Network.img

Debates and discussions



Dr. Mark Durie
Language Studies
Melbourne University
Parkville, VIC 3052


15 September, 1994

Dear Mark,

Warm greetings from Sydney! I hope you're enjoying spring; here it really feels like spring, with warm and sunny days.

I've read your article in The Age, which I think is a very welcome contribution to the public awareness and discussion of grammar in the educational context - with the exception of your interpretation of systemic theory, description and practice in the context of education. On the positive side from the point view of linguists and educational linguists using and developing systemic linguistics, I think that your involvement in these concerns could serve to counter the argument I come across on various occasions that systemic linguistics isn't real linguistics because it is being applied in an educational context. If you and other linguists of your stature show that the educational context is a valid one for linguistics, that will be a very significant step. But let me take up a number of points where you deal with systemic linguistics.

[1] "... have all introduced an approach known as systemics or functional grammar. Systemics is a set of terminologies for describing texts, developed from the work of Professor Emeritus Michael Halliday of Sydney University." This characterization is quite wrong in a number of respects.

[i] "Systemics" is not the same as "functional grammar". "Systemics" is short for "Systemic-Functional Linguistics"; that is, its phenomenon of study is not just grammar but the whole semiotic complex of language in context. There is a considerable body of work on the strata above and below grammar: on phonology (work by Halliday and others on English intonation, Chinese phonology, Irish phonology, Telugu phonology, Akan phonology, etc.) and on semantics ("discourse semantics") by Jim and many others developing his work and on context. Further, "systemics" includes not only accounts of this phenomenon but also on the one hand explorations of denotative semiotic systems other than language (as in the work by Michael O'Toole, Theo van Leeuwen, Gunther Kress, and Erich Steiner) and on the other hand metatheoretical principles and practices. "Functional grammar" is not the same as "Systemic-(Functional) Grammar", as is emphasized in Halliday's (1985/ 1994) Introduction to Functional Grammar. He introduces the systemic theory of grammar from the syntagmatic angle; this is informed by the paradigmatic, systemic work on grammar, but the systemic part is not set out in the book. He moves in through structures - configurations of functions. And it is of courses this work on 'functional grammar' that has been the main source research and teaching within faculties of education, not the systemic part of grammar.

[ii] Neither "systemics" nor any of its parts such as "functional grammar" is a set of terminologies for describing text. (a) Systemics includes a theory of language in context (as I noted above). A theory is, of course, a semiotic system; in Hjelmslev's terms, one can construe it as a connotative semiotic. Such a semiotic system is realized by a natural language - or by an alternative denotative semiotic, such as a programming "language" (see e.g. Matthiessen, 1993; Matthiessen & Nesbitt, in press). And included within that language will of course be "terminologies". These "terminologies" are thus part of the overall system that realizes a theory such as systemic theory, but on the one hand they are only one part and on the other hand they do not constitute the theory, they help realize it. The terminologies form systems, not sets; and these systems are not only lexical (as "terminology" suggests), but also grammatical - and so also semantic: the semantics of lexicogrammatical wordings, but also discourse semantics. If you analyse passages from a text such as An Introduction to Functional Grammar, you will find that grammatical motifs are critical in the realization of systemic theory - for example, the motif of construing order in language in terms of elaboration. It is possible to realize (aspects of) systemic theory in a denotative semiotic other than a natural language; in fact, this has been done in various partial computational implementations of systemic theory (see e.g. Matthiessen & Bateman, 1991, for discussion and references). Such realizations of systemic theory are patently not sets of terminology. And, of course, if systemics was "a set of terminologies" it could not possibly have been implemented in computational systems: such implementation requires far more than "terminology".

