Theoretical and Historical Background

Structure of Discourse

IDA Methodology

Applied Uses of IDA

Bibliography

 


Theoretical and Historical Background

The conception of integrational discourse analysis was first suggested by Viatcheslav Yatsko in 1995. Its full description can be found in V.Yatsko's monograph "Reasoning as a type of scientific discourse" (Abakan, 1998, in Russian). The methodology of analysis is available in English in V.Yatsko's paper "Compositional modeling: logical linguistic principles of a new method of scientific-information activity" published in Automatic Documentation and Mathematical Linguistics. 1995, VOL 29; NUMBER 6, pages 23-30. This site gives an outline of the conception.

The theory and methodology of integrational discourse analysis are the result of historical development of discourse analysis as a linguistic discipline. This development comprises two main periods. The first one, which in T.Kuhn's terminology may be described as a normal period, had lasted from the end of 60s till the end 1980s. This period is characterized by predominance of the ideas of communicative syntax, working out the notion of super-phrasal unit, studying lexical and grammatical manifestations of connections between sentences in a coherent text. At that time major contribution to the development of discourse analysis was made by such prominent scholars as W. Mathesius, F. Danes, W.Dressler, de Beaugrand. They introduced into linguistics the notions of actual division of a sentence, topic, and focus. The terms “text grammar,” “text linguistics,” “discourse analysis” had become generally accepted to denote a branch of linguistics dealing with analysis of coherent texts. By the beginning of the 1980s a lot of research had been done in the field all over the world, the results of which were reflected in numerous papers, proceedings of the conferences, monographs, textbooks (see in the Bibliography works of the above mentioned authors). The essential difference between discourse analysis and text grammar is that the former focuses on revealing psychological and social characteristics of a person through the analysis of the person's speech. A typical example is investigation of ethnic prejudice in Teun A. van Dijk works. Actually discourse analysis is a semi-linguistic methodology in which analysis of text structure is a means of revealing some socio-psychological characteristics of a person. The final aim of research done within scope of text grammar is revealing the structure of coherent texts. Integrational discourse analysis conception falls within scope of text grammar rather than discourse analysis and the term 'discourse' is used in its generalized meaning to denote any coherent text.

By the end of 1970s – beginning of 1980s it turned out that the communicative approach had some essential drawbacks. One of them is the inability to account for the sentences that are coherent, but senseless. For example the sentences It is summer. It is a table. It is difficult are coherent because of the repetition of subjects, nevertheless they make no sense. On the other hand there can be examples of sentences that don’t have manifestations of connections between them but seem perfectly understandable. For example: He wants to write a play for me. One act. One man. Decides to commit suicide. (Mansfield) Such facts made linguists differentiate between coherence and cohesion. The latter is understood as some deep level of relations between sentences or utterances though the understanding of this level (and of cohesion on the whole) is still vague. Perhaps the most general definitions of cohesion and coherence have been given by Bill Mann: “Cohesion has to do with how particular texts in effect have links within them, from one part to another. Various languages have different “cohesive devices” for expressing such links.”…Coherence has to do with an impression of wholeness. It is expected of texts, but not text fragments.” [RST archives, http://listserv.linguistlist.org/cgi-n/wa?A2=ind9912&L=rstlist&D=1&P=613]

The study of cohesion gave rise to a number of theoretical conceptions, which concentrate on relations between different components of discourse. One of the first such attempts was Grimes’s conception of rhetorical predicates as explicit organizing relations used in discourse . Here follows the list of the rhetorical predicates, specified later by McKeown.

1. Attributive:

Mary has a pink coat.

2. Equivalent:

Wines described as great are fine wines from an especially good village.

3. Specification (of general fact):

Mary is quite heavy. She weighs 200 pounds.

4. Explanation (reasoning behind an inference drawn):

So people form a low self-image of themselves, because their lives can never match the way Americans live on the screen.

5. Evidence (for a given fact):

The audience recognized the difference. They started laughing right from the very first frames of that film.

6. Analogy:

You make it in exactly the same way as red-wine sangria, except that you use any of your inexpensive white wines instead of one of your inexpensive reds.

7. Representative (item representative of a set):

What does a giraffe have that's special?… a long neck.

