Національна Академія Мистецтв України Інститут культурології

Therefore transformations as such can’t be evaluated as a way to some “deep structure” as the premise for textual integration. The conjecture of the kind would turn out to be erroneous. They can only provide conditions for textual growth or shrinkage, expansion or compression but the essence of text, the integrative nucleus can’t be disclosed with them. Moreover as to the nature of this suggested textual integrative nucleus one could doubt whether it remains invariant. Perhaps the most persuasive refutation of the transformational – generative conjecture about this abstract “deep structure” would be found in the writers’ creative workshop. Be this “kernel” structure comparable to an author’s draft, so it must encounter the resistance of stuff within the very procedures of its transformation in generating text. It is this resistance that essentially changes the primary draft (and, respectively, the deep structure), therefore it can’t be taken for something with a finite form, still more can’t it acquire the form of a constant statement¹⁵¹. There are numerous writers’ witnesses confirming the absence of something stable and constant that would precede the creative process¹⁵². The same reason of resistance concerns the problem of retaining the semantic identity of a sentence to its transformations, and it is textual referential ties of a sentence that determine this resistant power¹⁵³. Thus one has no reasons to take the integrative textual “nucleus” for a known abstraction; rather it must be regarded as a problem always demanding permanent exploration. It is not the whole this problematic object but only its separate abstracted attributes that make up the invariant retained in the transformations. Invariants can be only attributes of transformation that are still to be separated and abstracted. Therefore the mission of textual transformability can’t be the variable representation of a pretended invisible abstract deep structure, the very invariability of it necessitating proof. The genuine task of transformation would be also the developmental generative & interpretative work. In particular the paragon for such productive procedures can be found in translation that could be regarded as the development of the translated source and not its passive reproduction.

As to the cited samples of transformations as attempts to represent the invariant textual “nucleus”, it would be appropriate to regard them as the reflections of ancient rhetorical devices of amplification. If the supposed abstraction (taken as invariant) had really existed, then it would have coincided with a predestinated scheme chosen for amplification. Then the relation between deep and surface structures would have correlated with that between construction & decoration so that the generative process would have acquired an outlook of embellishing the predestinated scheme. Then one could remark that the cited statements haven’t taken into account and disclosed all the opportunities that such transformational procedure gives at one’s disposal (in particular it concerns the use of epithets)¹⁵⁴. No need to say that textual structure then would have become conceived in terms of reduction. Meanwhile “formation precedes transformation” (as S.D. Katsnelson has wittily noticed in his afterword to W. Chafe’s book), therefore schemes for amplification can’t be taken for the beforehand predestinated invariable abstractions. Then transformations are to be regarded as the developmental textual device where the “nucleus” would become problem to be explored and not the ready abstraction to be represented. Respectively the transformational invariant must here be conceived as the attributive (accidental) and not the substantial entity, therefore it can’t become some “preexistent essence” in the manner of a construction to be supplemented with decoration.

In this respect invariant should be described with the concept of function so that transformational grammar could be generalized and comprised with the functional grammar. Functional approach enables introducing morphological basis for textual transformations not only conceived as the predestinated metamorphoses but also represented within the textual integration. The very essence of function being the relation of the part and the whole, it becomes out of question to use functional concepts without the concept of integration. Therefore one obtains together with the functional approach the instrument for the disclosure of the common fundament of transformation and generation. A very succinct description of the procedure aiming at finding transformational invariant has been represented by A.V. Bondarko – the genuine founder of the functional grammar who involves here the concept of prototype as the basis for determining this invariant¹⁵⁵. A textual element can be selected with the conjecture of its invariable properties that are to be verified with the examination of textual transformational versions regarded as the representations of this supposed prototype. Apparently such transformational variants become then the covariants in regard to such selected prototype and behave as its periphrastic descriptions. One can say also of the primary and the derivative representations of invariant in regard to prototype’s representations. Transformational grammar discloses here its common roots together with the rhetoric means of homiletics where textual transformability is used as the developmental generative force. Together with prototype the concept of denotative role (coming back still to L. Tesniere) is to serve in making invariant more precise. To say of roles one has to discern the so called situation (as the initial source of textual contents) and its different interpretative representations in the same way as it takes place in dramatic play between the script and performances¹⁵⁶. Respectively one discerns the levels of the invariant as the denotation uniting different utterances of transformational variants and of the representations as autonomous textual segments taken irrespectively to their reciprocal ties as variants¹⁵⁷. It is within the interpretative representational level that the concept of rile appears as the functional destination of respective textual units¹⁵⁸. One can easily recognize the generalization of syntactic functions of subject, predicate etc. in these roles.

