Back to abstract list
Plenary abstracts
Abstracts (papers)

Plenary abstracts

Fries, Peter H., Central Michigan University, USA.
What Makes a Text Coherent?

The relation between coherence of a text and the features of the language of that text (such as the structure of the text, the cohesive ties within the text, cohesive harmony, etc.) has been a source of a great deal of discussion. The point is often made that coherence is a mental phenomenon while features of language such as cohesion are matters of language form.  This position can be criticized both for its view of coherence and for its view of features of language.  On the one hand, I will suggest that coherence is a social phenomenon jointly constructed by the participants in a social interaction.  On the other hand, features of language do not contribute to the perception of coherence simply because of their nature as form.  To be relevant to coherence, features of language must be interpreted.  For example, a cohesive tie does not exist simply because one of a set of words (e.g. a pronoun) occurs in a particular sentence, but because that word is interpreted with reference to the information provided by other portions of the same text.  Cohesive ties thus entail semantic interpretation.

In this paper, I will problematize the notion of text, and then examine a number of non-prototype texts (such as dictionaries and tests) to show that four factors influence the perception of coherence.

1. Is the language used referable to some sensible social interaction among the participants?
2. Does the text have an appropriate generic structure?
3. Are there features of language in the text which link the various portions to one another?
4. Does the producer of the text inscribe the text with appropriate and reasonably self-consistent evaluations?
 

Hoey, Michael, University of Liverpool, UK.
The textual dimension of lexical choice (and I’m not referring to cohesion)

In this paper I want to suggest that the choice of what to thematise in a text is constrained not only by the previous history of a word in the co-text but by the word’s  history of usage in the language as a whole. My argument will be developed in three stages. In the first I argue that there is a relationship between lexis and grammar such that lexical choices have grammatical constraints placed upon them while grammatical choices are constrained by the lexical choices that are being made. This relationship is in some respects ‘lexis as most delicate form of grammar’ but not as usually conceptualised. It is best handled as a manifestation of what Firth termed ‘colligation’.

The second stage of my argument will be that, as one manifestation of colligation, certain words favour thematic placement while others disfavour such placement. Where words are polysemous or homonymous, one sense is likely to be available for thematisation while the other is forced to avoid it. I shall argue that this is one aspect of a more pervasive sense-discriminatory function of colligation.

Thirdly I shall seek to demonstrate that certain words, once thematically placed, favour paragraph initial position in writing, while others avoid such positioning or favour paragraph initial position. Claims for the structural status of paragraphs can be shown to be false in the light of the evidence I shall adduce. On the basis of the demonstrable relationship between this aspect of colligation and paragraphing, it is possible to speculate on the way it might operate in speech.

Fourthly I shall seek to show that certain paragraph-initial words have a preference for text-initial positions. I show how the first sentence of a popular science text has several ‘text-initial’ choices in its first sentence and conclude by suggesting that we may need to talk of colligational prosody as a major influence on both grammar and texture.
 

van Leeuwen, Theo, University of Cardiff, UK.
Evaluation and legitimation in discourse – a multimodal approach.

The paper considers evaluation in the context of an emerging theory of discourse as the recontextualisation of social practice.

Building on the work of J.R.Martin and Basil Bernstein, the assumption is that in constructing representations of the world, we draw on specific discourses, that is, specific knowledges of social practices which have been recontextualised for the purposes of specific contexts. These discourses involve, not only a version of the social practices in question, but also a set of associated  evaluations, purposes and legitimations.

The paper first of all explores the borderline between legitimation and evaluation, arguing for the category of ‘moral’ evaluation, a form of evaluation which has a distinct legitimising function. In ‘moral’ evaluation, the to-be-legitimated element is referred to by means of an abstract expression which distils from it a quality that links it to a discourse of values. However, in Western societies evaluation has increasingly been separated from morality and this leaves its traces in discourse. Three key types of such non-moral evaluation are introduced, affective reactions, ratings and neutralised moral evaluations.

The paper then explores the realisation of discourses generally, and of evaluations in particular, in other semiotic modes. Colour, texture, and other aspects of materiality are interpreted as associated with discourses of value, and fulfilling a function of moral evaluation. The key example will be colour, as used in the realisation of discourses of childhood and discourses of family life.
 

Astracts (papers)

Andersen,Thomas, University of Southern Denmark.
The System of Theme in the Danish Clause.

This paper examines the primary textual ressource of the written Danish clause, i.e. the system of THEME.

A comprehensive system of THEME for any language should be able to account for Theme^Rheme structures in all possible instances, and this is precisely what the system of THEME, proposed in this paper, does with regard to Danish.

In Danish the notion of topical Theme (as opposed to interpersonal Theme) does not exist, since a purely interpersonal konstituent (e.g. an Adjunct of modality) fully can exhaust the thematic potential of the clause. Thus, the paper differs only between the two notions Theme and textual Theme and proposes a description of Theme based on the idea that the Theme in the Danish clause extends from the beginning of the clause and up to (and including) the first element that has a function either in MOOD or in TRANSITIVITY.

The system of THEME for Danish is by no means a simple one, since there are quite a few clause types – differing in structure – in Danish. The primary distinction is between main and subordinate clauses. Each of these clause types realizes the thematic potential in their own way. In the paper these different ways are mapped out.

As perspectives on the system of THEME, the paper examines marked vs. unmarked Theme  and Theme^Rheme structures in elliptical clauses.

Aronsson, Mia Bostrom, Goteborg University, Sweden.
Thematic Differences between Swedish Advanced Learner Writing and Native Speaker Writing.

