1974. Stuttgart - InterDisciplinary Aspects of Discourse Analysis.

By C. George SANDULESCU, Stockholm.

(Paper given at the Congress of Applied Linguistics which took place in Stuttgart in the summer of 1974.  Proceedings edited by Gerhard Nickel and Albert Raasch)

(EPIGRAPH:)(Roman Jakobson, 1973:)    The relationship between linguistics and the adjacent sciences awaits an intensive examination.

1.    This is an explanatory paper. It tries to put across the simple point that the main source of crisis in present-day linguistics is (a) the boundary problem, and (b) the lack of correlation studies. As such it is future-oriented, rather than the opposite.

2.    "The problem of interrelation between the sciences of man appears to be centred upon linguistics. [...]  Linguistics is recognized both by anthropologists and psychologists as the most progressive and precise among the sciences of man and hence as a methodological model for the remainder of those disciplines."

        Contemplating the history of the science of language is a useful and instructive way of peering into the future of this central discipline of the humanities. There are two distinct points worth taking up here: one is the fact that in its history there have been different language units lying in the central area of research preoccupations, and forming what may be termed a research focus. With the passing of time, there have been clear shifts in research focus; that such shifts have occurred is, of course, a commonplace, but the fact that they consistently occurred in a certain specified direction is far more relevant and symptomatic.

        Grimm's Law and Verner's Law as well as the work of Rasmus Rask illustrate one kind of preoccupations, emerging from phonetics and moving towards morphology. The total inability of diachronic linguistics to adequately cope with syntax is notorious.

        Research focus then imperceptively moved from the analysis of phones to the analysis of morphemes, and then to groups of morphemes. It was at about that stage that structuralism began to be remotely looming in the distance. As Taber (1966), Engler (1967), and Mauro (1972) interpret Saussure (1916), his emphasis was on the phrase; and research focus placed on the phrase roughly characterized the whole movement of surface structuralism, which could not provide and adequate description of the Sentence.

        Then came the dramatic hint at the "multistructure" approach, and the possibility of hierarchically organized layers of structure on a dimension of depth. It was thus that in 1957 Sentence Linguistics literally came into being.

        But after the first wave of enthusiasm was over it was gradually found out that there were many things that sentence linguistics could not do: first, it could in no way cope with inter-sentential relations; secondly, accumulated research tended to show that a 'ceiling effect' had occurred in syntax, which, in the words of John Sinclair (1974), now had to manage the intricacies of intonation selection, information organization, semantic structuring, sociolinguistic sensitivity, illocution and presupposition  (including the pragmatic variety), in addition to its whole array of traditional concerns. All these symptoms clearly point to the necessity of an additional level of linguistic description called discourse:  less than 20 years after the clear emergence of a multi-structure approach, as embodied by TG grammars, there is need of a multi-sentence angle in the discipline.

        3.    This consistent and unidirectional (from low to high) shift of research focus has led to the emergence in the last few years of both Text Linguistics and Discourse Analysis. The ultimate urge lying behind these new and similar (if not identical) research tendencies is the acute need to go beyond the somewhat artificial boundaries of the Sentence, hence one boundary problem. There has been considerable resistance to such 'expansionist' tendencies. The frontiers of linguistics were axiomatically laid down in statements like the one made by Katz & Fodor (1963), who declared that 'grammars seek to describe the structure of a sentence in isolation from its possible settings in linguistic discourse (written or verbal) [...] without reference to information about settings and without significant variation from speaker to speaker'. Situational contextualization and sociolinguistic sensitivity are thus emphatically left out. Zellig Harris (1951), too, discussing his brand of discourse analysis, has stated that "exact linguistic analysis does no tgo beynd the limits of the Sentence".  John Lyons (1969 : 176) echoed similar views by stating that "the Sentence is the largtest unit the linguist recognizes". Such clearcut statements, strongly programmatic in character, exclude all possibility of going beyond the Sentence.

4.    Such an attitude emphasizes formal aspects: any attempt at going beyond the Sentence must be justified by 'morphemic' evidence. But given the fact clearly acknowledged by many that there is an uneasy relationship between the formal and the pragmatic in linguistic description, one could roughly summarize the formal angle of vision in linguistics as a strong view animated by the desire to preserve pure the autonomy of the discipline in the sense given by Antoine Meillet as early as 1928, and strongly emphasized by Chomsky (1957, and subsequently), and Katz & Fodor (1963). By the side of the autonomous approach signalling the strong view, there has for some time been manifest an underlying trend towards integration. Hence, the second boundary problem.