8. Constituency (presentation of sub-parts or sub-classes):

This is an octopus… There is his eye, these are his legs, and he has these suction cups.

9. Covariance (antecedent, consequent statement):

If John went to the movies, then he can tell us what happened.

10. Alternatives:

We can visit the Empire State Building or call it a day.

11. Cause-effect:

The addition of spirit during the period of fermentation arrests the fermentation development…

12. Adversative:

It was a case of sink or swim.

13. Inference:

So people form a low self-image of themselves.

The conception of rhetorical predicates is based on confusing two essentially different types of relations: a predicative relation, i.e. a relation between the verb-predicate and its arguments, and a logical relation between judgements expressed by sentences. Attributive, equivalent, cause-effect, and adversative in the above list are predicative relations describing different semantic types of propositions whereas the rest of relations are logical relations between judgements, which determine the type of discourse (see Structure of Discourse).

One of the most recent developments is rhetorical structure theory (RST) formulated by Bill Mann and Sandra Thompson . The authors distinguish such relations as Background, Elaboration, Preparation, Contrast, etc. These relations occur between spans of text, one of which, more central to the authors purposes, is called a nucleus, and the other span, which is not so essential, is a satellite. For example, the Evidence relation occurs between Claim (nucleus) and Evidence (satellite), i.e. information intended to increase the reader’s belief in the claim; Summary is a relation between Text (nucleus) and a short summary of the text (satellite).

The main drawback of RST is that the authors don’t give any logical or linguistic distinctions of relations and spans of text, and the criteria for distinguishing between them are not clear enough. One might guess that the difference between Claim and Evidence is based on some (vague) semantic criterion, whereas the difference between Text and Summary seems to be based on some physical characteristics of text because a summary, being typeset in a different (smaller, as a rule) font, is usually separated from the rest of the text by some blank lines. Lack of comprehensible criteria makes the differentiation between relations and spans of text a purely arbitrary undertaking, and the number of relations seems to be unlimited. For example, why not distinguish the relation of Paragraph 1 between the first and the second paragraphs of text; the relation of Paragraph 2 between the second and the third paragraphs, etc.? Lack of logical and linguistic foundations diminishes the RTS’s scientific value.

Integrational discourse analysis conception presented on this web site is another attempt to reveal relations underlying the structure of discourse. The general idea constituting the IDA’s theoretical and methodological background is that when speaking or writing a person intuitively arranges judgments expressed by sentences in some logical order thus imposing some logical relations between them. It should be emphasized that these logical relations have little in common with the relations studied by formal logic. Actually one doesn’t need to have any knowledge of formal or traditional logic to write a letter or a scientific paper, and most of the authors are not aware of the rules formulated by formal logic. IDA’s goal is to reveal this intuitive logic of speakers and determine how it is manifested in linguistic structure of discourse. The specific feature of IDA’s methodology is exploration of four dimensions (or planes) of discourse structure: semantic, communicative, modal, and relational (see the next sections of this web page for detailed characteristics and examples of analyses).

T. Kuhn has distinguished two stages in the development of a scientific discipline: the normal period and the revolutionary period. The normal period is characterized by the predominance of one theoretical conception constituting the scientific discipline’s paradigm; the revolutionary period is characterized by the competition between a number of different conceptions. Discourse analysis seems to be experiencing the revolutionary stage of its development, and integrational discourse analysis conception is intended to make its contribution to the competition between different conceptions, to establishing of a new paradigm in the field.

Structure of Discourse


Within IDA discourse is considered as a multidimensional structure comprising four main planes: semantic, communicative, modal, and relational. Semantic plane of discourse is its propositional content; a proposition is an abstract monopredicative syntactic structure active and affirmative distinctly expressing a certain type of judgement and having only obligatory verb arguments.

At present IDA differentiates between 15 semantic types of propositions whose distinctive features are represented by table 1. A-model includes names of verb arguments and reflects semantic distinctions of a proposition. F-model includes names of syntactic functions of arguments (see explanatory notes below the table) and reflects syntactic distinctions of a proposition. The propositions are classified into those of the category of state and those of the category action. They are also divided into one, two, three member according to the number of obligatory verb arguments.