Texts produced by non-native speakers often differ from native speaker writing in various ways. When it comes to advanced learner writing, the deviations are often discussed in terms of differences in frequencies of certain categories of words or certain structures. Another explanation studied by, for example, Mauranen (1996) is that non-native speaker texts differ from native speaker texts as regards the thematic progression. An additional aspect, which will be discussed in this study, is that learners’ tendencies to show a preference for certain constructions, such as impersonal constructions or focusing devices, may effect the themes of their texts. The study is concerned with thematic differences in argumentative writing produced by Swedish university students of English and native speakers of English. It looks into how differences between Swedish advanced learner writing and native speaker writing may be reflected in the themes and what effect the differences may have on the texts. The study briefly looks into the possibility of transfer being involved in the thematic differences between Swedish advanced learner writing and native speaker writing as a result of differences between Swedish and English sentence structures. The material studied consists of argumentative essays taken from the Swedish component of the International Corpus of Learner English and comparable native speaker writing.

Reference
Mauranen, Anna. 1996. “Discourse Competence – Evidence from Thematic Development in native and Non-native Texts”. In Ventola, Eija and Anna Mauranen (eds). 1996. Academic Writing. Intercultural and Textual Issues. Amsterdam/Philadelphia: John Benjamins Publishing Company. pp 195-230

Banks, David, Université de Bretagne Occidentale, France.
Degrees of Newness.

According to standard systemic theory, the climax of new (focalized) information is signaled by the tonic accent. The tonic accent is also the site of tone or pitch movement. Systemic theory provides for five simple tones, fall, high rise, low rise, fall-rise, and rise-fall. My hypothesis is that different tones indicate different degrees of newness or focalization, low rise having the lowest degree of newness, and fall-rise and rise-fall having the highest degree of newness. This hypothesis is tested against some data, which consists of recordings of four readers reading extracts from newspaper articles. Although the nature of the data is such that it means that it is oral but non-spontaneous, it also eliminates some of the unnecessary complications of spontaneous oral data, and allows comparison between different individuals. As might be expected, the initial hypothesis is only a rough approximation of the true situation.

Bowcher, Wendy, Tokyo Gakugei University, Japan.
Theme-Rheme and Given-New Patterns in Radio Play-by-Play  Sports Commentary.

THEME and NEW structures represent two waves of informational prominence built into the clause. The 'carrier' of the Thematic wave of prominence is 'constituency', and Thematic prominence is realised by theordering of constituents in the clause. On the other hand, the carrier of the New wave of prominence is 'prosody', and the element that is considered to be 'New' is the element which is assigned tonic prominence by the speaker.

This paper presents an analysis of Theme and Rheme and Given and New in play-by-play radio sports commentating. The aim is to describe 1) the kinds of choices made in each of these systems, and 2) the specific patterns of informational peaks formed by these choices in the register of play-by-play talk.

The data used for this paper consist of two extracts from two different professional Australian Rugby League radio commentaries of the same game. A definition of 'play-by-play talk' is first offered. The talk classified as play-by-play is then analysed in terms of Theme and New structures. It will be shown that in radio play-by-play talk Theme and New carry complementary types of information. For instance, one of the findings from the analysis is that in terms of experiential information, both Theme and New informational peaks in play-by-play talk often feature 'a player', but when in Thematic position, the grammatical category of the player is that of participant, whereas when highlighted as New information, players are located in circumstances of location. Furthermore, on the whole, circumstances feature more prominently in the New information peaks than in Thematic position. One of the contextual motivations behind the organisation of play-by-play talk appears to be the effect of simultaneously relaying events; the talk is organised in a way which mirrors the sequence of events taking place in the game itself. That is, for instance, the location to which the ball is passed in the actual game is presented in the unmarked New position in the clause and not as the point of departure in the clause (i.e. as Theme).

Caffarel, Alice, University of Sydney, Australia
The Representation of a Second-order Semiosis in Camus' L'Etranger.

Numerous remarks, comments and reflections on l'Etranger have flowed at regular intervals since Sartre's first observations in Situations I (1943), "a clause from l'Etranger is an island. And we cascade from clause to clause, from nothingness into nothingness." The widespread interest in Camus' narrative method in L'Etranger was catalysed by his choice of the compound past as narrative past which 'stands out the established tendency' in French literary narration (See Hasan, 1985). The unexpected tense patterning in l'Étranger serves on the one level to structure the text temporally but on an other level to convey
higher-level meanings. We will see that the consistent choice and repetition of certain grammatical resources serve to convey a second-order semiosis, that is, 'meanings that are free from the control of the semantics and domain of choice in their own right' (Halliday, 1982). In this paper I would like to explore how the foregrounding of ideational (logical and experiential) grammatical patterns in L'Étranger contribute to the construal of Camus' philosophy of the absurd.

References:
Camus, A. 1942. L'Étranger. Gallimard. Paris
Hasan, R.1985. Linguistics, language and verbal art. Deakin
Unniversity,Victoria.
Halliday, M.A.K. 1982. The de-automatisation of grammar: from
Priestley's An
inspector call'. In Language form and linguistic variation: papers
dedicated
to Angus MacIntosh. J.N Anderson(ed.) Benjamins: Amsterdam.
Sartre, J.P. 1943. Situations 1.
 

Coffin, Caroline, Hewings, Ann & Mayor, Barbara, Open University, U.K.
The Textual and the Interpersonal: Theme and affect in student writing.

Effective academic argument is traditionally perceived as a dispassionate and objective process. As a consequence, university students, particularly those located within the arts and social sciences, are encouraged to produce written arguments that appear to be free of personal bias and judgement. Linguistic research (e.g. Coffin 1997, Hewings forthcoming ), however, has shown that the apparent 'objectivity' of such texts is merely a 'rhetorical trick' whereby the subjective stance or position of an individual student in relation to a particular issue is obfuscated and backgrounded through particular patterns of lexical and grammatical choices.