5. Linguists, whether they like it or not, "must become increasingly concerned with the many anthropological, sociological, and psychological problems which invade the field of language," said Edward Sapir in 1929. And it is in fact the balance of power between autonomy and integration that characterizes the development of present-day linguistics. The substantial concessions made to semantics and the wealth of research devoted to Analysis of Function and Analysis of Meaning clearly point to the fact that integrational tendencies are gaining ground, and the uneasy position -- hinted at before -- between the formal and the pragmatic is now in the process of favouring the latter. Once integration is taken up with all its far-reaching and often unpredictable consequences, when we have accepted two-term hyphenation, such as socio-linguistics, neuro-linguistics, and even para-linguistics, how are we to handle three-term varieties, like neuro-sociolinguistics ? All the more so as 'the problem of interrelation between the sciences of man appears to be centred upon linguistics'. Granted that one of the 'in-mates' often evinces domineering tendencies, pregnantly illustrated by various degrees of capitalization, e.g. SOCIO-linguistics, to what extent is the hyphenated field still "linguistics" ?

6. With the emergence in 1957 of the generative-transformational approach to language study, it is adequate to say that the scope of linguistics has been expanded in such a way as to include 'vertical' structuring, and notions linked to degree of depth have gained wider and wider circulation. It seems that in the early 1970s yet another expansion, this time coming from the applied subdisciplines is taking place in the field of the general theory of language: in addition to verticalization, the detection of linguistic structuring over stretches of language wider than the Sentence is being accorded increasing attention. But the expansion of what I prefer to call structural span entails considerable difficulties. On account of the growing evidence in support of the rejection of the Sentence as the highest unit of linguistic description, the notion of structural span may become an important linguistic variable. But in the process of dealing with this particularly troublesome variable, there is so far no evidence whatever that the conventional procedures for the investigation of more restricted structural spans, such as the Sentence, may have any validity. To put the idea in the words of Seymour Chatman (1974), 'there is the domain of the Sentence, and the domain beyond the Sentence; there is no particular reason why the rules of one should duplicate or even resemble those of the other'. Faced, in other words, with Discourse as a structural span rapidly emerging within our research focus to form a distinct topic of investigation, there is no reason so far to believe that what I provisionally entitle Discourse Mapping can best be handled by the already evolved generative procedures of the conventional brand. Given the considerable complexity of types of information and controlling factors structure in discourse, e.g. linguistic, para-linguistic, non-verbal, pragmatic, silence phenomena etc,  it is only via the concerted and co-ordinated efforts of a wide range of disciplines that minimal results can be obtained.

7. So much for the boundary problem. Casting now a cursory glance at the question of correlation studies, or rather the lack of them, we anticipate that Discourse Analysis can be solidly developed and make progress if and only if  thorough correlation studies are undertaken, on a genuine basis, between the following branches of science subordinated, superordinated, or simply non-ordinated, to linguistics proper. Given the restrictions of space, the synoptic presentation of a set of targets for correlation studies is the most profitable solution (though perhaps the least convincing. Brackets give sample sets of categories to be submitted to such correlative analysis.

(1)    SYNTAX    (e. g. competence & performance, pronominalization, sentence boundary, the 'ceiling' effect)

(2)    PHONOLOGY    (e. g. suprasegmentals)    

(3)    SEMANTICS    (e. g. presuppositions, esp. pragmatic presuppositions, factivity, topic & comment, discourse reference, truth value, performatives)

(4)    PSYCHOLINGUISTICS    (e. g. motivation, intention, psychological reality of structural depth)

(5)    SOCIOLINGUISTICS    (e. g. role, role structure, role repertoire, repertoire range; turn, turn taking, participant boundary, adjacency pair, [degree of] institutionalization)

(6)    THEORY OF LANGUAGE TEACHING    (e. g. domain, language for special purposes; transaction, sequence, exchange, act, move)

(7)    THEORY OF TRANSLATION    (e. g. contrastive analysis of target vs source discourse structure)