(b) Systemic theory of language is distinct from systemic descriptions of particular languages. I'll come back to the distinction between theory and description below; but here I would just like to note that "systemics" is a resource for doing many things other than "describing texts":


"Describing text" is thus only one of many activities for systemics, as any representative selection of systemic work will make clear (e.g., Halliday & Martin, 1981; Halliday & Fawcett, 1987). Even if you consider only Michael Halliday's work, you would see very clearly that "describing text" is only one strand among many. Among the various other inter-related strands, you will find considerable theoretical work on your very important point that "linguistic habits form people", where "Language is an essential part of the fabric of our personal life and our wider social context" (see e.g. Halliday, 1975, 1978 - further developed in work by Ruqaiya Hasan, Jim Martin, Jay Lemke, Clare Painter among others). And it is precisely the relationship of "describing text" to other activities that makes it significant; it is, for example, often a way into the system.

When I arrived at the University of Sydney, I had many years of personal experience of linguistics programs in Sweden and the U.S. One of the things that struck me about the systemic aspect of the program in our department was that students were introduced to far more language occurring naturally as text than anywhere else I had been - and that they had to be able to deal with these texts in a way I had not encountered before: they had to be able to interpret lots of texts by relating them to the system. But "describing text" was only one aspect of this activity; equally important was to describe the system through text. (I had of course met discourse-based accounts of grammar in West-Coast Functionalism: this approach was beginning to develop when I arrived in the U.S. But linguists never attempted comprehensive interpretations of texts and students were not responsible for dealing with texts in a comprehensive way, relating them to the system in toto.) In my own work, I have of course described lots of texts and continue to do so; yet if systemics was "describing text", I could hardly be said to be doing systemic linguistics (not that that matters as long as my work is of value!): you have probably not come across computational systemic work (see e.g. Matthiessen & Bateman, 1991); it is not concerned with "describing text" in any direct sense (although it does, of course, presuppose such work).

[iii] If we say that systemics is (among other things!) a resource for describing texts, the issue of what "text" means is, of course, central. In systemic linguistics, "text" is a technical term: it is language functioning in context of situation (Halliday & Hasan, 1976; Halliday, 1978; Martin, 1992). This means that it is (a) located stratally within semantics - it is a process/ product of meanings (realized lexicogrammatically, and then, in turn, phonologically or graphologically), and (b) located in terms of instantiation as instance of the meaning potential. I am not quite sure what you understand by "text" in your article - by itself, I think "describing text" would be read in a number of different ways by readers of The Age. But in systemic work, "text" is just one end of the continuum of instantiation; so a description of a text is always related to the overall systemic potential, the "system" (see e.g. Halliday, 1978, 1991). It is, for example, very different from explication de text or currently fashionable 'post-structuralist' work on text without reference to the system. So systemics is (among other things!) a resource for describing systems. What is perhaps a salient characteristic of systemic work is this special foregrounding of the relationship between system and text. I remember when I was learning linguistics in the 1970s, the puzzle of dichotomies such as parole / langue, competence/ performance, and Hjelmslev's way out of this puzzle. And Halliday's work develops Hjelmslev's insight.

[2] "Its terminologies are too baroque and not inclusive enough." Here I am very curious to know what you have in mind! Your input would be very valuable. In general, I have found over the years that there are so many different kinds of demand on terminology that often pull in different directions that it is very hard to satisfy everybody all the time. Many are quite content - perhaps particularly people coming from disciplines with traditions of extensive technical terminologies. Others will complain that terms have been taken over when they ought to have been replaced by new terms; yet others will complain that new terms have been created when existing ones might have been used. If "baroque" means that there are too many terms, I would say the situation is this: by and large, systemics is more comprehensive than other traditions, which tend to focus on some subsystem(s) of language in context. Consequently, it deals with phenomena that are not studied (or are construed in very different ways) and so not named elsewhere (cf. theoretical terms such as logogenesis, semogenesis, instantial systems, autogenetic potential, path augmentation, system traversal, selection expression, instantiation, fractal systems). But there has to be considerable variation here, of course: presentations for educational, computational, linguistic etc. contexts have to be different. Thus few or none of the theoretical terms listed above would be used in a presentation for say primary school teachers. In linguistics, I think it helps to have a background in European structuralism and functionalism as well as American work; but I at least have found that often I cannot take such a background for granted. (If we see systemic theory or any other theory merely as a "set of terminologies", then of course these terminologies will seem baroque - precisely because they are only a small 'part' of the overall theory, one that cannot make sense taken out of its context.)