 It was with this understanding of how written academic argument is traditionally valued that we approached a corpus of texts produced by internationally based non-native speakers of English. These texts were written as part of an English test designed to measure candidates' ability to successfully participate within English medium tertiary education. Specifically, candidates were requested to respond to a controversial issue by writing a short argumentative essay. The expectation was that a linguistic analysis of successful essays (i.e. ones that scored
high grades) would show a favouring of lexical and grammatical patternings construing the writer as an objective and neutral arbiter of knowledge. On the contrary, using the tools of systemic functional linguistics, a close analysis of thematic choices revealed that in many of the successful essays the subjective nature of the argument was made explicit. Such interpersonal flagging raises several questions:

To what extent do the texts measure candidates' control of academic argument and to what extent do they provide practice in the types of written argumentation traditionally favoured within Western based tertiary education?

To what extent are styles of argumentation influenced by the context (both cultural and situational) in which candidates write?

Given the changing environment of tertiary education and post modernist scrutiny of, and challenge to, the discursive construction of knowledge, is there a case for encouraging 'subjective' rather than 'objective' orientations to knowledge (as exemplified in the test papers)?

This paper will explore such questions by drawing on the results of a detailed SFL analysis of theme in 100 essays. It will also raise methodological issues concerning Theme analysis, for example, the units of analysis, theme delimitation and thematic progression.

Cummings, Michael, York University, Canada.
Cohesion and Genre in Continuous Text.

In Systemic Functional description, one of the fundamental components of textual cohesion is the iteration of a reference, whether realized by the repetition of the same lexical item, or by the use of pro-forms.  Some kinds of text are inherently more iterative than others.  This paper reports the results of a study of the iterative properties of different stretches of text belonging to different genre types (as defined by Systemic Functional Linguistics) within a more continuous text, in this case, a 19th-century novel.  A previous study (?Iteration and genre in a 19th-century novel?, forthcoming, 2001) used a computer program to divide the text of the whole novel into equal segments and count the iterations within  each segment.  The results were tabulated and correlated with the predominant genre type which each segment represented.  The current study improves on the methodology by segmenting the text by hand into passages exclusively characterized by a single genre type.  Then the iterative properties of the segments are determined by a computer method, and tabulated.  The results offer means of distinguishing narrative, descriptive, expository, monologue and dialogue genre types along the axis of iterative cohesion.

Davies, Martin, U.K.
"Two truths ... of the imperial theme."

Two truths are told,
As happy prologues to the swelling act
Of the imperial theme . . .
[Aside] This supernatural soliciting
Cannot be ill; cannot be good:

Macbeth, I.iii.127 ff.

Different ways of defining Theme have been in conflict ever since Plato, and in systemic-functional linguistics, despite Margaret
Berrys warnings against relying too much on definitions (Berry 1975:63), in the process of working out how to identify Themes in texts, we have often been in conflict over what definition to apply. The conflicts become unnecessary if it is recognized that     different definitions derive from different frameworks, the frameworks themselves deriving from different purposes and  philosophies. This paper tries to offer a "happy prologue", as a way of reconciling different views, in order that people can go about their business peaceably, and does so by approaching the questions of "aboutness" and "starting-point", or "ground", and by considering the question of "enabling" with reference to Algy's bulge.

Berry, H.M.: Introduction to Systemic Linguistics, Vol. 1 Batsford 1973, reprinted Nottingham (date unknown).
 

Fontaine, Lise, Université de Bordeaux 2, France.
Textual Challenges in Recursive Texts.

Recursion is defined in mathematics as the mechanism by which a function at step "n" is defined by its value at step "n-p". Email texts at level n that are defined by embedded subparts defined at level n-1, n-2 etc. are of a recursive nature. The result is an embedded dialogue.  A recursive text, as found in personal email messages for example, contains fragments that are intended to
represent the speech currently being addressed. The recursive nature of these email texts can be problematic in discourse analysis, posing specific problems for thematic analysis.   The first set of problems are faced when forced with the decision of what should be analyzed.  There are essentially three possible ways to treat the text before even beginning the analysis.  Each participant's speech can be taken in isolation, considering only what they actually said (or typed) and thereby converting it to a purely monologic text. The text of the message can be left intact, complete with referenced sections removed, rearranged and commented on. Finally, the text can be completely reconstructed, including the full utterances from the previous message. In the latter case, we must then consider possible limits in the depth of the reconstruction since the recursivity of the text could bring you back five or six messages if not more. This paper explores these problems with respect to text and proposes some solutions for textual analysis in recursive email texts using a Systemic Functional Linguistics approach.

keywords:  SFL, recursion, thematic progression, textual analysis, theme,
discourse, embedded dialogue

Ghadessy, Mohsen, University of Brunei Darussalam, Brunei & Gao, Yanjie, XJTU, China.
Simplification as a Universal Feature of the Language of Translation.

A common belief among a number of applied linguists working with parallel texts (texts from a source language, L1, and their translations into a target language, L2) is that “translated language is different from the original language” (Mauranen, 1998, p.160). A related research question is “Are translated texts different from comparable texts in the target language as well?” One way to answer the above question is to establish “translation universals” which make translated texts different from comparable texts in the target language. The process of simplification of translated language has been mentioned as one such universal feature (Baker, 1993,1995; Laviosa-Braithwaite, 1996). The purpose of the present study is to investigate one of the consequences of the process of simplification, i.e. a reduction in lexical density, in a number of texts and their translations from English into Chinese. It will be hypothesized that translated texts into English and translated texts into Chinese, in comparison with similar monolingual texts in the two languages, will be less lexically dense. A Systemic Functional Grammar (SFG) model will be used for defining and calculating lexical density. Some implications of the findings for teaching translation will also be discussed.

Gouveia, Carlos A.M., University of Lisbon, Portugal & Barbara, Leila, Catholic University of São Paulo, Brazil.
It is not there, but [it] is cohesive: the case of pronominal ellipsis of subject in Portuguese.

