The Interpreters’ Newsletter, No. 4 COHESION AND THE SYSTEMIC-FUNCTIONAL APPROACH TO TEXT: APPLICATIONS TO POLITICAL SPEECHES AND SIGNIFICANCE FOR SIMULTANEOUS INTERPRETATION By Sandra Gallina Permanent Interpreter EC Commission - Brussels — ee — — Introduction syntactical levels of language were gained This paper presents the results of a study on cohesion in some political speeches pointing to the advantages of the systemic-functional approach. The initial sections dealing with the evolution of textual concepts are addressed to readers who are not familiar with toxt linguistics. The=detail into which this review goes stems as mbch irom the need to stimulate interdisciplinary exchang as from the very warm response ot [fellow interpreters at the receiving end of Search activity whenever their attention is drawn to the theoretical aporia of simultaneous interpretation though simultaneous interpreters are called upon to translate in an on-line process, strictly speaking texts and texts alone are units of translation. This statement disposes of the mot- a-mot fallacy and stresses the significance of cognitive studies casting light on all the difficulties of text processing in simultaneous interpretation 1. Preliminary Remarks 1.1 The Study of Cohesion in Linguistics ‘The constraints on linguistic studies of those aspects that promoted the impression of unity at the level of text began to relent during the third quarter of this century, at the same time as remarkable insights into the morphological and 62 through the extension of the categories developed for phonology. Relations "beyond the sentence" had—fong been viewed as the preserve of thetoricians, aestheticians and pragmatists where llinguists were expected not to poach Admittedly, the range of combinatorial possibilities above the sentence baffled all atiempts at structural study. Yet the existence of fanguage features (e.g. co-teference, erepeminalization, choice of articles, word order, opic comment and intonation) that could not be explained within sentence boundaries prompted studies whose divisive issue was the appropriateness of sentence grammar to deal with these very aspects. In 1975, when prefacing M.A.K Halliday and R. Hasan’s book Cohesion in English , Randolph Quirk praised the authors for persisting in subserving “literary and other humanistic disciplines by extending their work to embrace stylistics and other aspects of textual studies" ( M.A.K Halliday and R. Hasan 1976, p. Vi) during a decade and a halt of intentional neglect. By 1985, though, the Swedish scholar N.E. Enkvist hailed the rapid growth in text and discourse studies as a dramatic breakthrough, adding that students interested in style and texts could "help themsetves to new succulent and nourishing dishes from the linguistic smérgdsbord". Sincethe publication of Halliday and Hasan's systematic study on cohesion, the question of what holds a text together has featured prominently amongst the plethora of studies subsumed under the labels of “text linguistics" or “discourse analysis". The bewildering array of terms coined by scholars formulating theories about the quiddity of texts betrays fundamental differences in concept. The devices appearing in the corpus studied and presented under the heading of “cohesion” might be seen as either quidditas or qualitas of texts, hence the classification of the concept of text and the unambiguous definition of cohesion prefacing ‘the paper in question. 1.2 Presentation of the Study The first part of the study set cohesion against its diachronic background of successive enquiries into the nature of the object ‘text’ and finally placed it within the system of language. The second part offered a detailed analysis. of the cohesive resources contained in the corpus which consisted of speeches delivered by U.K. and Italian Members of the European Parliament between 9th Sopt. 1985 and 1st Jan, 1988. in the light of the increasing relevance of text studies centred Upon cohesion for both the field of artificial intettigence (ct. Smith and Frawley 1983) and stylistios (cf Enkvist 1985), the conclusion of the paper strove to explain the significance of cohesive tesourebs in texts belonging to the genre of;|political language, a variety in which no such work had been published. 