Any text is a coherent stretch of speech which is a semantico-topical and syntactic unity. Sentences make up textual stretches on syntactic lines according to a communicative purpose in a particular communicative situation. As a result, a textual stretch has a unifying topic. So, in syntactic terms a text is a strictly topical stretch of talk centering on a common informative purpose.

    In 60 -70s of the 20^th century appeared a branch of linguistic research which consider that the sentence is not the highest level of language hierarchy as it was thought earlier. It was noticed that most sentences are relatively independent semantically and structurally: the full completeness is acquired by the sentence only in the element of higher level – in the text.

    The text began to be considered by many linguists as the highest lingual unit having its own categories and elements, and also rules of organization, i.e. having its specific grammar. In the grammar of the text it is said that the text has powerful integral potential and, therefore, structurally and communicatively it is the unity, the features of which can not be summed up to the features of constituent sentences.

    To the basic elements we can refer the blocks inside the text, being the chains of sentences, characterized by relative structural and semantic completeness and called either complex syntactic unity or supraphrasal unity or supra-sentential construction. Since sentences in these unities are joined by means of syntactic cumulation it stands to reason to call such sentence sequences cumulemes. The cumuleme is essentially a constituent part of one-direction sequence of sentences forming monologue speech. Besides one-direction sequences, i.e. cumulemes, two-direction sequences should be recognized that essentially build up constituent parts of dialogue speech. The component constructions – utterances in these sequences are positioned to meet one another, hence their name “occursemes” (of the Latin root meaning “to meet”).

    The new approach to the nature of the text has been proposed by the introduction of the notion of dicteme – the elementary topical textual unit. The dicteme occupies the highest position in the hierarchy of segmental levels of language. It can be expressed either by a cumuleme (a sequence of two or more sentences), or by one single sentence placed in topically significant position. The dicteme, as an elementary topical textual unit, is polyfunctional. In the text it performs the functions of nomination, predication, topicalization, and stylization.

The features of the text

In the framework of the given understanding of text, it has two main differential features topical (semantic) unity and semantico-syntactic cohesion. Under topical unity we understand impossibility to extract from the text any of it fragmant or in other case the text is destroyed or is reorganized into another text. Any way of extraction can lead to visible change of the text. Under cohesion we understand logical semantico-syntactic and formal structure of the text, leading to the reorganizing its components in any order or including into it some “foreign” fragments.

    Textual categories appear and function only in the text as a language unit of the highest rank. Textual categories reveal the cardinal and the most general differential features of the text.

    Today the list of textual categories is open: linguists name different textual categories because they approach the text from different angles. To the list of textual categories scholars usually refer cohesion, informativeness, retrospection, modality, causality, implication, the author’s image, and some others.