Reference has been traditionally considered as either exophoric or endophoric, with the latter being of two different kinds, anaphoric and cataphoric, depending on whether the reference is to previous text or subsequent text, respectively. This framework has proved adequate for English, and for many other SVO languages, but not for all of them, as is the case of Portuguese.

The purpose of this paper is to raise arguments for a re-definition of the reference system that may also account for languages such as Portuguese, a subject optional language. In Portuguese the absence of a pronominal subject is also a cohesive phenomenon at the level of reference, thus determining the need for the consideration of other types of reference apart from the traditionally established ones.

We shall be examining data from both the Brazilian and the European varieties of Portuguese drawn from the database of the bi-national Project "Discourse and Social Practice in Lusitanian and Brazilian Companies", involving the University of Lisbon, the Catholic Universities of S. Paulo and of Rio de Janeiro and the database of the "DIRECT Project — Towards the Language of Business Communication" (Catholic University of São Paulo).
 

Hansen de Chambouleyron, Ana, Universidad Nacional de Cuyo, Argentina.
Non-finite Clauses in Thematic Position.

Non-finite Clauses in Thematic Position.

This paper is part of a bigger project, called REDACTE, which aims at systematising linguistic tools to assist Spanish-speaking scientists to improve the writing of their papers in English and thus have more chances of being accepted in international forums. The purpose of this paper in particular is to analyse the conditions of occurrence of non-finite clauses in thematic position, in particular the performance of ING clauses.

The study of the specialised bibliography that was consulted showed that it is the To infinitive clause the type of non-finite construction that has been most widely analysed, while the ing and en clauses have not been so explored yet. The hypotheses claim that non-finite constructions, in particular ing clauses, have multiple discourse functions and establish more than one logico-semantic relationship when in thematic position.

The paper is written from a systemic functional perspective. The Textual metafunction, and in particular the thematic position, is considered. The study was based on the analysis of 82 research articles from 5 different disciplines and on previous works by Sandra Thompson and Peter Fires.

The results of the statistical analysis showed that the use of non-finite clauses is significant in certain cases. The hypotheses are satisfactorily proved.

The conclusions suggest in which cases and for what reasons ing clauses in thematic position are more or less used. There are recommendations to make the benefits of the textual metafunction more widely known and applied in academic environments to optimize scientific messages. There is a call of attention to the use of ambiguity.

Hasselgård, Hilde, University of Oslo, Norway.
Temporal and Spatial Adjuncts as Elements of Texture.

It has often been noted that adjuncts of temporal or spatial location, particularly in clause-initial position, can contribute to discourse structure (e.g. Hasselgård 1996). As Themes, location adjuncts are topical and marked. They give a temporal/spatial perspective on the clause message, and instruct the hearer to construe the appropriate ideational frame of reference (cf. Matthiessen 1995). E.g. All night in the example places the ideational content of this clause further back in time.

Location adjuncts can also contribute to the cohesion of a text (looking back rather than forward), for instance as part of a set of similar expressions (lexical cohesion, e.g. again and once more below) or by making comparative reference to a previously mentioned location. In the example, the thematic time adjuncts are part of a strategy to place each event in time, so as to restore chronology (cf. Enkvist 1976, Virtanen 1992).

Using material from the International Corpus of English I want to explore how such adjuncts contribute to texture in speech and writing. The focus is on circumstantial adjuncts of location, though adjuncts of extent and modal and conjunctive temporal adjuncts may also be relevant.

But she wanted, needed, to see Lesley again. All night she had tossed and turned, dreaming sometimes of being at a disadvantage, afraid, and helpless to help herself. In the dawn greyness she had listened to the birds' first brave cheepings and had given way to her overpowering urge to confront Lesley once more. And now she was heading for Manchester… <ICE-GB:W2F-003 #6-11>

References
Enkvist, Nils Erik. 1976. Notes on valency, semantic scope and thematic perspective as parameters of adverbial placement in English. In Enkvist, N.E. & V. Kohonen (eds.), Reports on Text Linguistics: Approaches to Word Order. Åbo: Åbo Akademi. 51–73.
Hasselgård, Hilde. 1996. Where and When: Positional and Functional Conventions for Sequences of Time and Space Adverbials in Present-Day English. Oslo: Scandinavian University Press, Acta Humaniora.
Hasselgård, Hilde. Forthcoming: The role of multiple Themes in cohesion. To appear in Aijmer, Karin (ed): Discourse Patterns in Spoken and Written Corpora. Benjamins.
Matthiessen, Christian. 1995. Theme as an enabling resource in ideational ‘knowledge’ construction. In Ghadessy, Moshen (ed.). Thematic Development in English Texts. London: Pinter. 20–54.
Virtanen, Tuija. 1992. Temporal adverbials in text structuring: On temporal text strategy. In Lindberg, A.C, N.E. Enkvist and Kay Wikberg (eds.), Nordic Research on Text and Discourse: NORDTEXT symposium 1990. Åbo: Åbo Academy Press. 186–197.

Lavid, Julia, Universidad Complutense de Madrid, Spain.
Building Textual Resources for Multilingual Generation: A function-typological approach.

This paper reports on the construction of textual resources for  multilingual generation (MLG) with special reference to the phenomena of theme and focus in English and Spanish. The theoretical framework for this work is Systemic Functional Linguistics (SFL), more specifically, the extensive cartographic grammar of the English language developed by C. Matthiessen (1995) and the computational context provided by KPML (Komet Penman Multilingual), an extensive grammar development environment and generation engine that supports large-scale multilingual development (Bateman 1997). The paper discusses the computational specifications used for these textual phenomena in the systemic-functional grammar of English "Nigel" contained in the KPML environment, suggesting extensions and improvements and ways to account for these phenomena in Spanish. Following recent research on MLG (Bateman et al. 1991, the approach adopted rests on the notion of functional typology, which is also the basic design principle behind the Nigel grammar. This design allows the integration of different languages into one multilingual grammar, where functional commonalities are represented systemically as shared options, and structural divergences capture the different realizations of these options characteristic of each language. Building on previous corpus-based empirical studies (Lavid 1998, 2001), the paper explores the functional commonalities with respect to the textual phenomena mentioned before, as well as the different realization probabilities characteristic of each language.
 

