New computer technologies for acquiring knowledge, including knowledge about language, open attractive prospects for solving practical and theoretical linguistic problems. The article examines the possibilities of the language corpus as a tool for researching the semantics of words. It was found that the linguistic corpus provides more reliable material for researching the meaning of words. The analysis of collocations obtained with the help of the corpus proves that words show different combinations, in which changes in meaning can be traced. Options for using the text corpus for research purposes are illustrated by the example of semantic innovations. Neosemantism contributes to the formation of new areas in the cognitive space of the modern Ukrainian language, deepening the verbalization of concepts already known to the Ukrainian language.
The orientation of Ukrainian studies to the use of new computer technologies for obtaining knowledge, a new way of research work marks a qualitatively new stage of its development. In the field of scientific research of linguists, databases about language objects areexisted, anda new subject of communication – a computer – needs a new apparatus for describing language, new procedures for its modeling, analysis and synthesis. At the same time, the language system, the products of its implementation in the language activity of modern Ukrainian society – various oral and written texts – appear in the computer environment in other ways, turn to the researcher with their new facets, which forces him to think about hitherto unimaginable problems, to look for new, other methods of representation of language information, to develop means of communication with a new participant in communication – a computer.
Any linguistic research begins with the selection of material and the arrangement of the file. Qualitative and quantitative parameters of the collected material largely determine the nature ofscientific work.
The creation of texts corpora of different national languages significantly simplified the preparatory, but extremely significant stage for the further development of the research. It is enough for the user to enter a word in the search engine of the corresponding corpus to get a complete index of its word forms in the nearest and extended contexts with appropriate addressing. Creating an empirical research base in this way ensures a minimum expenditure of effort and time for data collection, the volume of which will increase significantly at the same time. On the one hand, it simplified and facilitated, but on the other hand, it complicated the process of analyzing and comprehending so many examples.
The Internet and its quantitative data, which accompany the results of searches for individual words and phrases, testify to a fairly objective picture of the functioning of the language, since in these data and their "quantitative coefficients" of each word, a sample of source material for
2023 Copyright for this paper by its authors. multifaceted generalizations is revealed that deserves attention. A standard operation of searching for a text or a single word in the hands of a philologist can become a useful tool for lexicographic research, which will most likely end up with a lexicographic product – a dictionary article.
In research practice, the main approaches to the study of linguistic data available in corpus linguistics are widely used – corpus-driven (CDA) and corpus-based (CBA). The first of these approaches allows you to put forward scientific hypotheses, in another way, its task is to support mainly deductive research, the second has a pronounced inductive nature and fully corresponds to the directions of experimental and evidential linguistics. The best result, obviously, makes it possible to simultaneously use the advantages of both approaches, for example, to determine the word-forming or semantic potential of lexical items.
The reference to corpus data is associated with the name of J. Sinclair, who formulated the wellknown principle of idiomaticity: "The principle of idiom is that language user has available to him a large number of semi-preconstructed phrases that constitute single choices, even though they might appear to be analyzable into segments" [15]. Corpus material is used to study linguistic phenomena as a source of language study (according to F. de Saussure – langue). Corpus reflect real facts of word usage. Corpus data are used as proof of the existence of a particular phenomenon in the language or to test a particular hypothesis. Ch. Fillmore rightly noted the value of corpus data: "I don't think there can be any corpora , however large, that contain information about all the areas of English lexicon and grammar that I want to explore, ... every corpus I have had the change to examine, however small, has taught me facts I couldn't have imagined finding out in any other way" [18].
New methods of semantic analysis based on corpus materials are presented in the monographs of S. Th. Gries and А. Stefanowich [17], as well as D. Glynn and K. Fischer [20]. British linguists proved that the frequency of use of a word is closely related to the structure of the conceptual structure underlying its meaning [16].
The article uses the method of corpus research and the lexicographic method (analysis of explanatory dictionaries). All word usages are interpreted taking into account the meanings available in the dictionaries. The meaning of neosemantism is recorded in dictionaries, so we offer our own definition. In this way, we obtain a semantic grid covering regular and single interpretations, which can be characterized by chronology. In this way, we obtain the chronological dynamics of the lexical meaning.