In its turn invariant being regarded as the foundation for representing transformations as variants of some identical essence, it presupposes its reproducibility. Thus the fundamental antinomy of language returns in the outlook of reproduction as the invariable basis for transformations. Generally speaking an invariant (in difference, say, to invariable terminal editorial version) must belong to the metasystem of language as far as it deals with abstract attributes and is represented with the artificially concocted descriptors. Meanwhile it can also be conceived in another way, without involving the concept of metasystem, as the latent and possible properties that aren’t and can’t be manifested beyond separate variants but that can and must still be reproduced. Apparently in this respect invariant is comparable to the well known concept of type in literature that also is to be conceived only within its representations in separate individual personalities. Such mutuality of type & invariant can exemplify the peculiarities of the functional approach.

The functional approach enables avoiding the involvement of metasystem as it aims at disclosing the latency of functions. In its turn it is through invariants as the properties retained in transformations that functions can first be revealed. It is therefore the criterion of reproducibility & conservation that that decides upon the determination of functional invariants. Furthermore, an invariant can’t exist as a separate isolated feature (property) and always presupposes the involvement of the whole attributive space of the respective code. Invariant exists as a separate representative of a set of attributes integrated in the system of code. The very existence of invariant as the revelation of conservation & reproducibility always entails integration and subsequently the hierarchy of field structure with its “center - periphery” division. Meanwhile this integration of reproducible hierarchy of code is only the one side of invariants’ appearance. Invariants presuppose also the integration of those particular variants where they are revealed as the set of intersections of different transformations. The integration of the set of invariants presupposes also the integration of covariants that represent invariant and encircle it as its transformations. Each textual entity can exist only of invariants as the reproducible elements of code so that while producing any new text the invariants are necessarily involved. Each text can therefore be represented as the separate covariant of the set encircling the invariant. The opposition “invariant vs. covariant” reveals thus the fundamental antinomy of language.

To cope with the problem of reproducibility as the property of code (paradigm) A.V. Bondarko has suggested the concept of the functional semantic field (as a particular taxonomic class of <функционально-семантическое поле ФСП>) in opposite to the already mentioned concept of situation (or categorical situation <КС>) where the category is apparently supposed to be invariant represented in the given situation of text. The both of them are reciprocally tied and preclude any priority of one of them¹⁵⁹. There are at least two important distinctions from the usual taxonomy of semantic fields that ensue from functional approach: it goes about developmental historical properties and the inner ordered set of attributes associated with speech units¹⁶⁰. In difference to the commonly acceptable concept of semantic field as a reproducible taxonomic class it is functional categories and therefore integrative moments that play here the decisive role. Respectively categories as the definitive power involve here attributive space so that i t is attributes that become decisive for categories. Attributive definition of categories (resp. of taxonomic divisions) entails the necessity of attributive (and subsequently functional) approach¹⁶¹. While identifying categories and attributes one comes thus to their different representation within the Signifier and the Signified¹⁶² as well as within the speech – language dichotomy¹⁶³. In particular it is only within code that attribute marks the respective category as a whole whereas text always represents the representational selection of features¹⁶⁴. All it determines the priority of attribute as the foundation of category¹⁶⁵. Therefore attributive space is not a set of elements void of connections. The zero hypothesis of independent features is not valid for the verbal map of world. Rather it goes about the tissue of reciprocally intersected and interlaced classes where each one entails another¹⁶⁶. It is here that the principal distinction from the so called semantic component analysis lies: features are by no means independent and isolated units as in that case. Vice versa attributes conceived as functions must necessarily become mutually interconnected, the functions being the revelations of something integrated.