REFERENCES

Bateman, J.A., C. Matthiessen, K. Nanri, and L. Zeng. (1991) Multilingual text generation: an architecture based on functional typology. Proceedings of the International Conference on Computational Linguistics. Penang, Malaysia, pp. 389-407.
Bateman, J.A. (1997) Enabling technology for multilingual natural language generation: the KPML development environment. Journal of Natural Language Engineering 3 (1): 15-55.
Lavid, J. (1998) The relevance of corpus-based research for contrastive linguistic and computational studies: thematization as an example. IV-V Jornades de corpus lingüístics: els corpus en la recerca semàntica i pragmàtica. Publicaciones del Institut Universitari de Lingüística Aplicada.
Lavid, J. (2001) Using bilingual corpora for the construction of contrastive generation grammars: issues and problems. Proceedings of the International Conference on Corpus Linguistics 2001, Lancaster University, 30th March to 2nd April 2001.
 

McCabe, Anne, Saint Louis University, Madrid, Spain.
Thematic Progression Patterns and Text Types in History Textbooks.

This study takes an in depth look at the relationship between text types, or rhetorical patterns, and thematic progression patterns. Some researchers suggest a tendency for correlation between the two, e.g. explanation and exposition tend to prefer the simple linear pattern (Nwogu, 1990), while the constant theme pattern seems to occur frequently with descriptive (Nwogu, 1990; Bloor and Bloor, 1995; Holloway, 1981) and narrative (Fries, 1983) texts.
The present study examines the relationships between thematic progression patterns and text types as construed in a corpus of history textbooks. An analysis of these relationships points towards greater complexity in the correlation between text type and thematic progression patterns. It shows that text types can be matched with different thematic progression patterns, e.g. a stretch of narrative can be organized using an extended simple linear pattern, or a stretch of explanation using a predominately constant theme pattern, depending on the function of the text type within the overall textual configuration. This is illustrated with data from history textbooks in English and Spanish, which show that the concern in history writing with analysis and interpretation of past events calls for a complex interplay of rhetorical and thematic patterns.

Morley, David, University of Strathclyde, U.K.
Analysing syntactic function: elements of clause structure in systemic grammar.

Existing classification schemes for analysing elements of clause structure include quite different approaches to verbal complementation. Yet, every one of the varying criteria used to determine classes of clause element raises its own problems.

Working against the semantic background in systemic grammar, this paper develops further the discussion of clause structure in Morley (2000) and elaborates the concept of nominal, adjectival and adverbial syntactic functions which reflect ideational semantic elements. These functions are then applied to redefining the elements of clause structure. The resulting
framework incorporates a number of significant changes, such as a (revived) distinction between object and complement, the abandonment of the indirect object as an element of structure, the introduction of an 'adjectival element', and the separation of the 'circumstantial adverbial element' from adjunctive adverbials.

Reference
Morley,G.D.(2000) Syntax in functional grammar: an introduction to lexicogrammar in systemic linguistics. London & New York: Continuum.

Nieto Díez, Francisco & Arús Hita, Jorge, Universidad Complutense de Madrid, Spain.
English and Spanish Structures: The textual metafunction as a contrastive tool for the analysis of languages.

This paper uses SFG to explain differences in the way English and Spanish construct semiotic reality. Contrastive pairs such as (1-2) show important differences when looked at from an exclusively structural viewpoint. On the other hand, aspects of the textual metafunction will help explain such differences. For instance, the notion of end-weight explains why the unmarked English realization (1a) does not have a sequential equivalent in the same process in Spanish, where it is only realized through the sequence, marked in English, Actor ^ Process ^ Range ^ Recipient (1b, 2b): Spanish, as opposed to English, realizes the Recipient, a María, as a prepositional phrase regardless of its position. The increase in communicative weight caused by the preposition pushes this participant to the end of the sequence. The information structure, which could be affected by this movement, stays the same as in the English counterpart thanks to the reduplication of the Recipient through the pronoun le. This permits the Recipient to be in Given position with respect to the Range.

(1) (a) Peter gave Mary a kiss  (unmarked)
 (b) Peter gave a kiss to Mary  (marked)
(2) (a) ?Pedro le dio a María un beso (marked and bizarre)
 (b) Pedro le dio un beso a María  (unmarked)

 A number of examples will be contrasted to see how far the textual metafunction can be exploited as a contrastive tool.

Ormrod, Janet, ENST de Bretagne, France
Theme in Breton: A comparative study of a text in Breton and in French by P.-J. Hélias.

Breton has an extremely flexible syntax and is generally classified as a VSO language. However, it would be more accurate to call it an an xVSO language since the initial position in the main clause, the Theme in Hallidayan terms, is occupied by the information which the speaker wishes to focus on or to accentuate. In this (xVSO) - unmarked – structure, we can find Adjuncts, Modifiers, and so on as the initial element (occurring in around 60% of main clauses). This phenomenon is referred to as ‘topicalisation’ or ‘thematisation’ by some French linguists. The second position is occupied by the finite verb and the form of the verb is conditioned by the syntax of the initial element.

We shall study this phenomenon through a comparative analysis of the thematic structure of a text originally written in Breton, and later in French, by the same author Per-Jakez Hélias (taken from his famous chronicle ‘Marc’h al Lorc’h /Le Cheval d’Orgeuil/The Horse of Pride’). We shall investigate how the thematic structure typical of the Breton language maps onto the information structure, and how marked information is presented in the language.

Ribeiro Pedro, Emilia, University of Lisbon, Portugal.
Cleft Sentences in Business Meetings in Brazilian and European Portuguese.

