1978. "Text and Speech Act: A Propositional Analysis of Discourse." Denmark.

By C. George SANDULESCU, Stockholm.

(This paper was prepared for Det 4. nordiske lingvistmøde pa Hindsgavl Slot, Middelfart, Denmark, which took place between 6th and 8th January 1978.)




1.    The starting point is (a) a subset of Swedish texts, and (b) a subset of English texts in a clear Paraphrase Relation to each other (both within one and the same subset, and between the (a) and (b) subsets.

2.    On the basis of this informationally homogeneous set of texts, a system of theoretical constructs is devised, the most important of which are:

(i)    the Hypermorpheme hM, provisionally defined as a "conjoinerless" simplex sentence of the SVO type, e.g. 'The little student worshipped the Great Master (for a few hours)';

(ii)    the discourse Head dH, provisionally defined as a recurrent segment, evincing absolute hierarchical organization, directly derived from variable frequency of occurrence, and determining both discourse structure and illocutionary force at discourse level.

3.    The overt and the covert discourse Heads of a text can be assigned a whole range of Representational Fields on diagrams specially devised in order to acomodate both assertions and discourse presuppositions. Special attention is paid to the diagrammatical representation of REFERENCE versus PREDICATION (cf Searle 1989 : 72 -123, and Sigurd 1974 : 18).

4.    The conclusion of the study manifestly points to the existence of at least three distinct types of structures, both information i and grammatical g (out of a possible number of four, namely Sg(b), Sg(a); Si(b), and Si(a) ). It is the correlation between these widely different types of structure that should ideally form the object of a linguistic model.

        The basic task of Discourse Analysis is, then, to correlate below-the-sentence parameters with above-the-sentence factors, without losing sight of the balance between propositional content and illocutionary force.

        Another important conclusion is that the notion of discourseHead is a pragmatic category, largely dependent on the nature of the speech act.

        Finally, the last conclusion is that the paraphrase relation should be as important to the linguist as it is to the philosopher.

5.    REFERENCES: Familiarity with John Searle's Speech Acts, Cambridge U.P. 1969, and with Bengt Sigurd's Experiment med Text (PILUS, No. 25, October 1974) is advised for an easy understanding of the frame of reference. Passing mention will also be made to Nils-Erik ENKVIST's  notion of Theme Dynamics, as outlined in his Linguistic Stylistics, Mouton, 1973, pp. 115-27.

        The study is also grounded on the twenty documents circulated within the Working Group of Speech Acts at the Vienna Congress of Linguists in late August 1977, and on the papers presented in the Section Text Linguistics: DIALOGUE of the same Congress.

        (Comprehensive handouts accompany this presentation.)