2. The Need for a Concept of Text 2.1 From Sentence to Text Since the time of the methodological turning point when the text was recognized not only as an object deserving the attention of linguists but as primum datum almost two decades have elapsed and still neither a definition of text nor a specification of commonly accepted methods is forthcoming. The failure to achieve clear definitions is due to the disparate. tasks set for text linguistics and the gamut of morphic variations of the object text. Apart from non- verbal communicative acts, anything from a nursery rhyme to a signpost is described as a 63 The Interpreters’ Newsletter, No. 4 ‘text’ and a science of texts is expected to account for both similarities and dissimilarities (v. de Beaugrande and Dressler 1981). The terms ‘text’ and ‘linguistics’ have been increasingly juxtaposed to indicate studies extending methods of linguistic analysis to elements transcending sentence boundaries, but most scholars have emphasized the striking heterogeneity of the domain any such label is thought to refer to “der Gegenstand der Textlinguistik ist bislang noch nicht genau identifiziert worden, so daB ‘Textlinguistik’ in gewisser Hinsicht nicht viel mehr als ein Name far sehr unterschiedliche Betrachtungsweisen ist, ja sogar fur ganz verschiedene wissenschaftliche Disziplinen. "(Coseriu 1981, p.5). mThaychacs arises from a syncretic fusion of two definitions of text: a) the realization of a natural language, b) the superordinate of categories such as ‘novel’, ‘poem’, etc., independent of historical languages. Despite this confused state of affairs Conte (1977) distinguished three approaches in text linguistics: the first concerning studies analysing regularities beyond=sentences, the second concerning studies fecussing upon the construction of text grammars and the third concerning studies aiming at the erealion of comprehensive text theories. fnitially, text linguists regarded sentences as ihe prius since texts were mere sequences of sentences. As a result the distinction between texts and arbitrary concatenations of sentences was irremediably blurred. In the study of sentence constituents with relations across sentence boundaries ignoring textness is a treacherous procedure for "@ solo tematizzando fa struttura gerarchica di un testo, Ja sua coerenza semantica globale che si pud fare il passo dall'enunciato al testo." (Conte 1977, p.17) The second approach ted to the pursuit of comprehensive text grammars where the text was increasingly viewed as the actual realization of a language more akin to a speech act than to a sentence. The raison d’étre of text grammars was the failure to explain some linguistic phenomena inThe Interpreters’ Newsletter, No. 4 texts through sentence grammars, because “cid che legittima una grammatica testuale 8 una discontinuita fra enunciato e testo, la differenza qualitativa non meramente quantitativa fra enunciato e testo." (Conte 1977, p. 17) Sentences are elements of the virtual synchronic system: while the definition of a sentence relies on syntactic criteria, the definition of a text calls for different criteria as a text can infringe the rules of a natural language. Even the competence involved in understanding the correctness of a sentence differs from that involved in understanding whether or not something is a text and what type of text it is. By virlue of this specific text competence speakers can paraphrase, reduce or summarize aytext-once. they have grasped its gist. In the early stages research was evolving merely through changes in methodology’ with a passive adoption of a pre-theoretical notion of, ‘text’: rather than an act coming to fruition hie et nune , verbal products fixed in writing. Nor was the marked preference for literary texts a mere coincidence. The advent of studies. on conversation eroded some fundamentalypillars of text linguistics (e.g. the issue of pauses defimtting texts: what happened when several \sydakes spoke at the same time?) as syntactic eohesion and/or semantic coherence were useless tackling anacolutha. Furthermore, semantic coherence was not an intrinsic quality as toxts were the means whereby the speaker's communicative intention was conveyed for the receiver to recognize and accept it, acceptability depending upon a continuity between textual occurrences, context of situation and the knowledge of the world shared by the two speakers, in 2.2 The Systemic-Functional Model 2.2.1 Text and Context Arising from a long-standing tradition in ethnolinguistics, the systemic-functional model is ‘consistent in positing for extra-linguistic aspects a role that is no longer that of an ad hoc tool to be resorted to whenever texts display difficulties in interpretation. A ‘text’ is described in terms of ‘action’ rather than in terms of the language system, thus avoiding various difficulties as the text lies on the outermost layer of linguistic events and has no superordinate linguistic unit circumscribing its limits. The ‘constituents’ of a text are analysed in functional terms and regulated by contextual non-verbal tactors because of the relationship between a text, its context of situation and its context of culture. A text is not a supersentence belonging to an unbroken constituency chain from morpheme to text, but rather “a unit of language in use. It is not a grammatical unit like a clause or a sentence, and it is not defined by its size." (Halliday and Hasan 1976, p.1) Halliday’s concept of context of situation, to