Interdisciplinary digital ways of researching political media discourse

Bulgarova Bella Akhmedovna ORCID: 0000-0001-6005-2505 PhD in Philology Associate professor of the Department of Mass Communications, RUDN University named after Patrice Lumumba 117198, Russia, Moscow, Miklukho-Maklaya str.6. bulgarova-ba@rudn.ru Other publications by this author     Chinennaya Tamara Yurievna ORCID: 0000-0002-2621-6606 PhD in Philology Associate Professor; Department of National and Federal Relations; Russian Presidential Academy of National Economy and Public Administration 82 Vernadsky Ave., p. 1, Moscow, 119571, Russia. t.chinennaya@mail.ru Other publications by this author     Zyukina Zulfira Salihovna ORCID: 0000-0002-1583-9576 PhD in Pedagogy Associate Professor; Department of Russian Language and Linguoculturology ; RUDN named after Patrice Lumumba 6 Miklukho-Maklaya str., Moscow, 117198, Russia zyukina_zs@pfur.ru

Kozlovskaya Ekaterina Stanislavovna ORCID: 0000-0002-6308-725X PhD in Philology Associate Professor; Department of Russian Language No. 5 ; RUDN named after Patrice Lumumba 6 Miklukho-Maklaya str., Moscow, 117198, Russia kozlovskaya_es@pfur.ru

Introduction. In politics, language can be used not only as a tool, but also as an object of research. Political media discourse is an interdisciplinary subject located at the intersection of linguistic sciences, mass media (hereinafter referred to as mass media), politics and modern digital technologies ^[2].

Digital methods and tools for the study of political discourse help scientists identify and analyze the features of the semantic organization of a political text (including quotations, references, allusions, reminiscences) between its different types, ways of representing and constituting certain political processes or their interpretation, etc.

Literature review. According to the research results of S. Blassnik (Blassnig 2022), quantitative content analysis (manual, automated or semi-automated) continues to be the dominant method of studying political media discourse ^[10]. At the same time, qualitative and critical content analyses are also relevant today, which is confirmed by the work of R. Vodak, M. Khosravinik (Wodak, KhosraviNik 2013) ^[19]. Moreover, T.N. Lobanova (2020) emphasizes that almost all the "media platforms and Internet portals she studies allow for automatic content analysis by keywords" ^{[5, p. 30]}. M.M. Abbasi (2022) describes systems developed for analyzing text and emotions ^{[1, p. 4]}, which contribute to the disclosure of author's intentions in the analysis of discourse in automatic mode ^[1]. The collective monograph, written under the scientific editorship of I. P. Kuzheleva-Sagan (2017), focuses on the relevance of analyzing everyday discursive practices of network platforms, emphasizes that such a method "helps to trace how the new cognitive style of the social media world transforms ethnic virtual communication" ^{[8, p. 21]}, which proves the relevance interdisciplinary digital ways of media discourse research.

Methods. The following methods were used in the article: theoretical analysis of scientific literature on the identified issues, descriptive and comparative methods, content analysis, interpretation of the results obtained. The main research method in the article was a comprehensive comparative analysis.

The course of the study. Digital methods of media discourse analysis expand the boundaries of the study of corpus linguistics and other related disciplines. If earlier empirical studies more often highlighted the laws of building and distributing a single communication channel or platform, for example, political speech, press releases, party manifestos, talk shows, advertising, websites or social networks, then the growth of software and artificial intelligence tools allows you to analyze the discourse of large volumes of texts in a short time from different content.

The interdisciplinary aspect of the study of political media discourse is the attempt of scientists to study this issue in the context and mutual correlation of such sciences as:

1. Linguistics – statistical data on the number of definitions related to politics help this science determine the degree of kinship of languages, the presence of outdated words and expressions or neologisms in a political text, as well as differentiate the appropriate terminology, which allows you to create text corpora in order to supplement and improve dictionaries, glossaries, thesauruses, GOST, etc., for what programs such as Sketch Engine, TermoStat are used for ^{[6, p. 73]}, etc.

