The concept named "Compositional Syntax" was first formulated and further developed in the Minsk State Linguistic University and more exactly at the department of History and Grammar of English.
This theoretical framework is an attempt to develop a united theory of Syntax. The presentation of the semantics of the sentence is based on several assumptions following from the general properties of the semantic aspect of a sign. These assumptions are:
1.The semantic sphere of any sign is devided into the significational and the denotational aspects. The significational aspect of the meaning of a sign is determined by the sytem of signs and shows the way in which the Speaker present the phenomenon she/he speaks about. The denotational aspect of the meaning of a sign is determined by our knowledge of the Universe and our complex idea of the phenomenon we speak about and depends on our experence.
2.Because a sentence is a model of some fragment of the outer world its semantics is nesseceraly a construction, which means that both the significational and the denotational aspects of the meaning of a sentence must be structures, relations of special elements.
3.The structures of the significational and the denotational aspects of the meaning of a sentence must be independent of each other. This statement needs some comments. Usually it is taken for granted that the structure of the significational aspect of the sentence meaning is a reflection of the denotational aspect and the difference between them is only in the degree of generalisation. The structure of the denotational aspect, usually termed as "situation" or "the state of affairs/things", is thought to be the first step of generalisation and the significational aspect, usually termed as "proposition", is believed to be the final step. It is believed that while creating a sentence the Speaker has in mind the structure of the situation and generalises it up to the class of the similar sittuations and finding the necessary general stucture and then words it, adds modality (or in other words – predicativity) and finally produces a sentence. The Listener does it in the opposite way – receiving a sentebce the Listener discovers the proposition (that is he/she becomes aware of the significational aspect) and through it idetifies the situation the Speaker had in mind. This seems quite natural and logical, but unfortunately does noy explain different and rather numeruos occasions of metaphoric use of sentences. And by this I mean not only literary (poetic) metaphores but those we normally use often forgetting that they are metaphores, e.g. "Time flies", "John discovered inconsistences in Mike's story" &c. Such phrases should not exist, if we accept the model of speech production and understanding described above.[3] Besides the same situation can be presented through radically different propositions, e.g. "Jack built the house rather quickly">>"Jacxk's construction of the hhouse was rather quick". This evedence shows that these two structures (significational and denotational) must be at least independent of each other.
4.The meaning of a sentence is a resutlt of co-ordination of these two structures.
5.Since these two structures (significational and denotational) are independent of each other and yet are elements of a united semantic entity, the sentence semantic structure should include a special system or mechanism for correlating and co-ordinating these structures.
These assumptions allow us to unite the most important ideas of the four theories of sentence meaning.
The significational aspect being the result of the systemic relations of the signs comprisies a limited number of structures (propositions), which can undergo a limited number of modifications. They resemble the kernel structures postulated by N. Chomsky in his Generative Grammar. The principle difference is that the propositions are semanticstructures, while N. Chomsky treated them only as abstract forms. Since these structures (propositions) are limited, they can be calculated and formal and semantic operations can be found to identify them in sentences.
Being logical structures the propositions consist of the predicate and arguments which are determined by the nature of the predicate (L.Tesniér, Ch.Fillmore).
The situations (the denotational structures) are the strucrures of our experencce and thus cannot be calculated and can be found only by inductine analyses and generalisation.
The general meaning of a senntence is the result of composing the significational and the denotational semantical structures, so the meanings of sentences and their components should be described in two sets of terms and are actually combinations of significational and denotational meanungs (the resulting descriptions resemble the syntaxemic analysys developed by A.M. Mukhin).
The previous text was too absract to be understood, so we shall apply the ideas to actual sentances.
We should start with defining the significational structures (propositions). As we postulated before propositions are a limited (close) set of structures opposed within the group by very general semantic features wich opposition is reflected in their modificational abilities. It means that we can use transformational technique to define different types of propositions. The transformations used for this should be very simple applicable to any sentence and having only two possibilities: either it can be done it it cannot.
There is only one such transformation – change of the positions of the subject and a complement (George is my father>> My father is George). Appluication of this transformation to devides all sentences in two groups. One group includes sentences which do not allow this transformation: John asked Mike‡ Mike asks Joh;. Jack has a house ‡ The house has Jack. The other group consists of sentances allowing such transformation:Jack looks like Nisk = Nisk looks like Jack; Jane meets Nell = Nell meets Jane.
This difference demonstrates that the relations between the nominal elements of the sentences are different. If the transformation is possible, the nominal elements are equal. It means that the relations between the nominal elements has no direction and we shall call sentences permitting the transformation of transmutation indirected.
If sentences do not allow this transformation we can conclude that the nominal element occupying the subject position is the initial point of a directed relation with the direction from the subject to the object. This can be proved by the fact that if we change the positions of the subject and the object of such sentences we receive either a description of an opposite situation (John asked Mike > Mike asked John) or an impossible situation (Dick broke the cup > The cup broke Dick).
But if we compare the sentences of these two groups we can see that each of them falls in two more groups. The difference of these two new groups is that one of them can be a sencible answer to the question "What happens?", while the other group cannot (What happens? – Nell meets Jane; John asks Mike. What happens? ‡ Jack looks like Nisk, Jack has a house). The first group, the group which can be answers to the "What happens?", includes sentences that indicate the dynamic representation of the state of things, and the second group which, includes sentences that indicate the static representation of the state of things.
The analysis shows that we can have four types of prposition:
Dynamic and directed. These propositions do not permit the transmutation transformation and can be answers to the "What happens?" question./e.g. John gave Jane an apple/
Dynamic and non-directed. These propositions permit the transformation and can be answers to the "What happens?" question./e.g. John met Dave/
Static and directed. These propositions do not permit the transmutation transformation and cannot be answers to the "What happens?" question./e.g.Jkane has a car/
Static and non-directed. These propositions permit the transmutation transformation and cannot be answers to the "What happens?" question./e.g. Mike looks like Nick/
These four classes of propositions have their own valence, that is their own deep case frames. Three of the classes (dynamic and non-directed; static and directed; static and non-directed;) have only two deep cases
[1] Historically these forms are even twice past because they have the Past Singular vowel of strong worbs and the Past suffix of the weak.
[2] This is a
[3] In fact, if we take this procedure of communication and sentence production we could not speak at all. If the significational aspect of the sentence meaning is only the final step of generalisation, then each type of situations should have its own abstract representation, the signification structure (prposition). But then any new type of situations should demand a new prposition, but since there is yet no new prposition for that new type of situatio we cannot speak about it, because old propositions cannot be employed to represent new types of situation, and thus cannot develop a new rpopsition for it. Thus we could speak only about things we all already know and no new information can ever penetrate human societies.