The interpretation of language as a code led to the use of statistical methods of text research to
identify certain linguistic regularities in them. According to G. E. Miram, the main principle of
distributive theory is to study the textual behavior of language elements for their further
comprehensive study [
As the object of discussion of the raised problems, we chose neosemanticisms associated with the further aspectualization of certain concepts in the Ukrainian language. The appearance of new meanings in the semantic structure of a word is one of the powerful mechanisms of replenishing its lexicon in the Ukrainian language today. The term neosemantism is broader in terms of content than the generally accepted secondary nomination, semantic, or intraword derivation, since it covers both the formation of new meanings on their own ground in the structure of already existing words, and the appearance of new borrowed words-meanings (primary and secondary), and independent interword derivation with repeated use of the same model of word formation. Therefore, the term neosemanticization can be used to denote all the results of updating the semantics of a word at a certain stage of language development. In general, active semantic changes are stimulated by the need of the Ukrainian language to acquire new fields of hitherto unknown concepts.
Information about the peculiarities of the use of a word, its frequency and typical semantic and grammatical environment opens up for researchers the opportunity to determine its place in the structure of a certain linguistic register and to see a separate linguistic unit in the language system in general.
The goal is to investigate the mechanisms of the appearance of new meanings in words with an already fixed form, using data from the GRAK corpus (General Regionally Annotated Corpus of the Ukrainian Language).
Among the new applications of corpus linguistics that deepen the examination of traditional concepts in the field of lexical semantics, the problem of distinguishing between polysemy and homonymy occupies an important place. At the same time, for linguists who use the corpus on a regular basis, the focus of research gradually shifts, covering an ever-increasing range of questions: from collocation and phraseology to collimation and problems related to grammatical word classes.
The data obtained on the basis of the corpus show that one of the main factors is the word belonging to a certain lexical-grammatical class. It has long been known that the belonging of a word to a specific part of the language reflects the difference in the semantics of signs of different types and demonstrates the difference in the behavior of words in the text. The belonging of a lexical unit to this or that part of the language largely determines the semantics, and accordingly, the peculiarities of its semantic structure.
The problem of the specificity of polysemy of different parts of the language arises in connection with the consideration of certain regularities of the interaction of lexical and grammatical meaning. It has been repeatedly noted that the definition of the lexical meaning of a word already implies an indication of its grammatical characteristics.
On the one hand, the word is a unit of the lexical system, and on the other hand, as a grammatical unit, it has the imprint of both systems, the word is a part of the language, the lexical-grammatical category and those obligatory grammatical meanings that are characteristic of a part of the language.
The experiment consists in clarifying the semantic potential of nouns, adjectives and verbs on the material of the corpus of the Ukrainian language GRAK.
Nouns, adjectives and verbs are carriers of categorical meaning, realize different meanings of the semantic structure. Verbs and nouns, for example, realize the tendency to form a number of independent nominative meanings connected by parallel polysemy, adjectives, on the contrary, are characterized by the "diffusion" of the semantic structure, where the connection between LSV (Lexico-semantic variant) derivatives and the main meaning of the word can be clearly traced. This is characteristic of the radial type of semantic connection, where the radial-metaphorical meaning is important.
Let's pay attention to the semantic potential of adjectives and nouns. One of the indicators that the language is in the field of emotional tension of its speakers are new nuances in the meanings of adjectives that concentrate relevant, socially significant thoughts. The semantic potential of the words нафталін [naftalin] (Naphthalene) / нафталіновий [naftalinovyi] (Naphthalene), тефлоновий [teflonovyi] (Teflon) is indicative.
The GRAK corps provided the following statistics: Нафталін [naftalin] (Naphthalene) Found 796 documents.
According to the Dictionary of the Ukrainian Language in 11 volumes, нафталін [naftalin]
(Naphthalene) is "a white crystalline substance with a pungent odor, used to combat collar mites,
weevils, etc., to protect woolen products and fur from moths, as well as in technology and medicine"
[
In sources from 1929 to the end of the 20th century, we trace the use of the word in its literal sense. However, from the beginning of the XXI century. contexts realize the figurative use and meaning of "which has gone into the past, is irrelevant." Нафталіновий [naftalinovyi] (Naphthalene) Found 264 documents.
We record only a few sentences where the adjective naphthalene has a direct meaning. All other contexts of use have a negative emotional and evaluative shade of the meaning "irrelevant, which has gone into the past".
Contexts of use in Figure 1 naphthalene army, naphthalene hero, naphthalene "Regions` Party ", naphthalene smell, naphthalene team, naphthalene glamor of the film represent a new meaning of the adjective. Правильний [pravylnyi] (Correct) Found 87304 documents.