2. Psycholinguistics – digital programs focused on the study of linguistic behavior and the study of the principles of interpretation and verification of the conceptual components of political discourse, contribute to the disclosure of the "linguistic personality" of politicians and mass sentiments. For example, the results of psycholinguistic content analysis of media discourse help to shape public consciousness through the addressee's relevant choice of speech methods "to construct a text in accordance with the expectations of target audiences (addressees)" ^{[7, p. 2]} and, as a result, to fight for power. In particular, quantitative and qualitative analyses of the frequency of replication of keywords contribute to the disclosure of the implicit nature of the analyzed discourse through "identification of patterns in its structure and ... assessment of the frequency distribution of linguistic units" ^{[4, p. 223]}.

3. Sociolinguistics – the study of political media discourse allows us to determine how language changes between social groups, depending on age, during population migration and between communities ^[12] under the influence of a certain political situation in the country and the world, as well as to conduct a conversational analysis of social and political interactions through the prism of verbal behavior of political leaders ^{[4, p. 226]}. Political discourse research can be performed using software such as: Nvivo Leading Qualitative Data Analysis Software (QDAS); ATLAS.ti ^[4] – The No.1 Software for Qualitative Data Analysis; Quirkos – Qualitative Data Analysis Software made simple, etc.

4. Philosophy and history – the central themes for researchers are the development and changes in the language of politicians ^[12], indicating the political realities of a certain time (symbolization and value paradigms), the positioning and self-presentation of political leaders and their attitudes, the formation of a certain worldview and worldview of the supreme leadership, which are reflected in the works of philosophers, historians and historiographers and They are quoted by modern politicians in the mass media. Research on the above-mentioned issues is carried out through software such as Nvivo Leading Qualitative Data Analysis Software (QDAS); ATLAS.ti ^[4], T-LAB analytics, etc.

5. Pedagogy – the study of political media discourse helps psychologists, educators, linguists and publishers studying the process of reading and the formation of cognitive activity of future political scientists to differentiate and analyze the language and its phenomena in which students read and write, as well as to determine which dictionaries are relevant to them ^[12] (explanatory, information retrieval, dictionaries of quotations politicians, bilingual dictionaries of political terminology, etc.). Software such as, for example, Sketch Engine helps to study both materials on political topics created for students and schoolchildren, and essays written by the students of the training course themselves, presented in the media space. The analysis of discourse in academic circles and for educational purposes of political scientists is carried out in order to determine the attitudes, power relations and prospects of participants in political debates; the reasons for the growth or decrease in the number of migrants or refugee migration; how these conflicts are presented on news sites; how such content is distributed through blogs, etc. ^[12].

6. Digital linguistics – the study of the Internet corpus of user languages allows you to determine the public mood prevailing in social networks, messengers and other virtual platforms, as well as to determine the prevailing media content on political topics. For example, using the software (hereinafter referred to as the software) Provalis Research Text Analytics Software, the main tools of which are QDA Miner, WordStat, it is possible to analyze numerical and textual data, manage text content, analyze coding, surveys, image, moods and opinions of the audience, as well as analyze media framing ^[13].

Table 1 – Methods and tools for the study of political media discourse^[1]

Methods such as quantitative, qualitative and mixed content analysis, based on Table 1, are extremely popular in the digital space - to analyze the media discourse, developers include algorithms in the software that work like Google information search engines.

Content analysis is often conducted in a conglomeration with expert surveys or interviews with political figures, which allows you to study the motives or strategies of their self-presentation communication. In addition, this type of research is combined with data from panel or cross-sectional surveys, thereby helping to analyze the impact of communication, for example, on the formation of public opinions or attitudes.

Recently, scientists have been applying content analysis to data from "digital footprints", for example, in order to study the influence of specific communication content or elements of the style of posts on social networks on user reaction in the form of popularity signals.

The researchers' impression of such a method of studying political media discourse as content analysis made it advisable to compare the results of the works of scientists who turned to different tools: E.I. Beglova, O.Y. Shmeleva (2015), who used manual sampling and qualitative content analysis, the object of research of which was the speech of a political leader at speeches for 2012-2015; and Ya.-M. Eberl, P. Tolochko, P. Jost, T. Heidenreich, H.G. Bumgaarden (Eberl, Tolochko, Jost, Heidenreich, Boomgaarden 2020), who performed automated content analysis of Facebook posts, including the analysis of "digital footprints".