be traced back to Firth and thence to Malinowski, whose observations propounding a cultural and situational semantics were a stimulus to considering languages a potential form of meenig, is of an open set of options that members of a society had. The series of options Wasyembedded in the context of culture whilst specific choices occurred in the context of situation: a distinction that translated in linguistics the diorism of ‘power’ and ‘act’. Firth’s ‘geiitext of situation’, though, abstraction as it was not in rebus, Audiovisual record of surrounding props, but was an abstract representation of general categories relevant to the text [1] . In Halliday's socio- semiotic theory of language the situation is “the environment in which the text comes to life" conceived in even more abstract terms as a ‘situation type’. It is essentially a semiotic structure or a constellation of meanings deriving from the semiotic system that constitutes culture. Through this information the receiver of a text can make some predictions about text meaning, for texts are meaningful not because the receiver is unaware of what the producer is going to communicate (as would be the case with a mathematical model of communication) but paradoxically because he is aware. The process of selection of meaning options is regulated by Hymes' ‘native theory and system of speaking’, a communicative competence enabling members of a 64community to know when to speak, keep quiet, etc. Hymes' list of components (quoted by Halliday and Hasan 1976 and by Brown and Yule 1983) can be paraphrased as follows: form and content, setting, participants, ends (intended and achieved), key, medium, genre and interactional norms. Apart from the difficulties in establishing which were the relevant features for a description given the lack of a ‘system’ of situations, the text itself seems to be ignored as a determining factor. Halliday grouped situational factors determining texts into more abstract categories, i.e. field, tenor and mode, thus creating the concept of ‘co-text' to refer to the text. A context is not a lumpen mass: texts create their contexts. The situation is determined by: (i) field - the social action, what is going on, possessing a recognizable meaning in a society, typ complex of acts in some ordered cont where a text plays a part; it includes maiter as a specific manifestation; (i) the role structure, a cluster of meaningful participant relationships i permanent attributes and situationat speed ‘@ particular instance of it are levels of f (ii) mode - the symbolic organization, the! selected and the status of a text in relBligh) 49 field and tenor; it includes the medium (5 Edizio written). Collectively they serve to pre The Interpreters’ Newsletter, No. 4 whole they represent one function of language and each element has a role in a function. Utterances in aduit language are functionally complex and there is not a one-to-one correspondence of function and structure. indeed language functions are principles organizing the semantic level of language and taking grammatical forms. The lexicogrammatical system has a functional input and a structural output, hence the combination of functions through grammar. All languages are equally sophisticated and regardless of the complexity of social structures they are based on three elementary functions. Unlike philosophical or sociological theories of language, Halliday's systemic-functional model ‘expounds a taxonomy not a hierarchy of functions termed as ‘ideational’, ‘interpersonal’ and ‘textual’. Widely regarded as ‘the’ function of the Ideational function serves the of content encoding experience [2]. personal relations (including forms of ticipation in the speech situation) are by the interpersonal function. Finally, stablishes {inks with itself, hence the a textual function enabling a iter to produce a text [3]. The stratum shows four components fqiecriontial, logical [4], interpersonal and Textual), i.e sets of interrelated choices. The FBENSikogrammatical stratum maps the structures ‘register’ whereby the observer deri # pf@ggiiied by each component as output of the ed, el norms governing the particulars of a te: various subcategories of field, tenor and mode are linked to typical semantic properties of a text and not lexicogrammatical variants. 