The analysis of the semantic plane of discourse within the scope of IDA consists of reducing sentences in natural language text to the semantic types of propositions. This analysis meets with two main problems. The first one is that many sentences in natural language text are polypredicative, i.e. express more than one judgement. Such sentences are broken into propositions according to the number of predications. For example, the sentence John has a car is monopredicative and corresponds to the possessive proposition, whereas the sentence John has a car and a summer cottage is polypredicative and can be broken (decomposed) into two possessive propositions: John has a car, John has a summer cottage. Composite sentences that admit of syntactic decomposition are divisible; those sentences that don't admit of such decomposition are indivisible.

Another problem is that syntactic characteristics of sentences often don’t correspond to the semantic type of the proposition, i.e. to the type of judgement expressed by this sentence. For example the sentence John has blue eyes has the same syntactic characteristics as the possessive proposition (see table I), but expresses a qualitative judgement, and its deep structure can be represented by the model X is Yed (John’s eyes are blue). One can say that there is contradiction between this sentence’s A-model representing its deep structure and F-model representing its surface structure.

The communicative plane of discourse comprises manifestations of structural connections between sentences. Structural connections are based on repetition of some information and can be divided into lexical and syntactical. Lexical connections include lexical repetition, pronominal substitution, co-referent terms, hyponymic and hypernymic terms. Syntactic connections include ellipsis and syntactic parallelism. Sentences, united by structural connections and the same topic, make up a super phrasal unit. Thus the communicative plane of discourse can be represented as a number of super phrasal units.

Modal plane of discourse comprises different types of modi. A modus is the part of sentence’s structure expressing a modal relation, the relation between judgement expressed by the utterance and situation it denotes imposed by the speaker. IDA suggests two classifications of modi: (1) according to the type of modal relation and (2) according to the way a modus is presented in discourse. The first classification comprises the modi of believing, supposition, knowing, truth verification, evaluation, necessity etc. The second classification comprises exemplar, overt, and covert modi. Every modus distinguished according to the type of modal relation can have exemplar, overt, and covert forms. An exemplar form is characterized by the use of modal words and expressions in the function of predicate of the principal clause; an overt form is distinguished by the use of modal words as secondary parts of a sentence; a covert form occurs when a modal relation is manifested either in the logical relations between judgements expressed by sentences, or in the verb predicate fulfilling the predicative function. For example, an exemplar form of the modus of believing can be represented by the sentence I think he is decent (modal word think is a predicate of the principal clause, the modus introduces the proposition); its overt form can be represented by the sentence In my opinion he is decent (modal expression In my opinion is a parenthesis); its covert form can be exemplified by the sentence He is thought to be decent (the modal expression is a verb-predicate expressing predicative and modal relations at the same time). Different ways of expressing one and the same modal relation are accounted for by its role in discourse. IDA provides explanation for the reasons of expressing one and the same attitude in different syntactic structures. One of the important laws is that the higher is the syntactic rank of the construction in which an idea is expressed the more likely it is that this idea will de expanded in discourse. For example, the idea of a book being old may be expressed in a subordinate clause or in a separate sentence: (1) Tommy found a book, which was very old. (2) Tommy found a book. The book was very old. It is more probable that the text following the first variant will be about Tommy, whereas the text following the second variant will be about the book. The same is true to the ways of expressing a modal relation. The modus of believing expressed by the principal part of a complex sentence presupposes that in the text there should be a substantiation of speaker's judgment, of the truthfulness of the proposition, and there is no such a presupposition if this modus is expressed by a parenthesis.

Relational plane of discourse includes three types of discourse: description, narration, and reasoning that are correspondingly based on three types of logical relations: synchronic, diachronic, and causative-consecutive. A synchronic relation is a relation between judgements about simultaneous actions or states; a diachronic relation is a relation between judgements about temporal succession of actions or states; a causative consecutive relation is a relation between an inference and its substantiation. Consider the following example:

(1)  (a) The local coherence between sentences is not signaled by connectives.

(b). It is marked by various cohesion devices.