I also wonder what "not inclusive enough" means. Since you continue with a reference to Anglo-centricity and other languages, perhaps you mean that there are not enough descriptive terms for a variety of languages. But on the one hand, the terminological system is not closed - it is dynamic and open. As new areas of a language or new languages are described, new terms are introduced. For example, descriptive terms you will not find in IFG in accounts of languages other than English include: postpositive verb, Goods, centripetal, Interrogator, Negotiator, etc. etc. etc. - and many terms that are part of general linguistic terminology (construed systemic-functionally) such as adposition, ablative, advanced tongue root, evidentiality, polysynthetic. On the other hand, both descriptive and theoretical terms exist in languages other than English - something I have found very helpful in teaching.

[3] "The Anglo-centric bias of systemics makes it of limited value to school students who learn other languages, since it does not prepare them for the grammatical distinctions that foreign language teachers need to use." Let me first address the notion that systemics is Anglo-centric.

[i] From the very start, systemic linguists have distinguished between the general theory of language and descriptions of particular languages (e.g. Halliday, 1961). The general theory of grammar includes the theory of axial organization (represented by system networks with associated realization statements), the theory of the typological / topological complementarity, the theory of rank (but without a theoretically fixed hierarchy of ranks), the theory of metafunctions, the theory of the relationship between grammar and lexis, the theory of different modes of syntagmatic organization (segmental, prosodic, etc.), the theory of instantiation and the probabilistic nature of the system, the theory of inter-stratal realization, and so on. This theory of grammar is general across languages or "universal" - but people tend to be careful not to suggest the universalist ideology and theory of Chomsky's work and thus often avoid the term "universal". This theory is not at all "Anglo-centric" - by which I assume you mean biased towards English. It has been applied to a variety of languages - apart from English, there is grammatical work on e.g. Akan, Chinese, Finnish, French, German, Gooniyandi, Japanese, and Tagalog, with other languages being added such as Dutch and Vietnamese. The theoretical foundation was from the start anything but "Anglo-centric" - it drew on Firth's system-structure theory (applied to a variety of languages), which in turn drew on the Indian tradition; it drew on Chinese linguistics; it drew on the Sapir-Whorf tradition; and it drew on non-Anglo European work (in particular Glossematics and the Prague School). In fact, many of the points of divergence between Chomskyan theory and systemic theory have to do precisely with the avoidance of the English bias built into Chomskyan theory and its successors - configurationality / the translation of the Subject + Predicate model, Aux/ INFL/ tensed-ness, particular "thematic" roles, Subject as a syntactic primitive, pro drop, etc..

[ii] Categories such as Subject, Mood, Residue, Actor, Process, Goal, Theme, Rheme, Given, New; MOOD TYPE, MOOD TAG, KEY, PROCESS TYPE, PHENOMENALITY, THEME PREDICATION, INFORMATION; indicative, imperative, jussive; material, mental, intensive, meta-phenomenal; absolute theme, interpersonal theme, residue ellipsis; clause, nominal group, prepositional phrase are not theoretical categories; rather they are descriptive categories. That is, they are not part of the general theory of grammar (as categories such as VP, Subject, and Patient are in various other theories), but of descriptions of grammars of particular languages. Here systemic linguistics has been particularistic rather than universalistic. That is, a description of a given language should be constructed in its own terms, without imposing the categories of another language on it. So for example, while ASPECT is the grammatical model of time in an account of Chinese, the temporal system grammaticalized in English is not interpreted in terms of ASPECT, but in terms of (serial) TENSE. Similarly, categories such as Subject, Complement, Mood, Moodtag that are found in systemic descriptions of English should not be exported to other languages unless they can be independently motivated in terms of those languages.