2.2.2 Language Functions Lexicogrammatical Representations Speakers do not choose in vacuo as language potential is organized in networks of options each corresponding to one language function described in linguistic terms without resorting to psychological categories. Throughout the 1960s investigations looked for function as a correlative of structure; functions though are anything but secondary since in principle functional relevance permits the valid recognition of structure. The grammatical structure, on the other hand, is not an arbitrary configuration of elements since as a and 65 is in a network onto each other and produces @ text. Each component of meaning makes a contribution to the structural output {5). Meanings in all the three functions are generated simultaneously and mapped onto each other, for instance speakers do not decide on content first andior whether to encode it as a question. These are neither subfunctions of a ‘communicative’ function nor macrofunctions; they are metafunctions for they are abstract. Through this triadic system for the semantic level and the tripartite categorization of the situational determinants, the situation, the text and the semantic system can be linked. Situational determinants activate semantic components: field activates the ideational component, tenor the interpersonal and mode the textual. The two sets of categories are established independently andThe Interpreters' Newsletter, No. 4 actualized through the ‘code’ which is not a variety of language but the whole of those principles governing the choice of meanings by a speaker and their interpretation by a hearer. Codes are actualized through ‘register’ determining the semantic orientation of the speakers in particular social contexts. 2.2.9 Text: Structure and Texture Texts are the product of the meanings of all the components and the textual component makes all the difference between language ‘in the abstract’ and ‘in use! and between a random string of sentences and sentences realizing a text. The systemic-functional model has dealt with the issue of text studies from two (as would appear unrelated) directions that on closer scrutiny reveal their nature as fundamental and partly complementary traits of textness: structure’ and texture. The property of structure was Halliday and Hasan's (1976) fundamental ritetion | to distinguish both between complete and ineémplete: texts and between genre forms, Hasan (1978) posited the existence of ‘structural formulae’ against which the actual structure of a text was to be measured. A structural formule WSS a configuration of elements in a text stricture’ realized by lexicogrammatical units but) ¢reated on the basis of functions that depended omtite téxt genre, whence we return to the cottextual controls on structure. The other structural factor of the textual component is the thematic and informational distribution which is crucial in the distinction between texts and non-texts (6). Textness is the result of cohesion which is independent from linguistic structure and may extend beyond any structural unit, Cohesion is determined by mode and is relational rather than structural, in fact, it is alternative to structure. The features described in the last few paragraphs are not stylistic variants. Unfortunately most metaphors enshrine the view of @ system that generates a representation of reality, encodes it as a speech act and re-codes it as a text, Yet meanings are embodied in simultaneous networks of options and their structures are mapped onto each other so that realizates of these meanings appear prosodically 66 ‘throughout a text. When ‘listening’ to a text we ‘listen’ to a multilayered polyphonic structural composition. In polyphonic music the different melodies are mapped onto one another so that any specific chord is simultaneously an element in different melodies. This very ‘polyphonic’ quality motivates specific cohesive devices. 3. Cohesion in the Linguistic System 3.1 Cohesion, Connexity or Coherence? ‘Cohesion’, ‘connexity' and/or ‘coherence’ are terms adopted to indicate text specific properties. “The terms ‘connexity' and ‘cohesion’ refer to the (verbal) construction of a {ext, while the term ‘coherence’ refers fo the relation between the states of @ilaits expressed in the text. (Hatakeyama ef ali? 1985, p.58 underlining theirs) These definitions are the outcome of a different viewpoint from Halliday's. Cohesion is neithe:_necessary nor sufficient for a text to be fegardedias such since it is better conceived as a quafitas , Texts must ‘cohere’, i.e. refer to non- Tinguistié events in a way that for hearers is @fderly and structured. No advantage is to be gained by establishing constitutive formal featufes that @ text is to possess to qualify as a ‘text for texts are what text receivers decide to treat as texts (cf. Brown and Yule 1983, p.199 et seq. and for an interesting view on linguistic signs in general s. Ungerer 1991). Gonnectedness cannot be equated with grammaticainess [7] Unlike cohesion, coherence encompasses a relation between the text and the communicative setting (cf. Randquist 1985). Incoherence is the result of lack in prototypical knowledge to cover @ specific situation, not of a failure of linguistic form. When two elements are placed together any hearerireader will fee! compelled to co-interpret them as a whole and as relevant to the context in his/her ‘effort after meaning’. Brown and Yule (1983) reiterated Popper's