Synchronic relation, expressed by (1), can be identified by the following distinctions. 1) The unity of theme, cf. (1) (a) the local coherence, (1)(b) It (= local coherence). 2) Syntactic parallelism: (1)(a) and (1)(b) have similar syntactic structure: subject + passive voice predicate + prepositional object. 3) Impossibility to introduce adverbs with temporal meaning, cf.: ?First, the local coherence between sentences is not signaled by connectives, and then it is marked by various cohesion devices. In fiction the synchronic relation is often expressed in the description of some setting. Cf.: In the room there is a table. Near the table stands a comfortable armchair and *First in the room there is a table, and then near the table stands a comfortable armchair. Thus relational plane is correlated with communicative plane as a deep level of discourse structure with its surface level; logical relations constituting the relational plane are manifested in a number of linguistic features (lexical and grammatical) in the communicative plane of discourse. On the whole, three types of correlation between the communicative and relational planes of discourse can be distinguished. 1) Non-correspondence, contradiction between the meaning of connective words in the communicative plane and the nature of logical relation between judgements in relational plane. 2) Correspondence between the two planes, which occurs if the meaning of the connective word corresponds to the type of logical relation between judgements. 3) Inexplicability of logical relations between judgements, absence of connective words in communicative aspect.

Consider the following examples:

(2) (a) Abends, wenn es sich mSYMBOL 252 \f "Courier New" \s 13ьde gearbeitet hatte, kam es in kein Bett, sondern muSYMBOL 223 \f "Courier New" \s 13Яte sich neben den Herd in die Asche legen. (b) Und weil es darum immer staubig und schmutzig aussah, nannten sie es Aschenputtel.

The communicative plane of (2)(b) contains the causative–consecutive adverb weil, nevertheless, the judgements expressed by the sentence don’t make up reasoning, cannot be correlated with the major premise of syllogism, cf.: ?Everybody who looks dusty and dirty is called Cinderella. In spite of the use of the causative-consecutive adverb, (2)(b) expresses a diachronic relation denoting temporal succession of states of a person. Actually, causative–consecutive adverb weil in this sentence loses its lexical meaning and is used as a synonym of the adverbs with temporal meaning, cf.: First she looked dusty and dirty, and then she was called Cinderella. Thus, the use of weil contradicts the nature of logical relation between judgements expressed by (2)(b).

IDA differs essentially from the rhetorical structure theory. The RST is based on the assumption of existence of unlimited number of spans of discourse while the IDA differentiates between three types of discourse that can be subdivided into an unlimited variety of subtypes according to the configuration of their elements. Existence of contaminated types of discourse (description - reasoning, description-narration) is also possible. 2) Spans of discourse within the RTS are differentiated according to some semantic criteria whereas types of discourse within IDA are distinguished according to logical relations between utterances manifested by definite linguistic features.


Table. 1. Semantic types of propositions


Proposition 1

A-model

F-model

Proposition 2

Category

1

introductive-existential

There is Х

SP

one member

of

2

conclusive-existential

Х exists

PS


3

identificational

X is Y

SPn

two member


4

taxonomic


5

qualitative

X Yed

SPn

state

6

possesive

X hasY

SPOd


7

existential-locative

There is X in Y

PSAdl


8

locative

X is in Y

SPAdl


9

temporal

X is Тday

SPAdt


10

relational-qualitative

X is Zen with Y

SPOp

three

member


11

comparative

X is Zer than Y

SPnAdc


12

relational-identificational

X is Z of Y

SPnAtt


13

relational-taxonomic


14

processional

X works

SP

one member

of action

15

actional

X makes Y

SPOd

two member



Explanation of symbols: S – subject, P – simple predicate, Pn – compound nominal predicate, Od – direct object, Op-prepositional object, Adl—adverbial modifier of place, Adc – adverbial modifier of comparison, Adt – adverbial modifier of time, Att-attribute.