That is not to say that typology of linguistic systems is not seen as interesting or valuable. On the contrary. But in general, typology and comparison have to have a much richer base in comprehensive descriptions of particular languages than is often the case in typological work. As you know, I was trained in (among other things) typological work and I find it very interesting. But at the same time, I think a lot of it suffers from being based on fairly shallow, fragmentary descriptions of languages other than English - even when people make claims about English, I find that they can be pretty distorting; and here there is a danger of a real bias towards English precisely because fragments of incomplete descriptions that may very well have been influenced by English are put together in a typological mosaic. However, I think you and I would be in agreement on this point, wouldn't we? So the issue is rather how to avoid building in a bias towards English. Systemics does this, as I have noted, by separating general theory from particular descriptions and by favouring comprehensive, rich and particularistic descriptions which can then serve as a base for typology and the like. In our teaching we always make this very clear as we do in our writing. But I know that students who are not very familiar with a significant body of systemic work may not have grasped the general principle. For example, I have found people using systemics working on languages other than English can be quite confused about the distinction between theory and description and the role of a description of English in work on other languages. They seem to be making the classical mistake of looking for English categories (such as Range, Subject) in other languages. But it really shouldn't be necessary to make that mistake; many systemic linguists have emphasized the point over the years: see e.g. Halliday, 1993, for a recent statement.

So the description in IFG is particular to English. One might call it Anglo-centric - but this would be very misleading if this is seen as a failure: it is precisely intended to be a description of English, one that brings out its special features (such as the serial nature of the tense system or the interpersonal 'prosody' of Mood + Moodtag in the clause) instead of ironing them out in a universalist frame (or of taking over traditional analyses the way Chomsky did). A comparable work on say Chinese should be Sino-centric in this special sense. In this respect, IFG simply serves a different purpose from say Simon Dik's introductory book or Foley & van Valin (1984). In my experience, universalist proposals involving what would be descriptive categories in systemic work are not particularly convincing; and I do not feel that they guard against a bias in accounts of languages other than English. Rather they tend to blunt language-specific features (often with a bias towards English) - cf. Stan Starosta's contrast between situation-based case roles and grammar-based ones. Systemic generalizations about descriptions of particular languages and systemic typology are definitely part of the overall set of activities people engage in. Publications have appeared and are forthcoming. But systemic typology demands much more groundwork than a lot of general typological work precisely because it presupposes rich language-particular accounts.

In my own experience and work, I initiated a multilingual project in 1990, which is still continuing and expanding. The project involves Chinese, English and Japanese, with related work on French; and it is carried out in collaboration with projects in Japan (Japanese) and Europe (German, Dutch, explorations of Russian). I am supervising systemic work on Chinese, French, Japanese and Vietnamese at a graduate level; undergraduate students in my courses have written term papers on e.g. Chinese, Estonian, German, Japanese, Norwegian, Sinhalese. Jim has done ground-breaking work on Tagalog and is supervising a dissertation on Pitjantjatjara. I have travelled to China, Japan and India to meet systemic linguists working on languages other than English and found very interesting systemic accounts. Apart from the work mentioned above, I have also applied systemics to Akan phonology and grammar.

Systemic work on particular languages is informing work on other languages. This work serves to widen the descriptive imagination, precisely because descriptions of various languages are particularistic. So for example, current work on the textual domain of Vietnamese is being informed by interpretations of Chinese (rather than English) in the investigation of the relationship between the textual and the interpersonal; current work on Irish phonology is again being informed by work on the Chinese syllable; the work on French transitivity has taken place in the context of the transitivity systems of various languages, including Tagalog and Akan; the work on Akan transitivity suggested a different kind of complementarity between the transitive and ergative models from those found in Chinese and English; Bill McGregor's work on process types in Gooniyandi has provided a model of a fairly different kind of system from the one found in e.g. English and Japanese, opening the way for alternative interpretations in various languages. Systemic work on various languages is also part of the process of theory construction: languages other than English are certainly sites for theoretical work and for exchanges for non-English based theoretical traditions. For example, Trevor Johnston's (1992) work on metafunctions in Australian Sign is playing an important role in thinking about metafunctionally differentiated modes of expression and the iconicity of the different modes of expression (see e.g. discussion in Halliday & Matthiessen, forthc.).

