The syntax of sentence and text: a festschrift for Frantisek Danes. (Reviews).

The syntax of sentence and text: a festschrift for Frantisek Danes. By Svetla Cmejrkova and Frantisek Sticha (eds). Pp. 393. Amsterdam-Philadelphia: John Benjamins, 1994.

In the age of information explosion, it is difficult to keep track of all developments in any given field, particularly when the field is subject to paradigm pressure. That is one of the reasons (as well as political situation) why, for a long time, the achievements of the Prague School linguistics remained largely unknown (with notable exceptions) in the West. Although the book under review is a tribute to individual accomplishments of Frantisek Danes, it is also a sign of general recognition and appreciation of the whole linguistic school.

As usual and expected in a festschrift volume, contributions are quite diversified as to the content. Some are of a more general nature, and some very specific, though not, of course, without general implications. The number and variety of contributions make it difficult for the reviewer to give an exhaustive, adequate and balanced account of all papers, and selection is not easy. I have decided to give a brief account of only some contributions of more general character.

The contributions are organised into four sections: Prague School Functionalism. Functional Sentence Perspective and Thematic Progression, Text and Discourse, and Grammar and Semantics, all topics reflecting the enormous range of Danes' scholarship.

In the opening paper de Beaugrande discusses two approaches to language analysis: functionalism and formalism. He emphasises the largely underestimated proposals of the Prague school, among others the early awareness of the advantage of an interactive model of language over a modular one, as well as Danes distinction of three levels: the organisation of utterance, the semantic structure of sentence, and the grammatical structure of sentence.

One of the points de Beaugrande addresses is a methodological problem of linguistic analysis and language model. De Beaugrande writes that

a strict separation of levels would prevent an understanding of how any one level functions, even on us own terms. At best, a strict separation could be simulated after the fact once the data had been analysed from an interactive outlook. And this tactic might still entrain us in an arbitrary, often redundant reconstruction of parallel criteria and patterns for each level under different terms. (de Beaugrande 1994: 35).

It is, of course, trivial to say that any language entity is a multilevel construct, and additionally NOT a simple aggregate of levels and factors. Quite possibly it may be the "chicken-egg" problem, but I believe that the very identification, however intuitive and initially vague, of any single level or element is necessary to define its main function in which it necessarily interacts with other levels or elements, producing ultimately a complex structure.

Once levels or elements have been distinguished, they can not only be studied in their contextual interactions with other levels and elements, but also in their own terms, with other levels and elements kept as constant as possible. Thus separation should be looked upon as a convenient tool, perhaps an initial shortcut, not an end in itself.

Enkvist's "Centre and periphery, delicacy and fuzz" is a very central, delicate and non-fuzzy account of the major tendencies in modern linguistics, showing at the same time the relevance of earlier Prague School studies in the context of those new tendencies. In his extremely subtle study Enkvist addresses a problem similar to that raised by de Beaugrande, bringing to our attention once again the fact that quite a few "hot" ideas in contemporary linguistics were taken up much earlier by the Prague School. He quotes Danes (1966: 11) as writing what is regarded as one of the basic principles today:

All such insoluble dilemmas and useless disputes can be clearly removed if one gives up the common-notion of strict compartmentalization and simply admits an obvious idea, viz, that the organization of linguistic elements in the "system of systems" has a different nature, a different structure, than is usually assumed: the classes (and sub- classes) of elements should not be regarded as "boxes" with clear-cut boundaries but as formations with a compact core (centre) and with a gradual transition into a diffuse periphery which, again, gradually passes (infiltrates) into the peripheral domain of the next category.

Our conception does not, of course, deny the existence of classes or categories, but at the same time it does not force us into unambiguous decisions in those cases where the decision has not been made by the language ...

Enkvist also emphasises the processual aspect of language, referring to his own studies (e.g. 1986), as well as to Humboldt's energeia, and Mac Whinney's (1989) Competition Theory. He also recalls Schauber and Spolsky's (1986) ideas which he includes in what he calls "preference semantics". I think that in the context of the above mentioned proposals, due regard should be paid to de Beaugrande's various studies, perhaps particularly to de Beaugrande and Dressler (1980), where they underline the dynamic, probabilistic (preferences) and interactive nature of language, adding to the description the "problem solving theory" (decision-making in the Competition Model).

In his paper "Information Structure in Writing" Martin Davies discusses a very important problem of signalling information organization in written discourse concluding that

There thus seem to be grounds for believing that although there is no notation for intonation in the writing system of English, the grammar and lexis do make available sufficient information to enable readers - when reading - to discern the information structure the writer has in mind with sufficient precision to make it possible to know when they have got it right. And when they have made a mistake, they will KNOW they have, and will be able to say, "I'm sorry, I'll read that again. (Davis 1994: 88)

Wolfgang Dressler relates naturalness theory to de Beaugrande and Dressler's text model I referred to earlier. In particular he discusses some aspects of de Beaugrande and Dresslers standards of textuality in relation to FSP. Probably for lack of space, Dressler leaves undiscussed the most interesting part of the relation. He has advocated (e.g. 1990) general preference for binary relations (see also Szwedek 1985, 1990), which is manifested in the polar opposition between T proper and R proper. He does not however, take any stand on the fundamental, long standing assumption of FSP, i.e. CD and its graded character.