IDA Methodology


 Integrational discourse analysis comprises four procedures: interpretation, reduction, normalization, and canonization. Interpretation of discourse in its turn comprises three procedures. 1) Interpretation of modal plane, which consists of making explicit modi of utterances. A comprehensive classification of modi used in fiction was given by N.D.Arutiunova (1988). The modi typical of scientific discourse are those of believing, supposition, necessity, knowing, and truth verification (detailed treatment of these modi can be found in (Iatsko, 1996)). As modality can be expressed in different syntactic structures (parentheses, predicates expressed by modal verbs, adverbial modifiers, etc.), the aim of interpretation is to represent it as a modus expressed by the principal clause introducing a proposition expressed by a subordinate (nominal) clause. 2) Interpretation of communicative plane, which consists of making explicit lexical and grammatical manifestations of connections between sentences; replacing co-referent terms (pronouns in the first turn) by the standard names of the objects; replacing lexical and grammatical units with causative-consecutive meaning by the standard connective word therefore and placing sentences (or parts of composite sentences) in the order "antecedent–consequent". 3) Interpretation of semantic plane, which consists of breaking sentences having several predications into monopredicative utterances. Such sentences include: a) compressed structures. Compressed structures are sentences with modal verbs and modal expressions, and nominilized word groups (nominalizations), such as gerunds, substantivized participles, nouns with relative meaning fulfilling the function of subject; b) half-predicative structures, which include composite indivisible sentences and simple sentences with restrictive constructions; c) polypredicative structures which include compound sentences, complex divisible sentences, simple sentences with descriptive constructions.

In the process of interpretation compressed structures with modal verbs are decompressed into complex sentences with modi; nominalized constructions are changed into separate utterances. Cf. the modus of supposition It is possible in (3)(b), nominalized structure verbality in (4)(a), modus of knowing it is considered in (4)(c):

(3) (a) The repetition may be motivated by the production strategy ®(b) It is possible that the repetition is motivated by the production strategy

(4) (a) The verbality of the language of subject headings is the main reason for it to be considered a kind of natural language ® (b) The language of subject headings is verbal. (c) Therefore, it is considered that the language of subject headings is a kind of natural language

One must bear in mind that nominalized structures are expanded into separate utterances only if they are used as either antecedent or consequent in reasoning, i.e. express causative-consecutive logical relation. In other cases expanding nominalized structures into utterances makes no sense causing tautology, cf.:

(5) The boy's coming by on a bike was the initial setting of the first story ® The boy came by on a bike. The boy's coming by on a bike was the initial setting of the first story

Half-predicative structures are substantivized and compressed, indivisible complex sentences being compressed into simple ones. Cf. the restrictive attributive clause in (6)(a) and the attribute expressed by participle in (6)(b):

(6)(a). The story has information thatin normal everyday stories would not be mentioned ® (b). The story has information not mentioned in normal everyday stories.

If it is possible, participle constructions are changed into nominal word-groups.

Polypredicative constructions are broken into utterances (decomposed) in accordance with the number of predications. In the process of decomposition are eliminated conjunctions and connective words, which introduce parts of compound sentences and divisible complex sentences. Connectives with causative-consecutive meaning are not eliminated, but, as it has been already mentioned, are replaced by the standard connective therefore. Such an exception is made because the use of lexical and grammatical units with causative-consecutive meaning may be a manifestation of causal logical relation constituting the basis of reasoning. Cf.:

(7) A new topic is introduced in the usual way, but the next sentence again puts "he" in initial position ®  A new topic is introduced in the usual way. The next sentence again puts "he" in initial position

(8) The repetition, among other conversational and narrative functions, is motivated by the production strategy ® The repetition is motivated by the production strategy. Other conversational and narrative functions are motivated by the production strategy.

(9) Since "her" is a pronoun, the individual is known ® "Her" is a pronoun ® Therefore the individual is known

Syntactic decomposition often meets with difficulties. Consider the following examples.

(10) John and Mary ran away

(11) Worthing is a place in Sussex

(10) is a simple sentence which can be broken into two separate utterances because it has two subjects connected by the copulative conjunction, cf.: John ran away, Mary ran away. The possibility of breaking phrases with copulative conjunctions into separate sentences was noticed by L.Bloomfield (1994) who called such phrases endocentric, differentiating them from exocentric structures, which cannot be broken into separate sentences.

According to IDA, when interpreting the semantic plane of simple extended sentences one must take into account the following points: 1) place of a phrase in the communicative structure of a sentence; 2) connection of the sentence with the context; 3) syntactic rank of a phrase. Decomposition of an endocentric phrase is inexpedient if it is used as a topic of the sentence. Such a phrase denotes objects to which a common distinction is assigned. For example, the endocentric word combination in (10) contains a presupposition of a joint action (ran away together) or of a simultaneous action (ran away at the same time). Syntactic decomposition eliminates the presupposition, thus changing the meaning of context. Syntactic decomposition is possible if a phrase is used as a focus of the sentence. Cf.: Social psychology has paid attention to the cognitive aspects of persuasive communication and to the situational analysis of verbal interaction → Social psychology has paid attention to the cognitive aspects of persuasive communication. Social psychology has paid attention to the situational analysis of verbal interaction. If in the focus there is enumeration of similar characteristics of an object represented by co-referent terms, to break the latter into utterances is inexpedient, moreover, in the process of interpretation a synonym with the general meaning is picked up, while the other synonyms are omitted.