So you can imagine I find the notion of Anglo-centric bias misleading as a characterization of systemics in particular: there is clearly a danger of bias towards English in linguistics in general in accounts of various languages - but Halliday's position in laying the foundation for systemic linguistics was designed precisely to guard against this danger. (As you know, he has a extensive knowledge of a variety of languages; but it seems to me he has always resisted the temptation of displaying fragments of various languages that have not been investigated in depth yet and lack comprehensive accounts to back up typological generalizations.)

(By the way, when I on occasion hear from English-speaking linguists that Michael Halliday's work shows a bias towards English, I'm always amused by the contrast with Chinese linguists in China who have told me that there is a bias towards Chinese! By which they mean (among other things) that his descriptions of English have been enriched by the fact that he has worked on Chinese and that he has been influenced by Chinese phonological theory.)

[4] "When students go on to study linguistics at university they will find that these terminologies have little currency at that level either." But these "terminologies" include:

clause, sentence, noun, verb, adjective, particle, phrase; tense, transitivity, determination, modification; Head, Modifier, Subject, Adjunct; phonology, grammar, semantics, structure, class, function, feature, constituency

and many more. Surely these have general currency?

It seems to me that if terms of a systemic or Firthian origin such as collocation, delicacy, polarity, hypotaxis, parataxis, rank, register, metafunction (ideational, interpersonal, textual), Mood + Residue, modal responsibility, modalization, modulation, projection, expansion etc. etc. have no currency in linguistics departments in Australia, there is a serious problem - one of ignorance and neglect. I know you don't intend it to, but your statement could be misinterpreted as resonating with the polemics of "mainstream linguistics" vs. peripheral linguistics - with systemics as an example of the periphery. I think the whole notion of central vs. peripheral is very problematic - far too often it turns out that the "mainstream" is what is dominant in the U.S.; but apart from that, this is where the argument ceases to be concerned with what we can learn about languages & how we can achieve rich informative accounts of various languages and becomes concerned with theories about language instead. So the question is no longer what insight we can gain into a given language but what we can say and read in mainstream linguistics!

It seems to me that as we prepare to move linguistics into the 21st century, we can take stock of the various strands of development in this century and bring out similarities in various traditions that have emerged from rather different points of departure. There is an impressive list of such similarities, including (but not restricted to):



I don't mean that we should try to paper over differences. But I think differences tend to be overt and easy to observe, whereas similarities across different traditions tend to be much more covert so that the real intellectual challenge is to bring out similarities where only differences are thought to exist. Part of the problem lies in the reaction against the Chomskyan tradition as a whole or against earlier 'classical' accounts (such as 'classical' SPE style generative phonology) in the American context. Since the context was the Chomskyan one, alternative proposals tended to be framed as rejections of that context rather than as development of ideas within other contexts. In the process, I think a good deal of information was lost, for example in the failure to develop a network of references to earlier and current European work and to build on them rather than recreate them. As a result, the resonance with systemic-functional ideas is often not recognized in current non-systemic work. So it may seem that systemic ideas have little currency even though they are in fact often very close to non-systemic work.