cognitive assumption that human beings are primed for regularity when stating that meaning in a situation type is valued against the background of regularities with meaning predicted through context. The moreritualized the genre the greater the number of expectations fulfilled [8] Connestedness is an elusive concept for it is acquired by (or bestowed upon) a text through contact with a person. The author creates it but receivers either maintain or modify it so that texts as polyphonic creations are re-shaped at each ‘performance’. There is no gainsaying the creative contribution of text receivers; still they are guided by the cohesive devices deployed by authors. "To what extent isthe producer of a text responsible for its coherence, and how much inversely is construed (or induced) by the receiver?’ (Charolles 1985, p.86) Cohesive marks have no value unless perceived, thus the distinction between cohesion and coherence becomes a relative affal “Let us say, for example, that accept an anaphora as being the mark cohesion and that a particular series two sentences is cohesive because # first one is taken up by a co-refere pronoun in the second. Is the sequence two sentences cohesive in its own rig) or Is it coherent by virtue of a certa operation which we presume the hearBé¢|i7 i to carry out? It is an embarassin question." Indeed It is. Despite a certain ambiguity in’ Pip distinction between cohesion and cobierence, elements such as anaphoric relations must be regarded as part of the data. Relativity of interpretation cannot enter at this stage lest everything should become relative and scholars be left with nothing to study. Cohesive resources are language specific still they can be related to language in foto and expounded in the direction of language typology. A. study of cohesion cannot be based on merely formal expressions such as “the concept of cohesion is a semantic one; it refers to relations of meaning that exist within the text and that define it as a text” (Halliday and Hasan 1976, p.6} Since cohesion is a relational concept it is not through its mere presence in a text that an item Univers: The Interpreters’ Newsletter, No. 4 is cohesive but through its relation with another item, The fundamental principle is that of a ‘tie’, sc. ‘one occurrence of a pair of cohesively related items', whose salient feature is directionality. The relationship between two elements may be either anaphoric or cataphoric whenever the presupposed item either precedes or follows the presupposing one [9].The two elements may be found in immediately adjoining sentences (immediate tie) or in sentences that are far removed (remote tie) and in this latter instance even through some other intermediate cohesive item { mediated tie) thus leading to ties that are both remote and mediated. Relational patterns confer upon texts their rhythmical periodicity through both the type and the number of ties in sentences. The issue ipinges upon the cognitive principle of selective a variety of entities passes quickly periphery of the perceptual field and nds to remain attached to one entity Y sets in and the focus changes. dudy analysed cohesive resources PE thom into four types: reference, ellipsis or ellipsis proper), conjunctions and olc! cohesion. Reference designates the property that some #EBisplay in making reference to ‘something este”, their interpretation. It fosters continuity ‘FF HiBaning either anaphorically or cataphorically and can be classified as ‘personal’, ‘demonstrative’, ‘comparative’. Despite the terminology the classification is based on semantic and not grammatical criteria, Personal reference is brought about by pronouns that just like lexical items can be Heads or Modifiers within a nominal group even if their inclusion in limited sets is reminiscent of morphemes. The pronominal system transcribes the various roles in the speech situation thus not all items are ‘endophorically’ cohesive, i.e. referring to something in the text. It is generally the third porson that is cohesive through anaphora [10]. ‘Demonstrative’ reference is a form of verbal pointing whereby the referent is identified in terms of proximity to the speaker, though Halliday and Hasan (1976) extend this label to 67The Interpreters’ Newsletter, No. 4 include neutral instances of cohesive definite articles. Personal and demonstrative reference are resorted to for ‘extended’ reference, i.e. reference to a process or sequence of processes, or ‘text’ reference, i.e. reference to a stretch of text. Finally, reference may also be the result of, comparisen, ‘comparative’ reference, where cohesive items presuppose an item acting as a standard of reference. While reference is a relation in meaning (at the semantic level), substitution and ellipsis are mainly relations in wording (at the lexicogrammatical level). The difference between substitution and ellipsis is that in the