This substantiation of the necessity of attributive approach to the analysis of categories and the respective fields entails also the respective approach toe the introduced concept of categorical situations. To support the mentioned parallel between invariant and type it would be persuasive in this respect to refer to the cited M.V. Vsevolodova who has also suggested the concept of typical situation used as the generalization of textual situation conceived as the revelation of the latent invariable contents. The convenience of such generalization becomes apparent from its applicability to artistic text where categories are of diffuse nature and situations are in wide range determined with the position of the given textual segment within the whole so that the concept of type becomes preferable to that of invariant. The necessity of the notion of categorical situation is substantiated with the irreducibility of information to the additive combination of its ingredients so that situation is conceived from the very beginning as something correlated with integration. In particular this integrative essence of situation can be attested with the most declarative situational utterance that represents existential situation¹⁶⁷. Therefore one have grounds to say of signifying and denotative situations and of still further division of situational attributes¹⁶⁸. Such are especially the situations attached to modality, aspect, voice and other categorical attributes of verbs.

The reciprocity between field and situation based on the attributive representation of categories enables revealing the both sides of the fundamental antinomy so that text and code (together with the Signifier and the Signified) are taken in their mutual transitions¹⁶⁹. The transitions between inner and outer forms are the existential conditions of language. Therefore the conjecture about any kind of preexistence & priority of reproducible invariant is out of question: it coexists together with its representations being derived from them as well as they are derived from it. In its turn this reciprocity concerns also the mentioned semantic levels of the denotation and its representations conceived now morphologically as the inner and outer form¹⁷⁰.This statement of the coexistence as opposed to the preexistence of generative approach enables conceiving the transformability in wide range as the revelation the universal property of the homology of language¹⁷¹. Transformations as the covariants representing a prototype (from the viewpoint of functional approach) disclose the interpretative (resp. derivative and productive) opportunities given with this prototype as the invariant of the reproduction within a code and at the same time ensue from the conditions of the integration of text. Transformability discloses interpretability (with integration) and reproducibility as the functional properties. Transformations are not only admissible but also necessitated as the functional properties of language¹⁷². Therefore the coexistent periphrastic descriptions are to be taken for the normal existential form of textual entity. Such coexistence reveals itself in particular through correctional and explanatory transformational variants that arise as the constant satellite background in the manner of the possible phrasal synonyms and are virtually present within the textual presupposition¹⁷³. Besides, it is the necessity of commenting the situation that gives rise to transformation as the necessity to call the category which this situation concerns. In particular invariant is then to be conceived as the functional side of phenomena existing only as the metamorphoses.

Such morphological approach enables conceiving textual transformability as the disclosure of textual opportunities that must not necessarily be manifested. Transformational covariants appear together with the given textual variant as its virtual satellites or epiphenomena. There are admissible variants that are presupposed as the possible satellites coexisting virtually together with the given text. Functional properties always deal with this opposition “possibility vs. reality” that intersects in its turn with the opposition “latency vs. manifestation”. It is this space of possibilities that enables the coexistence of latently present different variants that can (but must not!) be derived from the given text¹⁷⁴. Obviously there exists also the poetic necessity together with these possibilities, and it is this necessity that precludes any deviation from the given textual variant. These “possible worlds” of language as the textual latent satellites partly intersect with code but don’t coincide with it as far as there are the unforeseen and therefore irreproducible possibilities.

Apparently “possible worlds” deal with attributive space, and it is attributive approach that presupposes the involvement of the respective functional conception. The difference between manifestation and possibility is the crucial point in the functional approach: A.V. Bondarko has suggested the division of function into the potential (Fp) and the resultant (Fr)¹⁷⁵. The productivity of such possible worlds (in opposite to code’s reproducibility) is attested with the developmental nature of the transition between these functional classes. This productivity is to be found in interpretative derivative activity as the generative productive procedure. Together with teleology the causal nature of functions becomes evident. It is the difference between possibility and result with the respective functional distinctions that entails the presence both of purpose and cause. The purposefulness of functions attests the revelation of integrative processes and the attachment of functions to textual entirety (as well as the entirety of code) whereas the transition from the possible to the real discloses the causation of textual generation that correlates with the mentioned division of functions into Fp and Fr¹⁷⁶. In its turn potential function is attached to form (as the opportunities of form) nearer than the resulting function where the interaction within the speech synthesis takes place¹⁷⁷. The distinction of goals and means plays the decisive role for functional approach and it immediately concerns the unity of the pair “Form – Function” where the purposefulness and the entirety disclose their inner interconnection¹⁷⁸. One can say of interpretative and representational relations that tie together the possible worlds of latencies and their actual revelation through textual situations. It is of importance that situational approach is the immediate consequence of the functional approach so that one could say of situational – functional analysis based on attributive representation. This functional attachment of situations reveals itself in the selection of special attributive features that become prevalent so that the notion of dominating situation arises¹⁷⁹. Another substantiation of the functional essence of situations can be found in the division of the situations into those with functional concentration (focus) and with the dispersed functions¹⁸⁰.