***

In my comments above, I have not touched specifically on the educational context of your discussion: I think here you will have continued discussions with Fran, Jim and others directly involved in educational linguistics. My own involvement has been more of a service role: I have taught courses in Australia and Japan attended by teachers and other people with educational concerns and I have had many discussions with such people (in my Australian courses, teachers come from Australia, Japan, China, HK, Indonesia); but I have never been active in developing educational linguistics - that is not one of my areas of expertise. I find it very interesting and I have learned a great deal from educational linguists about language, the linguistic construction of knowledge, and so on. I will just note one example where my experience in meeting teachers here in Australia has been completely different from the institutional situation you describe when you write:

"Despite its confident heralds, systemic-functional linguistics does not offer an optional answer to the problem of teaching English grammar in schools. Its real strength in the context of education is its deconstructionist and ultimately political critique of the way texts are used in society. Perhaps it is this political bent that has made systemics so attractive to teachers in education faculties and colleges throughout Australia. The resulting ideological fervor manifests in the energy with which systemics has been foisted on Australian schools."


I think it might appear as if systemics has been foisted on Australian schools from the inevitable superficiality of accounts in the media; but the reality is very different: from my observations over the last 15 years it has been a long process of comparing and contrasting various approaches, typically with the result that systemic work has been blended with other work. And from the very start, the work developed as a dialogue between educators and linguists - not at all as a one-way delivery of linguistic models to be adopted in schools.

In any case, in my own experience at Sydney Uni over the last six years (based mostly on engaging with "grass roots" teachers, not with the educational establishment) is that the real strength of systemic-functional linguistics in the educational context is not what you describe (which is merely one possible direction some people take) but rather that it empowers teachers to engage with language as a resource for the first time in the teaching-learning process. That is what seems to be the real breakthrough for many; and this is reflected in the essays they write - these essays are only rarely concerned with "deconstructionist and ultimately political critique of the way texts are used in society".

I would never in my own teaching foist anything on my teacher students (they do sometimes complain of having other non-systemic material that makes no contact with their educational context foisted upon them, but that is another story). On the one hand, neither I nor any of the systemic linguists I know believe in foisting - only in making available and accessible a body of ongoing work as a resource to be taken up if it is relevant. On the other hand, the teachers I meet are highly articulate and knowledgeable professionals: they know what they need I could certainly not foist anything on them they didn't want even if I had wanted to! They tend to be much more critical and demanding than undergraduate students; and they bring considerable professional experience to the courses. Again and again, I have had the experience of teachers seeking out our courses because of needs that have arisen in their educational context: they want insight into how grammar 'works', they want to teach writing in a variety of genres, they want to be able to select texts in an informed way for teaching languages, they want to be able to understand how knowledge is constructed in various disciplines, etc. etc.

***

Personally I rarely feel the need to assess whole theories of language or linguistic traditions in public, so maybe I have come across as not being interested in non-systemic work. But I am and I have been right from the start in the 1970s. Systemic linguistics had never been part of my departmental environment until I arrived in Sydney in 1988 - not at Lund University in the 1970s, not at UCLA and not in my contacts with linguists at USC (nor in the wider American setting). I grew up in an atmosphere of what might be characterized as choiceless awareness of various linguistic theories and traditions. We were encouraged (by Professor Bertil Malmberg and others) to explore a variety of approaches, try 'get under their skins' and understand them without judging. I came to understand different approaches as basically functional in their own contexts; what appeared to be baroque or bizarre from the point of view of some particular theory turned out to be motivated. Over the years I was taught generative linguistics, European functionalism, typological work, West-Coast functionalism, and so on and I believe I became something of a multi-metalinguist. I came across systemic linguistics in various systemic writings in the mid 1970s in Sweden and it really fascinated me (among other things it solved for me what I thought of as a great disjunction between the insights of the paradigmatic orientation of European structuralism and the syntagmatic orientation of American work) - I was encouraged to pursue it, but never attended a systemic course until I sat in on a ten week course given by Michael Halliday at UC Irvine. By then I had seen a good deal of American linguistics and leading linguists of the day, both generative and functional, and could interpret what Michael presented against that background. My immediate reaction was one of amazement: for the first time I was attending lectures by somebody who was revealing the overall organization of a grammatical system of a language, contextualized by higher-stratal and lower-stratal systems; and the focus was on language rather than on linguistics.