This adjective illustrates radial-chain polysemy using the example of the semantic structure of the word правильний [pravylnyi]: the meanings "true" (правильний шлях [pravylnyi shliakh] right way), "just" (правильне розв’язання конфлікту [pravylne rozviazannia konfliktu] correct resolution of the conflict), "good, just" (правильний батько [pravylnyi batko] trueparent) have a common component "which corresponds to reality"; the meaning "unerring" (правильний напрям [pravylnyi napriam] correct direction) in turn has a connection with the meaning "which corresponds to the established rules, norms" (правильні сівозміни [pravylni sivozminy] correct crop rotations), and it – with the meaning "regular, rhythmic" (правильне биття серця [pravylne byttia sertsia] correct heartbeat). All these meanings are fixed in dictionaries. Instead, we observe a change in the meaning and popularity of the adjective правильний [pravylnyi] in glossy publications and in certain circles of society. Phrases such as правильна їжа [pravylna yizha] (the right food), правильний ресторан [pravylnyi restoran] (the right restaurant), правильний одяг [pravylnyi odiah] (proper clothing) were not used before. Let's illustrate with more vivid examples of the use of the adjective правильний [pravylnyi] in this new evaluative meaning: Правила правильної жінки: пора зайнятися собою, а не пошуком принців! [Pravyla pravylnoi zhinky: pora zainiatysia soboiu, a ne poshukom pryntsiv!] (The rules of a correct woman: it's time to focus on yourself, and not on the search for princes!) This usage is close in meaning to Galician comilfo– Frenchсomme il faut–it should be, as it should be. And not without the influence of English right meaning "suitable, proper". Through the relevant literature and its not always successful translation, such colloquial, subcultural tracings come, because they are used by a certain stratum of society, which is "planted" on such literature, with such interests and instructions.
With the help of the correct word, publications try to form a new style of behavior: how to behave, what clothes to wear, what to eat, what to read. This whole system of rules is hidden behind the new use of the correct word. Anotherexample: На платформі ZOOM проходило навчання на тему: «Сучасний процес рекрутингу. «Правильні люди на правильні посади» (https://tax.gov.ua 29.04.2021) [Na platformi ZOOM prokhodylo navchannia na temu: «Suchasnyi protses rekrutynhu. «Pravylni liudy na pravylni posady»] (Training was held on the ZOOM platform on the topic: "Modern recruiting process. "The right people for the right positions"), where the right positions are positions in state bodies. In such contexts, there is manipulation of public consciousness.
The adjective conjugation is correct in the GRAK corpus in Figure 2: right diet, correct diagnostics, right date, correct shadows, right way, proper food ration, correct selection, correct oval, correct face, right life, proper nutrition.
The combination of the right diet, the right date, the right way, the proper food ration, and the proper nutrition reflect the formation of a new meaning of "suitable". Маніпуляція [manipuliatsiia] (manipulation) Found 26910 documents.
The noun acquired a new special meaning in political discourse. Ordinary native speakers may not guess the origin of this word: it comes from the Latin manipulus – "handful", which comes from manus "hand" and pleo "fill" and has the direct meaning "complicated action on something" and figuratively "machination". In the Dictionary of Foreign Words, edited by O. S. Melnychuk, this word is recorded with the explanation "1) movements of the hand or both hands to perform a certain task (e.g., in working with a telegraph key). 2) any complex action. 3) figuratively – a fraudulent prank." With changes, it was recorded in the later Modern Dictionary of Foreign Words by O. I. Skopnenko and T. V. Tsymbalyuk: "French. manipulation, manipulus – "handful" 1. The movement of the hand or hands associated with the performance of a certain task, e.g. M. doctor; 2) demonstration of tricks, which is based mainly on the dexterity of the hands, able to divert the attention of the audience from that; what should be hidden from them; 3) transferred Frauds, fraudulent pranks, e.g. M. with securities" [9].
Since manipulation is not a physical influence on a person, its object is psychological structures, then the new metaphorical meaning is “programming the thoughts and aspirations of the masses, their moods and even mental state to ensure the behavior that is needed by those who possess the means of manipulation. It is the art of controlling people's behavior by purposefully influencing their consciousness and instincts". The lexeme manipulation expands the scope of use by increasing the volume of semantics, which is characterized by a higher degree of generalization, for example: маніпуляції в темах [manipuliatsii v temakh] (manipulations in topics), маніпуляції фактами [manipuliatsii faktamy] (manipulations of facts), маніпуляції влади з виборами [manipuliatsii vlady z vyboramy] (manipulations of the authorities with elections), тиск та маніпуляції [tysk ta manipuliatsii] (pressure and manipulations), договорняки та маніпуляції [dohovorianky ta manipuliatsii] (negotiations and manipulations).