Attention then would be paid to the relations of text to propositional structure and to the aspects of textual irreducibility. To deal with this problem one would remind that the decisive element of propositional structure is predication as the special function that imparts asymmetry to text disclosing its field structure of the confrontation “center – periphery”. The discussed transformability’s properties show that such asymmetry doesn’t remain static and constant. The function of predication is the peculiar syntagmatic function absent within paradigm¹⁸¹. This textual attachment is supported also with the supposed absence of this function within the so called deep structures of semantics¹⁸². It entails in its turn the specification of the functional destinations with the divisions of modality, aspect, tense etc. that involves not only propositional structures but also textual situation as a whole¹⁸³. The particular place of predicative functions is attested also with their peculiarity as the denotative roles¹⁸⁴. It is here to add that the very fact of functional differentiation and succeeding opposition of textual elements imparted with predication results necessarily in the communicative dialogical relations arising between the predicated and the predicating. The most primitive verbal phrase can be represented as a dialogical cellule of question and answer represented with nominative sentences as in the example <The tree flourishes> → <– (What’s about) the tree? – (Oh, it) flourishes>: here the words in brackets (necessary for grammatically correct form) are omitted in “telegraph style” so that the predication really dissociates into a couple of cues. The most general consequence of such predicative split within the speech flow is that the speech’s segments are opposed as the proper and the alien ones. Predication provides thus the conditions for reflection as the division of textual elements into the classes that are in the relations of alienated speech. Therefore the communicative division of speech’s elements into those of the author’s utterance and the cited quotations of “direct speech” ensue already from this predicative dissociation.

Historically predication is the first step in transforming incorporated language structure towards more developed forms. It occurs when the binary division of the primary word-sentence comes to existence¹⁸⁵. The next step becomes the formation of pronominal attributes of predicate¹⁸⁶ so that the pronominal type of language structure presuming the mentioned deictic net of references precedes the nominative type in proper sense. These steps enabling the development of predicate don’t still presume the formation of verb as such¹⁸⁷. It is the pronoun that gains key role in separating verb from earlier predicate¹⁸⁸ so that predicate is endowed with pronominal functions while transforming into a verb. Moreover there are also grounds to see in the verb itself the initiative element of such transformation¹⁸⁹. Imparting the initial point for transformations verb gains an obvious leading role in building textual entity within the synchronous conditions in the nominative types of languages where it correlates with nouns as the dependent element¹⁹⁰. The leading role of predicates in the folding and unfolding procedures is caused with their role as intermediary layer that presumes further growth to narratives’ plots that they fold. The possibility of such expansion is based upon the ruling properties of predicates that evoke other lexical units continuing them so that the verbs’ ability for growth and expansion reveals itself most productive¹⁹¹. In opposite to abstract substantives verbs endow textual entity with synthetic forces aiming at concrete map and thus contribute in overcoming this abstractedness.

One ought to stress that to the already discussed deficiencies of the semantic studies belongs obviously the lack of attention towards verbs in special vocabularies. Thus the verbs are either excluded or substituted with substantive forms in the both mentioned dictionaries of poetic images [Павлович, 1999; Иванова, 2004]. The same concerns the semantic dictionary where also the preference for substantives entailed such substitutions as любовь (love) instead любить (to love), течение (flow) instead of течь (to flow) [Караулов et al.]. Meanwhile it is the semantic derivation and respectively the generation of poetical images where the verbs give especially rich and valuable stuff. It suffices to refer to prefixes in the Slavonic languages that radically change the meanings of the roots: Ukr. розгубитися (to lose > to get confused), Russ. загнуть (to bend > to invent), утрясти дело (to shake > to arrange the affair), Pol. naciskac (kogo) (to press > to compel), Old. Sl.