I am in full agreement with Michael's, Jim's, Ruqaiya's, and other systemic linguists' conception of systemic linguistics as an open, dynamic semiotic system, permeable from outside; and I would think you would take a similar view of your work. (I know the systemic position has been criticized as softness & lack of principle, but I think it's easy to clarify that it's not.) All of us have always developed systemic work in dialogue with others; in my own case, this has also included researchers from computer science and systems science. And systemic scholars come from a variety of backgrounds. In the computational context, the systemic contribution has always been permeable and mixed: it has been woven together with other traditions into working systems. It is definitely my impression that the same thing has happened in the educational setting. So I cannot see why contributions to the educational context in e.g. Victoria shouldn't be a blend of complementary approaches - different approaches from linguistics as well as from other disciplines. It seems to me it would be unfortunate if a debate unfolded where alternatives are set up as all or nothing candidates rather than complementarities. So it's sad that The Age framed your contribution as a volley, based on the tired old metaphor of discussion as war. It is precisely the kind of thing that will make people defensive and unwilling to explore complementarities in approaches.

Well, Mark, it would be great if we could follow up our exchanges from the past. I greatly appreciate your willingness to engage. We may disagree - that's as it should be, of course; but we can do so without misrepresenting each other's positions and frameworks. That is very valuable indeed. I think there are many issues we can explore - not only the current context. Certainly I would hope that an interpretation and assessment of systemic-functional work will be part of that exchange.

I am sending you some articles that relate in various ways to the issues I have touched on. Looking forward to future discussions,



