The article argues that antonymous words are not isolated in the language but are part of the general scheme of thematic classification of the vocabulary, entering into close ties with representatives of synonymous and polysemantic groups. Proof of this position is the availability of antonymous-synonymous blocks and paradigms of meanings of antonymous words-polysemants in the language. This fact serves to understand the antonymy of lexical as a relationship between members of opposing dichotomous series, as well as expands the boundaries of closed binary structures, which are traditionally used to make antonymous connections. Antonymous relations are established to connect two opposites in pairs, as well as to combine several units into one polynomial structure. Words are found to have more antonyms than the corresponding number of meanings. This proves that some semes of one word are opposed by two or more synonymous antonyms. From the above-mentioned, it follows that the antonyms of polysemous words can be stated only within one family. Such a framework can be defined by limiting the range of antonymous units to a third common concept, under which two antonyms are summed up; establishing the antonyms of polysemous words through the contextual indicative minimum; distinguishing between the direct and figurative meaning of the word. The antonymous connections within the framework of the synonymous-antonymous paradigm are established to demonstrate a qualitatively heterogeneous character, differ in the strength of the manifestation of the relations of opposites and the frequency of practical implementation. The analysis of the manifestations of antonyms between the members of the dichotomous synonymous series allows distinguishing the following types: radial, linear, and cross radial. It is proved that the quantitative indicator of the formation of certain antonymous pairs from among the members of synonymous series depends on the volume of the synonymous series being compared, from the nature of semantic differences of the words of each series, from the emotional, evaluative, and stylistic characteristics of the members of the series, and the distribution of the words with relatively opposite meanings.
Unconscious semantic processing of polysemous words is not automatic