former the antecedent (11) must replace the substitute and in the latter the antecedent must be supplied. A major element of cohesion is conjunction. Unlike other links it is ‘non-phoric' as conjunction ties do not presuppose antecedents. The tie concerns the way a sentence can be connected to the rest of the text, Conjunctive adjunets. (simple/compound adverbs and prepositional phrases) extend to the whole of a sentence unless Tepudiated. In the study conjunction was. classified as follows: additive, adversative, casual and temporal with several subtypes. The tables displaying cohesive ties for, eac! speech showed that lexical cohesion was lenuciel for this genre of the corpus. The study) folldwed. the systemic-functional concept of loxis and Coseriu's principles of syntagmatic siricturé (affinity’, ‘selection’ and ‘implication’). The thread of discourse is generally maintained through reiterations that just like general nouns (ie. items with generalized reference: ‘people’, ‘thing’, ‘person’, etc.) must be accompanied by reference items to be cohesive. Reiteration can be the result of the use of the same item, a synonym ‘or a superordinate with a cline emerging in this study between general nouns, superordinates, synonyms and repetitions [12]. Still the most complex aspect of lexical cohesion is the association of items that regularly co-occur. Collocations were classified as cohesive on the basis of Hatakeyama et alii 's (1985) concept of, ‘reference conform thesaural’ relations. There are no specific lexical items that are cohesive, cohesion being established only with reference to their occurrence in a text. Lexical 68 cohesion may be viewed as the mere result of a text developing along certain lines, yet this would be a facile assumption for some texts, though sticking to the point, display an extremely limited number of cohesive ties. The crucial consideration is the ‘textual history’ of a term in the lexical environment of a text with all the forms thereto related, without giving priority to relatedness in the language system. Proximity in the lexical system would be determined by the probability of two words co-ocourring whilst in a text it would be the material distance in terms of sentences separating the items. The cohesive force of a tie is proportional to proximity in these two senses. Items with high overall frequency and entering into collocation with several items (e.g. ‘know’, ‘dire’, etc.) produce ties of very low cohesive force if_at all. 4, Results The first consideration following the analysis of the speeches in the corpus was their peculiar referential cohesion. Apart from the total absence of instances of possessives as Heads, there was a verysiow number of ‘mixed’ personals, i.e. the ‘nous-vous’ relationship of inclusion/exclusion of text receivers. Ellipsis was expected to play a greater role but in fact instances of this cohesive 4id jaée negligible. Aleit limited in number, conjunction is Sighiticant in that it gives clues to higher level conceptual structures. While additives followed by causal ties prevailed in Italian speeches, causal conjunction followed by adversatives was the main feature of English speeches. Given that adversatives of the more simple type generally contribute to foregrounding, English speeches display a higher degree of vividness. The limited use of temporal conjunction ties is the corollary of preference for lexical cohesion. In these texts lexical cohesion was mainly the outcome of ‘nominalisations', a primary source of ambiguity, opacity and lack of conciseness. Lexical cohesion is one of the reasons why texts cannot be studied through an enlarged sentence grammar. Each sentence can be studied syntagmatically, on the basis of its combination of elements, or patadigmatically, on the basis of the choice relations between the actual elementsand other alternative elements in the system (paradigmatic substitution). Unlike sentences, texts display ‘syntagmatic substitution’ prompted by the rhetorical canon of non-repetition but supported also by the psycholinguistic phenomenon of ‘semantic satiation’. The different density of cohesive resources within/between texts is the outcome of two different phenomena: progression and focalization. Focalization is a concept reminiscent of Longacre’s peak (1981), i.e. rhetorical underlining to ensure that a ‘topic does not go by too fast’. For instance, this aim can be achieved through the insertion of short sentences in an environment of long ones. With reference to cohesion the effect is most conspicucus when remote ties are included in an environment of immediate ones, or the number of ties goes up unexpectedly. Progression concerns str text where the receiver is led onwards segment to another. The absence of devices is matched by a paucity of col a corollary to the introduction of elements or predicates. Non-progressio is not to be equated