The aim of this paper is to look at the way cleft sentences (as formal grammar used to call them) are used in four business meetings, two of them involving Brazilian speakers and the other two involving Portuguese speakers. Ishall be trying to determine how these specific types of wordings contribute to the thematic progression of the discourses in question, to the
change or maintenance of discursive topics and to the overall objectives of the meetings. A further objective will be to relate the use of cleft sentences with the specific characteristics of the participants involved in the meetings.

I shall be using data from the bi-national Project Discourse and Social Practice in Portuguese and Brazilian Enterprises, developed by teams of researchers from the University of Lisbon and the Catholic Universities of Rio de Janeiro and S.Paulo.

Rodriguez, Julia, Universite La Sorbonne- Paris V, France.
Thematic Progression and Field, Tenor and Mode in Two Folktales.

This paper tries to explore the thematic progression in relation to field (Ideational metafunction), tenor (Interpersonal metafunction) and mode (Textual metafunction) in two narrative texts : A Proper Gander by James Thurber, aimed at adults, and Mother Eagle and the Hunter, from a book for older children of folkstories.

These texts were chosen to discuss the hypothesis that they both belong to the same genre, if they both follow the same text structure then they should both follow more or less the same thematic organisation. And if that is not the case, to check if variables like Field, Tenor or Mode have any influence.

Although the sample for this study is too small to draw any final conclusions it shows that there is a connection between genre and thematic progression in a text, and between genre and field, tenor and mode. This may be interesting for teaching purposes.

For the student of English as a second language genre analysis can be more useful if it is seen in relation to these three variables (Field, tenor and Mode). Writing effective texts in a special genre having in mind the different purposes, audiences and medium can help students to choose from the vast panoply of options within the same genre.

Rowley-Jolivet, Elizabeth, Université d'Orléans, France.
Visual Textual Patterns in Scientific Conference Presentations.

In science, the visual semiotic is an important resource for making meanings, some of which are less efficiently conveyed by the linguistic semiotic alone (Lemke 1998), and scientific disciplines have developed particular visual languages and visual coding strategies to satisfy these communicative needs (Kress & van Leeuwen 1996). In the oral research genre of the conference presentation, the visual channel of communication is omnipresent, with slides or transparencies being projected throughout the speaker’s monologue; in the international conference situation, where the majority of speakers and audience participants are non-native English speakers, and where no written text can be referred to, the linguistic semiotic cannot be relied upon to the same extent as in the research article to communicate meaning. As a result, the visual channel carries a particularly heavy ideational, interpersonal and textual burden. It is therefore important, for a better understanding of this discourse genre, to take into account its visual dimension.

Scientific conference papers are « audio-visual texts » of a rather particular kind : the verbal track is formed by the speaker’s linear monologic commentary, and the visual track by a succession of ‘stills’ (the slides and transparencies). To fulfil the communicative purpose of the research presentation, not only does the correlation between the two tracks need to be close and continuous, - unlike the much looser correlation that can pertain in certain documentary films (van Leeuwen 1991) - but the logical relations within each track also need to follow a coherent development in order to construct the speaker’s claim. This paper explores some of the visual conjunctive relations between the stills in the visual track in order to show how the larger textual patterns which create coherence in text - General-Particular, Problem-Solution, Claim-Evidence, Matching relations ...(Hoey 1983; Winter 1994) - are realised visually. The data used for the study comprise video recordings of 90 conference presentations given at international scientific conferences in three fields (geology, medicine, and physics).

References
Hoey, M. On the Surface of Discourse. London : Allen & Unwin, 1983.
Kress, G. & T. van Leeuwen. Reading Images. London : Routledge, 1996.
Lemke, Jay. « Multiplying meaning : Visual and verbal semiotics in scientific text ». 87-113 in  J.R. Martin & Robert Veel (Eds). Reading science.London : Routledge, 1998.
van Leeuwen, T. « Conjunctive structure in documentary film and television ». Continuum 5 : 1, 76-114, 1991.
Winter, E. « Clause relations as information structure : two basic text structures in English ». 46-68 in  Malcolm Coulthard (Ed.) Advances in written text analysis. London : Routledge, 1994.
 

Saki, Mohamed, Université de Bretagne Occidentale, France.
Thematic Progression Patterns and Generic Identity.An analysis of the distribution of thematic progression patterns in abstracts.

The present paper sets to investigate whether there are any systematic correlations between the concept of thematic progression-which concerns the ways texts develop and present their ideas- and texts, exhibiting a quite strong generic identity, viz. abstracts. This issue is addressed by analysing how the overall schematic structure of abstracts, which enables academics to achieve specific rhetorical purposes, enters into a dialectic relationship with the thematic progression patterns. The analysis will be mainly concerned with finding out whether there are any systematic relationships between each of the large elements of the abstract macrostructure and a given progression pattern.
 

Schiess, Raimund, University of Frankfurt am Main, Germany.
Reading the Television Screen: Text, texture, screen design.

Many contemporary television newscasts do not show just 'pure', photographic images but combine them with superimposed written text and other graphic, electronically generated elements (logos, dividing lines, frames, etc.). Building on the work of Kress and van Leeuwen (1996, 1998), I treat this type of television screen as a complex sign that engages and interrelates different semiotic modalities; such screens require on the part of the viewers both traditional reading skills and 'visual literacy'. In addition, the visuals of tv newscasts interact with the sound, thus producing a highly complex multimodal text that integrates spoken and written language, visual images and graphics. The aim of the paper is to show how the tv screen layout coheres internally, i.e. within the visual mode, and how it relates to the tv sound track, i.e. spoken language.
The data are drawn from CNN's coverage of the 'Florida Recount'. CNN makes frequent use of inserts, i.e. short stretches of written text (narrative and non-narrative) that appear at the bottom of the screen. The station also employs montages of moving images showing events happening simultaneously at different locations; these separately framed images are embedded within a graphic background and combined with written text. CNN's coverage is a prime example of how meaning is made across modes and a good starting point to investigate the interrelationships between different semiotic resources.