Changes in the denotative component of the meaning of the lexeme manipulation consist in the loss of several differential sems and the appearance of the "socio-political" and "psychological" sems. Oksana Zabuzhko aptly defined the result of the development of a new meaning of this lexeme: The manipulator presses the buttons not of our strengths, but of our weaknesses. He takes advantage of our sins. I hate people who exploit weaknesses. They know what they are doing. But those who are manipulated refuse rationality. They are driven by subconscious impulses. This is what is called the crisis of rationalism, which exists in our time (Gazeta po-ukrainsky, 19.09.2019).
Concordance of the noun manipulation in the GRAK corpus in Figure 3 indicates the emergence of the terms "social and political", "psychological": The election manipulation of the 2004 made visible the legislation in force at the time, and I consider it to be quite legislation. This is not a serious approach from the point of view of responsible politics, but it is very convenient for those who will later resort to manipulation. "Perhaps some have rightly decided that Lutsenko, despite all his activity, has not been of any real use for quite some time, while his bright image of a martyr languishing in prison is a more playful an object for campaign manipulation," Zaitsev reflects.
According to the Dictionary of the Ukrainian language in 11 volumes, the lexeme маніпулятор
[manipuliator] (Manipulator) is "1. A person doing various manipulations (in 1 sign.); // A circus
magician skillfully manipulating various objects 2. A device for transmitting telegraphic signals.
3. special The device on the control panel, in the control room, etc. to regulate complex production
processes" [
Cиндром [syndrom] (Syndrome) Found 15929 documents.
In such a generalized deterministic meaning, the verb to articulate becomes a synonym of the verbs to express, to reveal, to define, to outline, to formulate, to form, to demonstrate with the qualifiers exactly, expressively, clearly, and therefore, expands its lexical and syntactic compatibility.
Variants of using the corpus of texts for research purposes will be illustrated on the example of
neosemantism, a синдром [syndrom] fixed with the meaning: "complexof symptoms characteristic of
a certain disease" [
Similar "syndromes" distinguish phenomena undesirable for speakers of the Ukrainian language, which they assess as a threat to their peaceful, stable life.
Conjugation of the noun syndrome in the GRAK corpus
The collocation of the noun syndrome in Figure 4 reflects the new meaning of "complex of features distinctive for a certain phenomenon, object": desacralization syndrome, war syndrome, rejection syndrome, vagrancy syndrome, indifference syndrome, Chornobyl syndrome. Артикулювати [artykuliuvaty] (to articulate) Found 1114 documents.
The verb артикулювати [artykuliuvaty] (to articulate) is recorded in standard dictionaries only in the sense of "pronounce speech sounds." However, the contexts of use in the corpus testify that it is capable of expressing a new, figurative meaning "to clearly define, outline something, express one's attitude towards something." In this sense, there are reasons to consider it a hidden borrowing from modern Western European languages, cf. English articulate – in the sense of "express clearly, explain; formulate". The words артикль [artykl] and the Polish and old Ukrainian артикул [artykul] (article) have a common etymology with articulate (lat. articulus – member, section). So, you can logically explain, formulate, emphasize, clearly outline anything, for example, position, interests, nation, as evidenced by the information from the GRAK corpus.
Conjugation of the verb to articulate in the GRAK corpus in Figure 6: articulate interests, articulate opinion, articulate this, articulate questions, articulate self, articulate to the world, articulate things, articulate problems, articulate position, articulate constancy. All phrases represent a new, figurative, meaning "to clearly define, outline something, express one's attitude towards something." meanings, and the completeness of the concordance is determined by the exhaustive nature of the description of the corresponding corpus. Because of this, there is a fundamental need for grammatical (morphological) information in the concordance, which helps to characterize and distinguish forms, and the need for semantic analysis (description or interpretation of meanings) in the dictionary.
The object and unit of dictionary description is the meaning and the dictionary article, and for the concordance it is the word usage and the corpus. Its completeness requires that it be a register-type dictionary.
In addition, the dictionary approach seeks to highlight some essence and interpret it. This is a normative explanatory dictionary, which contains a large amount of semantic information, semantic and stylistic markings, and the concordance reproduces the completeness of the thesaurus type (a dictionary that maximally but fully presents the words of the language with examples of their use in the text), it is the first derivative of the corpus, from list of words (dictionary), it is distinguished by the presence of morphological characteristics of word usage.