with focatization term referring to elements holding) receiver's attention leading to " satiation’. Most of the speeches in the days) TAPS displayed a surfeit of non-progression. - my pes Univer 5. Text as a Product vs. Tex Free: Process The analysis contained in the study is a posteriori focussing on the text as a product. Still a text may be studied in terms of process as interpretation studies generally do. As a socio-semiotic process any text can be described as a ‘semiotic encounter’, a channel to exchange meanings making up the social system. Text producers are 'meaners’ whose acts of meaning do maintain or re-shape social reality which Is constituted of indeterminate and unbounded meanings. The dynamic aspect of reality qua reality in fier shows its incessant re- structuring, periodicity and occurrence in time and space. Texls as processes incessantly shift their relation to the environment. Scholars idealizing this aspect away concentrate either on the system or on the single text. The temporal 69 sty? ing or subsequent information. The Interpreters’ Newsletter, No. 4 dimension so crucial for interpretation studies is lost sight of. As stated in the introduction to this paper, the difficulty of simultaneous interpretation is due to interpreters translating texts as on-line processes. Even though texts are units of translation, interpreters have to make do with smatler units, Hypostasizing these units is one of the areas covered by interpretation studies. Successful inroads seem to have been made through Searle's (1969) concept of speech act requiring the co-presence in the minds of the producer and the receiver of all those constituents [13] that even though appearing in sequence can be traced back to a unitary mental moment. Scholars seem to be convinced that a speech act does not exceed the boundaries ‘of what psychologists describe as short-term memory, Avoiding any term that may entail a whole series ms {is there a quantitative or a difference between short and long ory, if at all? Is there a break of between the two or do they shade into 17), speakers do have the ability to ations existing in a ‘unit’ that is being |. For all the difficulty concerning unit is, texts are unlikely to be processed in ep. It is precisely their articulation in .ch acts that prompts the use of anaphoric and ‘ataghoric cohesive ties, for it is these very ties ‘@ontain condensed and thorough reference to fr cohesion is a fundamental element simultaneous interpretation and in text processing in general. Unless the condensed information of cohesive items were available, textual solidarity would be irretrievable. Within sentences cohesive items epitomize the whole of the text process. ‘Adopting Halliday and Hasan's (1976) metaphor, a sentence whose presupposed items are unknown is like a picture that is blurred though complete. Such a sentence clicks into focus no sooner than its cohesive valencies are saturated. Inevitably, simultaneous interpretation is affected to an extreme degree by the type and the density of cohesion displayed. Through the 'modéle o'efforts: (cf. Gile, 1988) an interesting study might show if there is a link between the difficulties ‘experienced by interpreters and differences in cohesion. Time is an essential factor inThe Interpreters' Newsletter, No. 4 interpreting. Therefore, in any research work on interpretation, cohesion ought to feature prominently for cohesion is one of the most manifest symptoms of a text developing in time. FOOTNOTES [1] The context of the present text is totally different from what surrounded the act of writing it. [2] In performing this function language provides a structure for experience, hence the need for a conscious effort to see things differently from what our language might suggest to us, [8] The terms ‘system’ or ‘structure’ must not be used as synonymous with ‘function’. The linguistic system is organized around a set of abstract functions and the term structure refers ‘Semantic Component 1 experiential ideational \ logical interpersonal textual {6] Thematic and informational considerations are central to an understanding of psoudodlett and cleft constructions as Collins (1991) showed, [7] The issue is more pointed than it seems for in real lite coherence-grammaticalness is not an absolute value - either there is or there is not - but rather a relative quality. Incoherent texts might be ruled out as ungrammatical but to a certain degree. The text receiver might think that it is unacceptable ‘for the time being’, till an explanation is given for the oddity. Beyond sentence boundaries there are indeed no formal constraints establishing which sequences are acceptable or not. [8] Political language is a species of the genus of formal language: a restricted code because of the limitations on options at all levels. Independence of elements compatible in praesentia guarantees adaptability