Silvestre, Carminda, ESTG-IPL/ULICES, Portugal.
Discourse Markers as Texture Creating Devices in Business Meetings.

The aim of this paper is to identify the different functions of discourse markers in a corpus of spoken business language constituted by four business meetings of four different Portuguese companies. I shall be particularly interested in researching the way discourse markers contribute, as cohesive devices, to the texture and coherence of discourse, in spoken Portuguese.
Following research developed in socio-interactional linguistics I will also try to relate the use of the discourse markers identified to the different participants and leaders of the business meetings, as two of the meetings are lead by women and the other two by men.
The research undertaken is to be developed within a project entitled “Discourse and Social Practice in Portuguese and Brazilian companies”, involving the University of Lisbon, Portugal, and the Catholic Universities of São Paulo and Rio de Janeiro, Brazil.
 

Spence, Robert, Universität des Saarlandes, Germany.
The Textual Metafunction as a Didactic Problem in Applied Translation Studies.

When it comes to teaching and learning translation, the textual metafunction poses a number of conceptual, terminological and procedural challenges. Most of the conceptual challenges arise in the context of extending the notion of linguistic structure beyond its traditional domain of constituency and dependency, and can thus to some extent be reduced to issues of linguistic representation. The terminological challenges are in part a function of the typological differences between the (native and foreign, source and target) object languages involved in the translation tasks being didactically simulated, but are also in part a function of the typological differences between the (linguistic and translatological) metalanguages that have become dominant in the academic cultures natively associated with those object languages. The procedural challenges are not reducible to a simple function of the conceptual and terminological ones, but are related to more fundamental differences between the forms that didactic or, in Bernstein's provocatively broader terms, "pedagogic" discourse takes in those academic cultures; in the present case, we are dealing with the mismatches between a German institutional setting (whose preferred form is the "methodische Einführung"), French students (who have been taught to expect an "initiation raisonnée") and Anglo-Saxon teachers (who are inclined to want to offer a "practical introduction"). The transculturality inherent in (didactically simulating, and pedagogically transmitting) the translation process provides a context within which to criticize these institutional preferences, acquired expectations, and
unconscious inclinations.

On the basis of a corpus of French-English translations done under examination conditions by French-speaking students at the University of the Saarland, we present a statistical overview of the problems the students most frequently encounter in dealing with the textual-metafunction features of French texts that are to be put into English -- in particular in relation to Theme-Rheme and Given-New structures -- and list some of the conceptual and terminological difficulties their teachers have in recognizing and classifying these problems-- attested marginalia include "word order", "sentence structure", "stress", "emphasis", and (the thoroughly opaque:) "text". Anglo-Saxon didactic empiricism apparently requires French-speaking students of French-English translation to auto-contextualize these (and many other) scribbled comments, as part of the process of their learning to culturally mediate between the
continental and the cross-Channel. In the spirit of the Hallidayan attempt to reconcile Firth with Saussure-via-Hjelmslev, we offer a methodical, rational and practical alternative to the current classificational chaos.
 

Sriniwass, Sridevi, University of Malaya, Malaysia.
A Systemic-Semantic Investigation of Textuality through the Resources of Ideational Lexis.

This cross-disciplinary research is associated with a Systemic Functional Linguistic content analysis of academic texts in scientific discourse.  The investigation of the textual function of the semantic system is in terms of how lexis is organised in chemistry to display cohesion.  The cohesive nature of the text in lexical patterning is demonstrated by  capturing  the cohering lexical items syntagmatically and paradigmatically in system networks  (Fawcett, 1988; Halliday & Martin, 1981).    The networks in chemistry, in this study, present the ideational lexis as choices which systemically relate to other choices in a degree of delicacy.  This study seeks to show the textual patterning of ideational lexis and how lexis embodying concepts in chemistry may be represented systematically using a Hallidayan system-structure theory of language.  The texts selected for the  semantic interpretations of lexis are drawn from the genre of chemistry textbooks, used at the undergraduate level at the University of Malaya and depict  the kinds of academic scientific texts  frequently consulted to supplement lecture notes.  This study also seeks to explore  the reintegration and the adaptation of ideas on the framework of lexical cohesion provided by  Halliday (1994), Halliday & Hasan (1976), Hasan (1984) and Martin (1981, 1989 & 1992).  The traditional  taxonomic relations of hyponymy, co-hyponymy, repetition, synonymy, antonymy, meronymy and co-meronymy are applied  to identify the establishment of cohesive relations.  This  research, which was carried out in 1995 through 1996, is also related to the more recent interpretations of lexis in Tucker’s (1998) work.  In summary, this study shows that the explication of a text in terms of its lexical organisation demonstrates how  meaning relations are  made manifest  in the creation of texture in text.

Steiner, Erich, Universität des Saarlandes, Germany.
Translated Texts: Some properties derived from modelling understanding as grammatical de-metaphorisation.

A first type of source for properties of translated texts are systemic typological differences between languages. We shall give a quick overview of these for English and German, adopting a systemic i.e. paradigmatic, rather than structural i.e. syntagmatic perspective. A second type of source can be seen in (intended or unintended) register change in translation. A third type of source are properties of translated text deriving from the relationship and process of translation as such, rather than from properties of the languages involved: Translations, especially English-German, are often (but not necessarily correctly) claimed to be

-longer than their originals;
-less expressive in terms of affect;
-shifted in terms of text type from the persuasive and argumentative region of the cline of text-types to the more descriptive region;
-disambiguated and more explicit than originals;
-marked due to typologically-motivated influences of the source language;
-arbitrarily incomplete relative to their originals

In some strands of work, universal tendencies of normalization, simplification, explicitation and levelling-out are put forward as hypotheses to be empirically tested.