Another limitation is the syntactic rank of a phrase. It is inexpedient to expand the attributive phrase in Sussex in (11) (cf. locative proposition Worthing is in Sussex), since this phrase is subordinated to the predicative noun place. The communicative purpose of the speaker in (11) is to show that Worthing is a place, but not, e.g. a personal name. (11) expresses a taxonomic judgement about the class of object, answering the question What is Worthing? The idea of location in this sentence is of a supplementary character, to express it in a separate utterance is to change the meaning of context. One more example can be the sentence Peter took 5 big yellow apples, in which expanding attributes into sentences causes changing the meaning of text, cf.: Peter took 5 big apples, Peter took 5 yellow apples (=Peter took 10 apples).

Text reduction includes defining obligatory and optional utterances and eliminating optional ones. Relation between obligatory and optional utterances is that one of subordination: obligatory utterances are subordinating, optional – subordinated. Subordinate relations are manifested in communicative, nominative and modal planes of utterances, and are identified according to the following rules. (1c) An utterance, the theme of which is not expanded in the communicative aspect of the text, is optional in relation to the utterance, the theme of which gets such expansion. Themes of obligatory utterances are usually indicated in the text's title. (2c) An utterance, which denies a quality of the object, is optional in relation to the utterance, which assigns a quality to the object. (3c) Utterances, which express single judgements (judgements about concrete single objects), are optional as they play a supplementary role in the structure of discourse. These utterances are usually introduced by special parenthetical constructions, such as for example, in particular. (4c) If sentences or parts of a composite sentence are connected by the constructions indicating equivalence of the expressed judgements (such as that is, in other words), an optional utterance is that one, which expresses the judgement in a less explicit way. The degree of judgement's explicability is determined by the context. (5c) If sentences or parts of sentences are connected by the constructions indicating the relation of addition between judgements (such as apart from that, besides), an optional utterance is that one, which is introduced by the construction. (6c) Utterances, which are given in brackets (embedded constructions), are optional. (7s) An utterance, the referent of which is a part or a kind of the object described in the text is optional in relation to the utterance, the referent of which is the whole object. (8m) Utterances with the modus of supposition are optional as they introduce judgements with low epistemic rank (judgements, the truthfulness of which hasn't been properly verified). (9m) Utterances–quotations, utterances with the modus of truth verification, and utterances with the modus of believing of the author of the cited document are optional, if they don't make up an antithesis of reasoning. (10c) utterances, having lexical and grammatical manifestations of connections with the utterances, which have been defined optional according to the previous distinctions (1-9) are optional.

Text normalization includes elimination of lexical and grammatical distinctions of modal, communicative, and stylistic aspects; elimination of optional verb-adjuncts; identifying deep syntactic structure of obligatory utterances.

Text canonization includes defining logical relations corresponding to a certain type of speech and making up a compositional model. Making up a compositional model of reasoning consists in representing judgements as elements of causal relation (major, minor premises and conclusion of syllogism). In the natural language discourse some elements of syllogism are usually not given. In this case two variants are possible. 1) Explication of the major premise on the basis of the minor premise and conclusion, which are given in the text. 2) Explication of the conclusion on the basis of major and minor premises.

Applied uses of IDA


 The structure of discourse within IDA provides a sound basis for comparative analysis of different texts and can be used in different applied research and development programs, specifically in solving some problems in text summarization.

Bibliography


Grimes, J.E. (1975). The Thread of Discourse. The Hague :Mouton

Mann, W.C., Thompson, S.A. 1988. Rhetorical Structure Theory: Toward a functional theory of text organization. Text, 8 (3). 243-281

Arutiunova N.D., Shyriaev E.N. (1983), Russkoe predlozhenie, Moscow: Russki Iazyk.

Bach, Emmon (1989), Informal lectures on formal semantics, Albany, N.Y.: State

University of New York Press.