It is worth noting that all modern dictionaries are based on corpora, so there is also a feedback
relationship – lexicographic and grammatical knowledge is reflected in the corpus in the process of
creating a system of linguistic annotation of the corpus, but the dictionary and concordance are
derivatives of the text – a "chopped test" [
The concordance makes it possible to analyze large arrays of text to identify patterns in the use of words or phrases in the language. Based on the obtained results of the concordance, it is possible to understand the meaning of the word from the context and carry out an analysis of its usage. Let's analyze the concordance of the word form on zero and the adjective zero.
На нулі [na nuli] (At zero) Found 694 documents.
Of these, 50 documents cover the period up to 2022, all others for 2022, and most of them have the meaning "located closest to the front line, to the enemy". In the Dictionary of the Ukrainian Language in 11 volumes, the ambiguous word zero has the meaning: "1. The number 0, which means no value, and when it is substituted for another number on the right, it means increasing it tenfold. Нуль уваги [nul uvahy] (Zero attention) – no attention.
2. A conditional value from which the deduction of values similar to it (time, temperature, pressure, etc.) begins.
3. In the pre-revolutionary school, the lowest score for the evaluation of knowledge and behavior.
4. transferred That which has no value and meaning» [
Concordance of the noun At zero (frontline) in Figure 7 testify to the formation of a new meaning of the noun at the zero (frontline) of the "leading position":
Guys are at zero (frontline), we need two Motorolas.
My friend, the major, was at zero (frontline) somewhere for 3 months, now he has moved to Lviv and is actually teaching paratroopers how to fight here.
For 102 brigades from Ivano-Frankivsk, which are "at zero"(frontline), it was possible to collect 80 thousand hryvnias quickly.
Many of them voluntarily joined the Armed Forces, spent months "at zero"(frontline) and during vacation have the opportunity to meet their families who are abroad.
Critical thinking is at zero (frontline), as are causal relationships. "After six months at zero (frontline), it's hard to find something I'm not ready for," says the singer.
Let all troubles, bad mood and evil intentions of enemies be at zero (frontline) in the daily reporting.
The fact that everything is calm in Kyiv region now, there is no sense of war in the city, people live as usual in everyday life – go to the beach, shops, cinemas – this is all thanks to the guys who are at zero (frontline) now in hellish conditions and are taking the fight.
Thanks to our volunteers once again, who organize special courses, including online, where you can learn and master the big control of drones, learn some life hacks, how to use them at zero (frontline).
Discussion
In the process of studying the development of polysemy, one cannot limit oneself to only the symbolic aspect of a word. In the process of studying the semantics of a word, it is impossible to be limited only to its symbolic and onomasiological aspect, that is, to the analysis of the logical and subject content of the word.
It is necessary to investigate how the experience fixed in the meanings of words is the basis of the words generated by speech and how a new reality enters the language as a fact of its system.
The relationship between language and speech is so important that one necessarily implies the other. Language is the source of all the subjective that penetrates the objective language system, it is the channel through which human practice is realized in speech. In this sense, speech is primary to language.
That is why corpus linguistics primarily reflects speech and is of great importance for the identification of lexical units.
The development of the meaning of a word includes the following indicators: frequency in the original corpus of texts; the number of lexical meanings, the sequence of meanings in the normative dictionary. And this is significant; the most important values are first recorded in the dictionary article, and then all the others in order of decreasing usage. The synchronous vector of a word makes it possible to obtain the general distributional and statistical characteristics of a word in a certain language field and to find trends in the development of its meanings.
The mechanism of determining the meaning of a word consists precisely in assigning a word to a certain paradigmatic series, to a conceptual field. This is evidenced by the fact that in many cases the lexeme is diagnostic, not connected to the analyzed word by a syntactic connection, but capable of signaling one of several possible paradigmatic series for this word.
For example, in the sentence Війську потрібна свіжа кров, або Ротацію тим, хто «на нулі» (https://www.pravda.com.ua/columns/2022/12/20/7381565/) [Viisku potribna svizha krov, abo Rotatsiiu tym, khto «na nuli»] (The army needs fresh blood orRotation to those who are "at zero") military vocabulary, although not syntactically related to the word на нулі [na nuli] (at zero), indicates that the last word should be understood not as "number", but as "to be on the front lines."