of language to reality. Formal language with its limitation of choices 70 to the representation of syntagmatic relations in language. [4] Unlike other components the logical component is characterized by its expression through recursive structures such as parataxis and hypotaxis (including apposition, coordination and reported speech). Still the distinction between experiential and logical within the ideational component is not merely due to different modes of expression. The logical element, while ideational in its origin, derives from the speaker's experience of the external world and once it is built into language it becomes neutral with respect to other functions: all structures, regardless of their functional origin can have logical structures built into them . [5] Here is a simplified table: Grammatical Structure constituent, segmental recursive prosodic culminative result in a blurred picture of the individuality ndGhistoricity of events transformed into ‘scriptural examples’. [9] Anaphora and cataphora acquire significance as an opposition within the system. The ordo naturalis would be for the presupposing item to follow the presupposed one, hence the markedness of the option of cataphora. An illuminating insight into the issues involved in studying anaphora is given by Berretta (1990), Conte (1990 and 1990a) and Kleiber (1990), while the significance of cataphora within sentences is analysed by Kesik (1989). [10] Exceptionally even ‘I' can be endophorically cohesive, of. Halliday and Hasan's (1976) example: ‘There was a brief note from Susan. She just said "I'm not coming home this weekend.’ In the example ‘I! refers to Susan. [11] Antecedent is a general term covering both anaphorically and cataphorically cohesivepresupposed items. [12] The moot issue concerning lexical cohesion is identity of reference. Halliday and Hasan (1976) posited that cohesion existed between two lexical items imespective of their teferential relation, thus it was that the occurrence might have identical, inclusive, exclusive or unrelated reference. [13] Needless to say that the present author would reject any such notion of constituents as a lower rung in the ladder leading to text. References BEAUGRANDE, Robert-Alain de, and DRESSLER, Wolfgang U., Introduction to Text Linguistics , London, Longman, 1981. BERRETTA, Monica, "Catene anaforiche in prospettiva fuzionale: antecedenti difficili*, Rivista di linguistica, Vol. 2 (1), Torino, Rosemberg & Sellier, 1990, pp.91-120. BROWN, G. and YULE, G., Discourse New York, C.U.P., 1983. CHAROLLES, Michel, “Text connex coherence and text interpretation proce: Sdzer (ed.), 1985, pp. 1-15. COLLINS, Peter Craig, "Pseudocteft constructions: a thematic and infor interpretation", Linguistics , Vol. 29 (3) New York, Mouton de Gruyter, 1991, 519. AB 1s Edizioni erence: Aspects, Methods and Results, in English, English Language Series, London, Longman, 1976. HASAN, Rugalya, “Text in the systemic- functional model", in W.U. Dressler (ed.), Current Trends in Text Linguistics - Research in Text Theory , Berlin-Now York, Walter de Gruyter, 1978, 228-246. HATAKEYAMA, RATSHUKIKO, JANOS, PETOFI, SOZER, “Text, connexity, cohesion, coherence", in Sézer (ed.), 1985, pp. 86-105. KESIK, Marek, "La cataphore", Linguistique Nouvelle, Paris, P.U.F., 1989, 77-90. KLEIBER, Georges, “Sur l'anaphore associative: article défini et adjectif démonstratil", Rivista di linguistica, Vol. 2 (1), Torino, Rosemberg & Sellier, 1990, pp. 155- 170. LONGACRE, Robert E., " A spectrum and profile approach to discourse analysis", Text, 1 pp.337-359 UIST, Madeleine G., "The barely visible aspects of textual connectedness", in ), 1985, pp.189-218. Speech Acts -An Essay in the of Language , Cambridge, C.U.P., 1969, iT Raoul N., and FRAWLEY, William J., Hunetive cohesion in four English genres", ), 1983, pp.347-374, ~SOZER, Emel (ed.), Text Connexity, Text CONTE, Maria-Elisabeth (ed.), La tihd-weligre pte wg, Helmut Buske, 1985. testuale , Milano, Feltrinelli, 1977. CONTE, M.-E., "Foreword", Aidila 1H OGLE linguistica, Vol. 2 (1), Torino, Rosemberg & Sellier, 1990, pp.3-7. CONTE, M.-E., “Pronominale Anaphern im Text", Rivista di linguistica, Vol. 2 (1), Torino, Rosemberg & Sellier, 1990a, pp. 141-154. COSERIU, Eugenio, Textlinguistik, Eine Einféhrung , Tibinger Beitrage zur Linguistik, Tubingen, Gunter Narr Verlag, 1981. ENKVIST, Nils Erik, "Stylistics, text linguistics and composition”, Text, 5 (4), 1985, pp. HIV. GILE, Daniel, "Le partage de fattention et le ‘modale d'effort’ en intérpretation simultanée", The Interpreters’ Newsletter, 1, Trieste SSLM, University of Trieste, 1988, pp.4-22. HALLIDAY, MAK, and HASAN, R., Cohesion 7 UNGERER, Friedrich, "What makes a linguistic successful? Towards a pragmatic interpretation of the linguistic sign", Lingua, Vol. 83 (March), North Holland Amsterdam, Elsevier Science Publishers B.V., 1991, pp. 155-181.