We shall present some attempts at quantifying the effects of the assumed process of translation as such, focussing on processes of “understanding”. It will be argued that “understanding” of source texts involves grammatical de-metaphorization, a process which will not be completely undone in producing the target text. This will lead on to hypotheses to be tested in some corpus-based studies of translated texts in a project involving electronic corpora in Saarbrücken.

Our remarks here will be formulated against the background of work by Halliday and Matthiessen and by Martin on Grammatical Metaphor, on work by Hawkins and by Doherty on typological relationships between English and German and their influence on translation, on previous work by Baker and by Toury on properties of translated text, as well as on some recent work by Hansen, by Steiner and by Teich on empirical investigations of translation and other types of multilingual text production.
 

Thomas, Alain, University of Guelph, Canada.
Textual Cohesion, Pauses and hesitations in the speech of advanced FSL students.

The proposed paper reports on a large-scale on-going study on the linguistic progress made by French-as-a-second-language students who spend their third year of university in France (the "experimental" group, 48 subjects), by comparison with their class mates who choose to stay and study French in a regular program offered by a Southern Ontario institution (the "control" group, 39 subjects). In both cases, progress is measured by means of a one-hour test of oral and general proficiency, administered both at the beginning (pre-test) and the end of the academic year (post-test).

This particular part of the study is based on the students'reading of two pages of text and on their recording of a semi-direct 12-min. oral test where they express themselves spontaneously in French on familiar subjects. Comparisons between pre-test and post-test results were used to determine individual progress and group performance. Speed of delivery,
pauses and hesitations were quantified to evaluate textual cohesion – the focus of this paper - and to highlight differences between the experimental and the control groups. Results point to textual cohesion as one of the main benefits of the year-abroad experience in the area of
speech production.

Thompson, Geoff, Univertsity of Liverpool.
The Role of Conjunction in Text as Exchange.

It is generally accepted that all utterances are intended (in the unmarked cases) to be seen as connected in some way with other utterances around (see e.g. Hoey and Winter, 1986); and any investigation of the texture of text needs to take account of this. There are two broad perspectives on how the connections are made: either through sequences of moves (e.g. question and answer) or through conjunction (e.g. cause and effect). Within SFG, the first of these is handled under the interpersonal metafunction, while the second is handled under the ideational (logical) metafunction; and the two are seen as having different reflexes in the textual metafunction. However, there has been steady, if less salient, focus on internal conjunction: that is, conjunction seen as construing aspects of the interaction between speaker and hearer (see e.g. Davies 1979, Verstraete 1998). In this paper, I will present contrasting text analyses in order to illustrate how different areas of conjunction may be seen in either logical or interpersonal terms. On the basis of these analyses, I argue that internal conjunction, rather than being an interesting but minor extension of external conjunction, is at least equally important, if not more fundamental. This argument is presented as part of an overdue re-orientation of linguistics away from its traditional bias towards content as the core, in favour of an overall privileging of the interpersonal perspective.
 

Davies, E. (1979) On the Semantics of Syntax: Mood and Condition in English London: Croom Helm.
Hoey, M. and E. Winter (1986) ‘Clause relations and the writer’s communicative task’ In B. Couture (ed.) Functional Approaches to Writing: Research perspectives London: Pinter, pp 120-141.
Verstraete, J-C. (1998) ‘A semiotic model for the description of levels in conjunction: external, internal-modal and internal-speech functional’ Functions of Language 5/2: 179-211.
 

Van Mol, Maarten & Mick O'Donnell, Language&Computing nv., Belgium.
Medical Discharge Notices: Relating generic structure to recurring linguistic patterns.

This paper will explore the generic structure of one genre of text, medical discharge notices. We first outline the generic staging of the genre, which is reasonably regular. These results are derived from the study of a large corpus. We then explore each stage of the text (e.g., case history, admission details, technical investigations, etc.) in terms of the regularly recurring features that are exhibited, including typographic, lexical, syntactic and semantic features. We then attempt to explain the recurring features in terms of the function that the generic stage is serving in the discourse.

This ongoing work is part of a larger project, with a goal of using computers to automatically recognise the structure of these texts.

Veltman, Rob, UK.   - WITHDRAWN
On Clarity / Sur la clarté dans le discours.

Taking my cue from Rivarol ( 'Ce qui n'est pas clair n'est pas francais' 1784:49 in Classiques Larousse), I want to examine the notion of 'clarity' as an important discourse and textural phenomenon, which has served to evaluate languages as well as texts. Thus texture may have phylogenetic/synoptic consequences for language systems, as well as logogenetic/dynamic ones for texts. The notion of 'clarity' has resonances with grammatical metaphor and naturalness of grammars.

Webber, Pauline, Italy.
From Spoken Science to Published Research Article:a Comparative Study

Most systemic linguistics consider that choices on the register plane are controlled by the genre plane. In fact, investigating genre can reveal a great deal about the different communicative purposes of texts and register variations in specialised discourse. In scientific English most genre analysis has concentrated on written texts, so that the knowledge of sets of choices available in spoken science is still limited.
The aim of this study is to analyse and compare an oral presentation of  research and a written article from the same medical sub-field and on the same topic. Nwogu’s (1997)model for the global structure of a medical research article was used to analyse several articles analysed by students in the medical faculty as part of their self-access study assignments. This model was accordingly modified and applied to the index article and to the text of the oral presentation.
There were marked similarities in the overall structure of the two genres, but the oral presentation contained more overt discourse signals and boundary markers, probably to facilitate real-time processing. On the register plane, there were notable differences in mode, as the role of the slides was a more important structuring device in the spoken material than were the visuals in the written article. The tenor was also different, as the co-presence of participants at the conference led to greater use of  politeness and other features which are part of the interactional ritual of the conference.