The specificity of the keyword has a great influence on the diagnosis of the meaning of the word. Not every word connected with another can indicate the meaning of the latter. Words with very wide connecting possibilities (lexemes with extremely developed polysemy and words with unclear meaning) have no diagnostic power at all. The appearance of a new meaning is determined by the context of the word's functioning and its internal semantic possibilities, because the core of meaning a set of signs of something – remains in all phrases. The social context provides objects with such sets of features.
As Sidney Landau points out, a common problem when working with a citation index is the lack
of examples of specific vocabulary; a common problem when working with the corpus is an excess of
examples ("hits") [
The results of using corpora of language bases in lexical semantics demonstrate that the significance of this approach is not limited to corpus resources, that is, the material used by researchers. Today, corpus linguistics is acquiring more and more signs of a new linguistic paradigm. Despite the fact that the data of corpus linguistics mainly clarify the existing descriptions of linguistic phenomena, without radically changing them, the novelty and importance of this approach in lexical semantics should be seen in the fact that an empirical research base is formed and new opportunities for the analysis of paradigmatic and syntagmatic qualities of linguistic units.
Obtaining a modern model of the Ukrainian language system involves not only the analysis and description of the form and semantics of lexical units, their functioning in texts, but also the creation of reliable tools for monitoring the course of its dynamics. Іnternet and language corpora are the main source of information about a language in our days.
As we can see, an important feature of Internet data as a linguistic source is the possibility of direct access to contexts in which the use of the desired word is recorded. This is important for the transition from quantitative to qualitative characteristics, from an abstract invariant represented in the quantitative coefficient of a word to its semantic, stylistic, and other variants.
URL:http://irbis-nbuv.gov.ua/ulib/item/UKR0000989. [8] V. Starko. The Ukrainian Semantic Lexicon. 2022. URL: https://github.com/brownuk/dict_uk/tree/master/data/sem. [9] O. І. Skopnenko, T. V. Tsymbaliuk. Modern dictionary off oreign words, Dovira, Kyiv, 2006, 709 р. [10] J. Zhan, H. Zhao, Span Model for Open Information Extraction on Accurate Corpus.
Proceedings of the AAAI Conference on Artificial Intelligence, 34(05), 2020, P. 9523-9530. [11] H. Hoherchak, N. Darchuk, S. Kryvyi, Representation, analysis, and extraction of knowledge from unstructured natural language text, Cybernetics and Systems Analysis, 2021, Volume 57, № 3., P. 164–183. https://doi.org/10.1007/s10559-021-00373-7 [12] H. Hoherchak, Knowledge Based and Description Logics Applications to Natural Language Texts Analysis, Proceedings of the 12th International Scientific and Practical Conference of Programming (UkrPROG 2020), 2021, Volume 2866, P. 259–269. [13] D. Rothman, Transformers for Natural Language Processing (2nd addition), publishing Packt, 2021, pp. 384. [14] W. Che, Y. Lui, Y.Wang, B. Zheng, T. Liu, Towards Better UD Parsing: Deep Contextualized Word Embeddings, Ensemble, and Treebank Concatenation. – Proceedings of the {CoNLL} 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. – Association for Computational Linguistics, 2019, P. 55–64. [15] J. Sinclair. Lexical Grammar, in: Nujoji Metodologija. 24 (2000). pp. 191–203. [16] D. Divjak. Mapping Between Domains. The Aspect-Modality Interaction in Russian, in : Russian
Linguistics. 33 (3), 2009. pp. 249–269. [17] S. Th. Gries. Corpus-based methods and cognitive semantics: the many meanings of to run, in : S. Th. Gries, A. Stefanowich (Ed.), Corpora and cognitive linguistics: corpus-based approaches to syntax and lexis, Berlin, 2006. рр. 57–99. [18] Ch. Filmore. Corpus linguistics or computer-aided armchair linguistics, in: Directions in Corpus
Linguistics: Proceedings of Nobel Symposium 82. 1991. рр. 35–38. [19] M.-W. Devlin, K. Chang Lee, K. Toutanova, BERT: Pre-training of Deep Bidirectional
Transformers for Language Understanding, NAACL-HLT, 2019, P. 4171–4186. [20] D. Glynn. Corpus-driven Cognitive Semantics Introduction to the field, : Dylan Glynn, Kerstin Fischer (Ed.), Quantitative Methods in Cognitive Semantics: Corpus-Driven Approaches, 2010, рр. 1– 42. doi: 10.13140/RG.2.1.2325.1365 [21] Ye. Karpilovska. Introduction to Computational Linguistics, TOV Yuho-Vostok, Ltd, Donetsk, 2003, 264 р.