The object of the study by E.I. Beglova and O.Y. Shmeleva (2015) was the speech of V.V. Putin at nine speeches, as a result of which about one hundred and forty linguistic units were analyzed. In addition, the researchers were tasked with the following tasks: to identify those means of language that ensure interaction between the state apparatus and Russian citizens; specific methods of information presentation; ways of effective communication between authorities and citizens; to establish the presence or absence of a correlation between the president's speech and official documents. As a result of the research conducted by E.I. Beglova and O.Y. Shmeleva (2015), the following conclusions were made: the head of state evaluates the activities of other democratic states favorably; the president pays great attention to the importance of traditional values; the political leader of Russia uses "contrasts" and "opposition" as the predominant speech strategies ^[3].

J.-M. Eberl, P. Tolochko, P. Jost, T. Heidenreich, H.G. Bumgaarden (Eberl, Tolochko, Jost, Heidenreich, Boomgaarden 2020) analyzed the "digital footprints" of social network users and voter survey data using Microsoft Graph API software as part of a multilevel approach of negative binomial regression. The object of the study were six political parties that participated in the Legislative elections in Austria in 2017: the Greens, the Peter Pilz List, NEOS, the Social Democrats (SPÖ), the rebranded New People's Party (ÖVP) and the right-wing Freedom Party (FPÖ). The authors analyzed about 51,000 messages from users of social networks and about 315 accounts in order to identify in which context people mention the names of political parties, with which topics they are associated ^[11].

The advantages and disadvantages of each of the methods (manual and automated) of the two above-mentioned studies should be differentiated according to the following criteria: verifiability of the results obtained; objectivity; number of research objects. In the process of analyzing the work, the following results were obtained:

- the conclusions of E.I. Beglova and O.Y. Shmeleva (2015) ^[3] demonstrate the nature of subjectivity and low verifiability on the following grounds. Firstly, if you give the initial data to another researcher, the results may be completely opposite to what linguists received, since they are subjective in nature. In addition, the analysis includes only one object. In addition, the authors' conclusions do not fully correspond to the tasks set, since, for example, language tools have not been identified that ensure effective interaction between citizens of the Russian Federation and the state apparatus. The reason for this is that the work does not include feedback from the public. In other words, there is no fundamentally important determinant that would allow us to draw long-term conclusions.;

- J.-M. Eberl, P. Tolochko, P. Jost, T. Heidenreich, H.G. Bumgaarden (Eberl, Tolochko, Jost, Heidenreich, Boomgaarden 2020) ^[11] analyze a larger data set (51,000 social media posts) and apply mathematical models (reflected in the formulas used), resulting in An empirical study can be repeated by another scientist and get similar results. In addition, the authors draw conclusions based on the analysis of six objects (six political parties), while E.I. Beglova and O.Y. Shmeleva (2015) conducted a content analysis of the discourse of one object (the political leader of Russia, V.V. Putin) - due to the fact that the manual sampling method does not allow analyzing a large number As an automated method, it provides less verifiability of the results obtained, unlike the digital method of data processing.

Conclusions. Methods and tools for the study of political media discourse are being integrated into digital technologies. Among the popular software used for the analysis of the linguistic corpus, the following should be highlighted: Sketch Engine, TermoStat, Nvivo Leading Qualitative Data Analysis Software (QDAS); ATLAS.ti, T-LAB, Provalis Research Text Analytics Software.

Methods and tools for the study of political media discourse include: quantitative, qualitative and mixed methods of content analysis; linguistic, statistical, mathematical and graphical tools. Quantitative content analysis is differentiated into manual, automated or semi-automated.

New digital methods of content analysis are more objective, structured and verifiable, which is why the manual method of content analysis is no longer actively used. It is being replaced by new digital technologies and tools with integrated artificial intelligence, which is already actively used not only in methods of analyzing and systematizing political media content, but also in its generation, both textual and visual.

According to the concept of modern search engines and software, digital processing of Internet text is performed mainly by keywords - definitions that make up the semantic core of an utterance or article.