatives d�Antoine Culioli. In: Cotte, P., Joly, A.,
O�Kelly, D., Gilbert, E., Delmas, C., Girard, G., Gu�eron, J. (Eds.), Les th�eories de la grammaire
anglaise en France. Hachette, Paris, pp. 63–96.
Halliday, M.A.K., 1988. On the Language of Physical Science. In: Ghadessy, M. (Ed.), Registers of
Written English: Situational factors and linguistic features. Pinter, London, pp. 162–178.
Halliday, M.A.K., 1994a. An Introduction to Functional Grammar, second ed. Arnold, London.
Halliday, M.A.K., 1994b. The construction of knowledge and value in the grammar of scientific discourse,
with reference to Charles Darwin�s The Origin of Species. In: Coulthard, M. (Ed.), Advances in
Written Text Analysis. Routledge, London, pp. 136–156.
Halliday, M.A.K., 1998. Things and relations: regrammaticising experience as technical knowledge. In:
Martin, J.R., Veel, R. (Eds.), Reading Science, Critical and Functional Perspectives on Discourses of
Science. London, Routledge, pp. 185–235.
Halliday, M.A.K., Hasan, R., 1976. Cohesion in English. Longman, London.
Halliday, M.A.K., Martin, J.R., 1993. Writing Science, Literacy and discursive power. Falmer Press,
London.
Halliday, M.A.K., Matthiessen, C.M.I.M., 1999. Construing Experience through Meaning, A Language-
based Approach to Cognition. Cassell, London.
Hospers, J., 1967. An Introduction to Philosophical Analysis, second ed. Routledge & Kegan Paul,
London.
Joly, A., O�Kelly, D., 1990. Grammaire syst�ematique de l�anglais. Nathan, Paris.
Kress, G. (Ed.), 1976. Halliday: System and Function in Language. Oxford University Press, London.
Lapaire, J.-R., Rotg�e, W., 1991. Linguistique et grammaire de l�anglais. Presses Universitaires du Mirail,
Toulouse.
Lapaire, J.-R., Rotg�e, W., 1992. R�eussir le commentaire grammatical de textes. Ellipses, Paris.
Lapaire, J.-R., Rotg�e, W., 1993. S�eminaire pratique de linguistique anglaise. Presses Universitaires de
Mirail, Toulouse.
Locke, J., 1688 [1965]. An Essay Concerning Human Understanding. Dent, London.
Martinet, A., 1962. A Functional View of Language. Oxford University Press, Oxford.
Martinet, A., 1970. El�ements de linguistique g�en�erale. Armand Colin, Paris.
Martinet, A., 1979. Grammaire fonctionnelle du Franc�ais. Didier, Paris.
Matthiessen, C.M.I.M., 1998. Construing processes of consciousness: from the commonsense model to the
uncommonsense model of cognitive science. In: Martin, J.R., Veel, R. (Eds.), Reading Science, Critical
and Functional Perspectives on Discourses of Science. Routledge, London, pp. 327–356.
Morley, G.D., 2000. Syntax in Functional Grammar, an Introduction to Lexicogrammar in Systemic
Linguistics. Continuum, London.
O�Connor, J.D., 1967. John Locke, second ed. Dover, New York.
Peeters, J., 1999. Th�ematisation et focalisation: deux principes distincts et compl�ementaires de
construction du sens. In: Guimier, C. (Ed.), La th�ematisation dans les langues. Peter Lang, Bern,
pp. 45–61.
Pollock, J.-Y., 1997. Langage et cognition, Introduction au programme minimaliste de la grammaire
g�en�erative. Presses Universitaires de France, Paris.
Popper, K.R., 1959. The Logic of Scientific Discovery. Basic Books, New York.
Popper, K.R., 1994. The Myth of the Framework. In: Notturno, M.A. (Ed.), Defence of science and
rationality. Routledge, London.
Recanati, F., 1981. Les �enonc�es performatifs. Minuit, Paris.
Souesme, J.-C., 1992. Grammaire anglaise en contexte. Ophrys, Gap.
Souesme, J.-C., 1998. Approche linguistique d�un passage de The Partner de J. Conradiania Anglophonia
4, 167–179.
Thompson, G., 1996. Introducing Functional Grammar. Arnold, London.