All the best,

Christian



References

Bateman, J. 1989. Dynamic systemic-functional grammar: a new frontier. Word 40.1-2 (Systems, Structures and Discourse: selected papers from the Fifteenth International Systemic Congress). 263-286.
Bateman, J. & S. Momma. 1991. The nondirectional representation of Systemic Functional Grammars and semantics as Typed Feature Structures. Technical report, GMD/ Institut für Integrierte Publikations- und Informationssysteme, Darmstadt and Institut für Maschinelle Sprachverarbeitung. Universitaet Stuttgart.
Davey, A. 1978. Discourse Production: a computer model of some aspects of a speaker. Edinburgh: Edinburgh University Press.
Ghadessy, M. (ed) 1988. Registers of written English: situational factors and linguistic features. London: Pinter.
Ghadessy, M. (ed) 1993. Register analysis: theory and practice. London: Pinter.
Halliday, M.A.K. 1961. Categories of the theory of grammar. Word 17.3: 241-92.
Halliday, M.A.K. 1975. Learning how to Mean: explorations in the development of language. London: Edward Arnold (Explorations in Language Study).
Halliday, M.A.K. 1978. Language as social semiotic. London: Edward Arnold.
Halliday, M.A.K. 1984. On the ineffability of grammatical categories. A. Manning, P. Martin & K. McCalla (eds.), The Tenth LACUS Forum 1983. Columbia: Hornbeam Press.
Halliday, M.A.K. 1988. On the language of physical science. In Mohsen Ghadessy (ed.), Registers in Written English: situational factors and linguistic features. London: Pinter (Open Linguistics Series).
Halliday, M.A.K. 1991. The notion of 'context' in language education. In Thao Lê & M. McCausland (eds.), Language education: interaction and development. Proceedings of the Internation Conference held in Ho Chi Minh City, Vietnam 30 March - 1 April, 1990. Published by the University of Tasmania at Launceston.
Halliday, M.A.K. 1993. Towards a language-based theory of learning. Linguistics and Education. Vol. 5.2.
Halliday, M.A.K. 1993b. Systemic grammar and the concept of a "science of language". In Language, , text, context. Tsinghua University Press.
Halliday, M.A.K. & R. Fawcett (eds). 1987. New Developments in Systemic Linguistics Vol. 1: theory and description. London: Pinter.
Halliday, M.A.K. & R. Hasan. 1976. Cohesion in English. London: Longman.
Halliday, M.A.K. & Z. James. 1993. A quantitative study of polarity and primary tense in the English finite clause. J.M. Sinclair, M. Hoey & J. Fox (eds), Techniques of description: spoken and written discourse. London & New York: Routledge.
Halliday, M.A.K. & J.R. Martin. (eds.) 1981. Readings in Systemic Linguistics. London:Batsford.
Halliday, M.A.K. & J.R. Martin. 1993. Writing science. London: Falmer Press.
Halliday, M.A.K. & C. Matthiessen. forthc. Construing experience through meaning: a language-basesd approach to cognition. Berlin: de Gruyter.
Hasan, R. 1985/9. Linguistics, Language and Verbal Art. Deakin University Press (Language and Learning). 1989 edition by Oxford University Press, London.
Hasan, R. 1986. The ontogenesis of ideology: an interpretation of mother child talk. T Threadgold, E A Grosz, G Kress & M A K Halliday. Language, Semiotics, Ideology. Sydney: Sydney Association for Studies in Society and Culture (Sydney Studies in Society and Culture 3). 125-146.
Hasan, R. 1990. Semantic variation and sociolinguistics. Australian Journal of Linguistics. Volume 9. 2: 221-277.
Henrici, A. 1965. Some notes on the systemic generation of a paradigm of the English clause. Working Paper for the O.S.T.I. Programme in the Linguistic Properties of Scientific English. Reprinted in Halliday & Martin (eds.).
Johnston, T. 1992. The realization of the linguistic metafunctions in a sign language. Language Sciences.
Kasper, R. 1987. Feature Structures: A Logical Theory with Application to Language Analysis. University of Michigan Ph.D. Dissertation (Computer and Communication Sciences).
Kasper, R. 1988. Systemic Grammar and Functional Unification Grammar. In Benson & Greaves (eds.). Also as ISI/RS-87-179.
Martin, J.R. 1992. English text: system and structure. Amsterdam: Benjamins.
Matthiessen, C. & J. Bateman. 1991. Text Generation and Systemic Linguistics: experiences from English and Japanese. London: Pinter.
Matthiessen, C. & C. Nesbitt. in press. On the idea of theory-neutral description. In R. Hasan (ed).
Matthiessen, C. 1991. Language on language: the grammar of semiosis. Social semiotics: a transdisciplinary journal in functional linguistics, semiotics and critical theory, Vol. 1:2.
Matthiessen, C. 1993. The object of study in cognitive science in relation to its construal and enactment in language. Cultural Dynamics 6.1.
Matthiessen, C., K. Nanri, L. Zeng. 1991. Multilingual resources in text generation: ideational focus. In Proceedings of the 2nd Japan-Australia Symposium on Natural Language Processing, Japan, October 1991.
Mellish, C. 1988. Implementing systemic classification by unification. Journal of Computational Linguistics. 14.1: 40-51.
Nesbitt, C. & G. Plum. 1987. Probabilities in a systemic grammar: the clause complex in English. In R.P. Fawcett & D.J. Young (eds).
Painter, C. 1984. Into the mother tongue. London: Pinter.
Painter, C. 1993. Learning through language: A case study in the development of language as a resource for learning from 2 1/2 to 5 years. Ph.D. dissertation, Department of Linguistics, University of Sydney.
Painter, C. in press. The development of language as a resource for thinking: a linguistic view of learning. In R. Hasan & G. Williams (eds), Literacy in society.
Patten, T. 1988. Systemic text generation as problem solving. Cambridge: Cambridge University Press.London: Longman (Applied linguistics and language studies).
Patten, T. & Graeme Ritchie. 1987. A formal model of Systemic Grammar. In Gerard Kempen (ed), Natural Language Generation. Dordrecht: Martinus Nijhoff.
Winograd, T. 1972. Understanding Language. Edinburgh: Edinburgh Universtiy Press.