Bach, Emmon (1982), Syntactic theory, Washington, D.C.: University Press of America.

Bach, Emmon (1964) An introduction to transformational grammars, New York: Holt,

Rinehart & Winston.

Bach, Emmon, Harms, Robert T, ed. (1970) Universals in linguistic theory, London: Holt,

Rinehart and Winston.

Bessmertnaja N., Wittmers E., (1979), Ubungsbuch zur Textlinguistik, Moscow: Vyssaja Skola.

Chomsky, Noam, (1972), Studies on semantics in generative grammar, The Hague: Mouton.

Dijk, T. van, (1972). Some aspects of text grammars, The Hague: Mouton.

Fillmore, C., (1987). Fillmore's case grammar, Heidelberg: Groos.

Fleischer, W, Michel, G., (1977). Stilistik der deutschen Gegenwartssprache,

Leipzig: Bibliographisches Institut.

Gak V.G. (1969)  K probleme sintaksitcheskoi semantiki. Invariantnye syntaxitcheskie

znatchenia i struktura predlozhenia. Мoscow, p. 77-85.

Iatsko, V., (1998), Rassyzhdenie kak tip nauchnoi rechi, Abakan: Publishing House of the Katanov State University of Khakasia.

Kuhn, Thomas S., (1962). The structure of scientific revolutions, Chicago: Univ. of

Chicago Press.

Ogden, C.K., and Richards, I.A., (1927) The meaning of meaning. New York: Harcourt, Brace and C., Inc.

Padutchieva, E., (1974), O semantike syntaxisa, Moscow: Nauka.

Tesnière, L., (1959), Elements de syntax structurale, Paris: C.Klinckersieck.

Apresian, Y.D. (1986) Integralnoe opisaniye iazyka i tolkovyi slovar Voprosy iazykoznaniya № 2. P.57-70.

Apresian, Y.D. (1992) Lexikograficheskiye portrety Nauchno-tekhnicheskaya informatsia. Series 2. № 3. P. 20-33.

Arutiunova, N., (1988). Tipy iazykovyh znacheniy. Moscow: Nauka.

Baldinger, K. (1980). Semantic theory. Towards a modern semantics. Oxford: Basil Blackwell.

Bally, C. (1965). Liguistique generale et linguistic française. Bern: Franke.

Bloomfield, L. (1994). Language. Delhi: Motilal Banarsidass.

Cook, W. A. (1998) Case Grammar Applied. Arlington: The University of Texas Press.

Daneš, F. (ed). (1974). Papers on functional sentence perspective. Prague: Publishing House of the Chechoslovak academy of sciences.

Dijk, T. Van. (1972). Some aspects of text grammars. The Hague:.Mouton.

Dijk, T. Van. (ed). (1985). Handbook of discourse analysis Vol.1: Disciplines of discourse. London : Academic.

Dressler W. Einfuhrung in die Textlinguistik. Tubingen : Niemeyer, 1973..

Fillmore, C. Fillmore's case grammar. (Groos, Heidelberg, 1987).

Fleischer, W, and Michel, G. (1977). Stilistik der deutschen gegenwartssprache. Leipzig:.Bibliographisches Institut.

Iatsko, V. (1996). Loghiko-lingvisticheskie problemy analiza i referirovania nauchnogo teksta. Abakan: Katanov State University of Hakasia Press.

Iatsko, V. (1998). Rassuzhdenie kak tip nauchnoi rechi. Abakan: Katanov State University of Hakasia Press.

Kuhn, Thomas S. (1962). The structure of scientific revolutions. Chicago: Univ. of Chicago Press.

Mathesius V. (1969). Le  Cercle de Prague. Paris: Ed. du Seuil.

Ogden, C.K., and Richards, I.A. (1927).The meaning of meaning. New York: Harcourt, Brace and C., Inc.

Skorohodko, E. F. (1983). Semanticheskie seti i avtomaticheskaya obrabotka teksta. Kiev, Naukova Dumka.

Tesnière, L. (1959). Elements de syntax structurale. Paris: C.Klinckersieck.

Vendler, Z. (1968). Adjectives and nominalizations. The Hague: Mouton.

Vendler, Z. (1967). Linguistics in philosophy

. Ithaka: Cornell University Press.