This is precisely the problem taken up in the next contribution by Firbas on Dancs" view of givenness as a graded phenomenon. Firbas bases his notion of "givenness" on the relation of retrievability. This raises the question of the length of the text stretch from which information remains retrievable, activated. Apart from the fact that the problem had been earlier addressed by Osgood (1971), it is worth noting that I see a significant change in the interpretation of FSP. Progression which is such an important part of CD seems to have taken on a new dimension. Instead of a number of degrees of contextual dependencies, Firbas presents a clearly prototype effect description; he admits that "in principle the opposition of context dependence and context independence is preserved" (cf. Szwedek 1985, 1990); however, it is possible that within the category of context dependence (givenness) there are more and less prototypical instances, giving the impression of gradual distinctions. That important modification is in perfect consonance with DressIer's (1990) "preference for binary relations" (however, cf. also Szwedek 1985, 1990). By the way, I think that the "binary relations" view is a very promising proposal; apart from the fact that we can find lots of binary relations (oppositions) (marked-unmarked, given-new, topic-comment, things-relations, figure-ground, foreground-background, things-relations, temporalatemporal relations, etc.), all of them can be traced to the basic cognitive process described by Langacker (1986) as comparison: "Fundamental to cognitive processing and the structuring of experience is our ability to compare events and register any contrast or discrepancy between them acts of comparison continually occur in all active cognitive domains, ..." (Langacker 1986:101). Langacker further writes that an act of comparison "has the general schematic form S T, where S can be called the standard of comparison and T the target." Thus any act of cognition is dichotomous in nature involving in its foundations a contrast between two entities. This would account for the role of "binary relations" in language. That fu dichotomy would then be subject to modification depending on the nature of both the standard and thc target, leaving the impression of gradation. This is what I advocated for many years with respect to information organization in the sentence (e.g. Szwedek 1976. 1985. 1990). Langacker also remarks that "the comparison of two events need not involve them as unanalyzed wholes; it may instead pertain only to certain facets of them or their manifestation in particular domains" (Langacker 1986: 104). He finally concludes that "This capacity for selection is also quite important for understanding semantic and grammatical structure".

Luelsdort's contribution is a reinterpretation of thematic progression in terms of determinacy grammar, a new theory, judging by references. It makes quite difficult, technical reading, even more difficult by either lack of clarity or inconsistent use of symbols. For example, D is used for determinacy (throughout the paper), but also for directionality (p. 152); R is used for rheme, as well as resilience (p. 148), P for progression. as well as preceding (p. 152), B for position (p. 152), as well as binder (p. 153). Petofi's contribution is in consonance with his earlier encyclopaedic approach to text analysis. His study is extremely interesting, full of technical details, showing all complexities of text constitution as basis for interdisciplinary studies, among others, because of its multimedial character. This new approach he proposes to call Semiotic Textology to refer to the most broadly comprehended text analysis and text description.

In conclusion, that interesting volume can be recommended to linguists of many persuasions, as it not only increases our knowledge of linguistic past, but also very much so about the linguistic present.

REFERENCES

Beaugrande, Robert de -- Wolfgang Dressler

1980 Introduction to text linguistics. London and New York: Longman.

Dressler, Wolfgang

1990 "The cognitive perspective of "naturalist" linguistics models". Cognitive Linguistics 1: 75-98.

Enkvist, Nils Erik

1986 "Linearization, text type, and parameter weighting.", in: Jacob L. Mey ed.).

Langacker Ronald

1986 Foundations of Cognitive Grammar. Vol. I. Stanford, California: Stanford University Press.

Mey, Jacob L. (ed.)

1986 Language and discourse: Test and protest. A festschrift for Pctr Sgall (= Linguistic and Literary Studies in Eastern Europe 19). Amsterdam and Philadelphia: Benjamins.

Osgood, Charles

1971 "Where do sentences come from?", in: Danny Steinberg -- Leon Jakobovits (eds.), 497-529.

Steinberg, Danny -- Leon Jakobovits (eds.)

1971 Semantics: An interdisciplinary reader in philosophy, linguistics and psychology. Cambridge: Cambridge University Press.

Fisiak, Jacek (ed.)

1990 Further Insights Into Contrastive Analysis. Amsterdam: Benjamins.

Pieper. Ursula -- Gerhard Stickel (eds.)

1985 Studia Linguistica Diachronica et Synchronica. Berlin: Mouton de Gruyter.

Szwedek, Aleksander

1976 Word Order, Sentence Stress and Reference in English and Polish. Edmonton, Canada: Linguistic Research.

1985 "Sexpartition versus bipartition", in: Ursula Pieper and Gerhard Stickel (eds.), 801-809.

1990 "What is topic? A contrastivist's view", in: Jacek Fisiak (ed.), 499-506.