(8)    POETICS    (e. g. context, style, stylistics, 'literaricity', macrostructure, casual vs non-casual utterance)

(9)    DISCOURSE ANALYSIS    (e. g. cohesion, coherence, congruence, connectedness, appropriateness, phoric reference)

(10)    SEMIOTICS    (e. g. verbal vs non-verbal in the whole range of discourse, syntactics & pragmatics, encoding processes, sign production)

8. Discourse Analysis and Text Linguistics also require correlation: though they have almost identical areas of investigation, both the theoretical apparatus and the respective methodologies are developing along diverging lines, the primary outcome of which is uncertainty of communication.  One of the important points to be taken into account as part of the correlation studies is the question of degree of reliance on either a data-bound approach or an intuition-bound approach to language study.

9. The present paper postulates, rather than demonstrates, that a dynamic interdisciplinary approach to discourse is possible and necessary on the basis of fundamental concepts such as Discourse Mapping, which is a complex socio-psycho-semiotic/linguistic category, not necessarily implying generation. Given the tremendous complexities of spoken discourse, particularly as compared with the apparent straightforwardness of written discourse, though 'writing is only a substitutive system,' as Roman Jakobson (1973 : 29) stresses, it is as yet far too early to say what the actual and specific parameters  of discourse mapping are. It has been stated by Chomsky (1964 : 112) that linguistics has to do ultimately with 'theoretical psychology';  in the light of developments in the last ten years in sociolinguistics, semiotics, pragmatics, it is perhaps high time to rephrase the goals of linguistics as having to do with 'theoretical social psychology'. And still lingering on the very last lines of Current Issues, it may be more correct to say: it is necessary to go far beyond the over-restricted framework of modern generative linguistics and the narrowly-conceived abstractionism from which it springs.

10. Is there 'a clearcut division between theoretical linguistics and applied linguistics,' as asserted by Janos Petöfi (1973 : 13)  ?  One important conclusion of the present paper is, on the contrary, that there are no clearcut divisions in  linguistics. It is precisely this kind of answer which tips the scales in favour of either autonomy or integration, excluding or including the whole area of applied study.

        The second conclusion is the obvious necessity to conduct genuine and thorough correlation studies, which, in its turn, is the direct outcome of an imperceptible switch from an autonomy approach to a more realistic integrational approach: we are at present witnessing this shift as a slow and steady process operating in one direction only.

REFERENCES: to Interdisciplinary Aspects of Discourse Analysis.

Chatman, Seymour    Report given at the First Semiotics Congress, Milan, June 1974.

Chomsky, Noam    Syntactic Structures  The Hague: Mouton 1957.

Chomsky, Noam    Current Issues in Linguistic Theory    The Hague : Mouton 1964.

Harris, Zellig    Structural Linguistics    Univ. of Chicago Press. 1951. Preface to the 4th impression.

Jakobson, Roman    Main Trends in the Science of Language. London: George Allen & Unwin. 1973. 75 pp.

Katz, Jerrold & Jerry Fodor    'The Structure of a Semantic Theory', in: The Structure of Language, Readings in the Philosophy of Language. Prentice Hall. 1964. pp. 479-518.

Lyons, John    Introduction to Theoretical Linguistics. Cambridge: University Press. 1969.

Petöfi, Janos & Hannes Rieser, eds.     Studies in Text Grammar. Foundations of Language, Supplementary Series, Vol. 19     D.Reidel Publishing Co. 1973. 348 pp.

Saussure, Ferdinand de    Cours de linguistique générale.  Edition critique par Rudolf Engler 1967.  Otto Harrassowitz, Wiesbaden, fascicule 1: 146 pp.; fascicule 2:  316 pp., fascicule 3: 515 pp. 

Saussure, Ferdinand de    Cours de linguistique générale. Edition critique préparée par Tullio de Mauro 1972  (translated from Italian) Payothèque, Payot, Paris 1916, 510 pp. 

Sinclair, John et al. Towards an Analysis of Discourse, The Language used by Teachers and Pupils. Oxford University Press. 1974.

Taber, Ch.    The Structure of Sango Narrative.  Hartford Studies in Linguistics. 2 vols.  Hartford Seminary Foundation. Hartford, Conn. 1966.