ARTIFICIAL INTELLIGENCE: A NEW ONTOLOGY OF E.F. TARASOV'S THEORY OF SPEECH COMMUNICATION

Tarasov A.E.

Moscow City Pedagogical University
Moscow, Russia
e-mail: alexgarmisch@yandex.ru

Introduction

Any theoretical or practical research in psychology faces the challenge of direct—bypassing speech—contact with the living brain. It is understood that "translating" thoughts into words in natural language inevitably leads to partial or complete "distortion." Therefore, psychologists strive to understand how to isolate a "thought" so that "the spoken thought does not become a lie."

The task is not limited to the search for "truth." Modern civilization is a complex, multi-level system of social and industrial interactions that requires a well-thought-out and predictable organization of joint activities.

Activity itself, as a form of mental activity, began to be considered in the 1930s in response to the social demands of mass production. While up to 100,000 people interacted simultaneously in major battles during World War I, the industrial revolution of the interwar period required the organization of joint activities—taking into account supply chains and the division of labor—by over a million people.

Against the backdrop of the expansion of mass production, the problem of organizing mass labor without the use of a military management system began to emerge. Such a system required the training of a significant number of people, a non-trivial task even under a totalitarian state. Therefore, the idea of creating theoretical constructs that would provide, on the one hand, a simple logical framework for describing both individual and collective (joint) activity, and on the other, would form the psychological basis of a subject-object activity paradigm, seemed promising [22]. Such a theory was developed by S.L. Rubinstein [14] and subsequently developed into a coherent model by A.N. Leontiev [9].

Although activity theory provided the prerequisites for a broad theoretical understanding of collaborative activity and the management of mass behavior, as a psychological theory, it did not allow for an examination of the practical mechanisms of organizing mass production and lacked the tools to analyze interactions beyond the subject-object system. In other words, activity theory adequately described the interaction between the agent (the human subject) and the object of activity (nature-object), although even profane experience shows that object-based activity can take a transformed form.

An attempt to solve this problem was undertaken by a group of linguists led by A.A. Leontiev [12], who proposed a theory of speech activity. This theory postulated that the organization of joint activity can be considered a special, transformed form of objective activity, in which one of the communicants acts as the object of the activity.

This theory effectively linked subject and speech activity, making it possible to solve numerous practical problems, such as combating various forms of illegal activity [19]. The theory of speech activity provided a heuristic previously unattainable in any scientific discipline. It became possible to evaluate the course of thought processes in an individual based on the results of speech activity. In effect, researchers developed a speech research framework that, on the one hand, did not require interpretation as psychoanalysis, while, on the other hand, the accuracy of statistical research methods—associative experiments and semantic differentials—significantly increased compared to research conducted in the early 20th century [25].

Based on the tenets of speech activity theory, numerous theoretical and methodological approaches were developed that improved the effectiveness of many processes related to speech, from school education to public behavior management. It can be confidently stated that the 1970s and 1980s were the era of the triumphant advance of speech activity theory.

However, with the accumulation of empirical material, certain limitations of the theory of speech activity began to emerge.

Firstly, the theory of speech activity inherited the problem of activity motivation from activity theory. Being convinced Marxists, the authors of activity theory did not address what they considered an obvious point: "For what purpose, in fact, does objective activity begin and take place?" Motivation for objective activity was essentially reduced to the base of Maslow's pyramid [11]. Meanwhile, the question of self-actualization was reduced to the construction of a communist society.^[1].

Secondly, speech activity theory has grappled with the problem of text as an object of study. In the early stages of speech activity theory's development, text was examined within the framework of the Shannon-Weaver model. Communication was viewed as a process of "information transfer," with the primary focus not on the content of the information, but on its distortions due to imperfect communication lines [5]. The Shannon-Weaver model adequately describes communication using templated messages that require decoding. Moreover, a simplex (one-way) communication system is used by default.

Naturally, the ontological problem of text as such arose. The problem of the hermeneutics of sacred text has been known since ancient times: over time, the connection between the bodies of speech signs and their meaning was lost, requiring the creation of a new text that would ensure understanding for a new generation of communicants. The next generation of communicants needed a new text to understand the original text and the changes introduced by previous interpreters—a procedure known as the "hermeneutic circle.".

Empirical experience shows that during the process of communication—organizing joint activities—communicants reduce the volume of text presented, taking into account their shared ethnic (cultural) experience. This reduction in text is especially noticeable when translating into a foreign language—certain concepts and images common in one culture must be recreated for speakers of another ethnic culture [15].

Moreover, some of the text relies on certain unconscious constructs inherent to any ethnic group and perceived as given—archetypes (cultural stereotypes) [2]. Archetype theory has attracted the interest of linguists in the realization of cultural aspects of communication and, as a result, has generated an urgent need for theoretical understanding of the interaction between the communicant and the surrounding reality in the process of life.

Third, one of the problems with speech activity theory concerns the nature of interaction between communicators. In the Shannon-Weaver model, the line of communication is broken at the moment of "message decoding," and the communicators' reactions to the information received fall outside the model's scope. Moreover, this model lacks an analysis of the entire interaction process.

In the activity theory, the subject-object activity paradigm is implemented: the subject (person) has consciousness and reflection, and the object (nature) does not have the ability to influence the subject in any way and, in general, a feedback scheme is sufficient to describe the subject’s reflection on the object.

In the case of communication, a problem arises: who is the subject and who is the object of activity? The principle of feedback proves insufficient for constructing a model of activity. In speech activity theory, this problem is solved by introducing the principle of dialogicity into the model [3]. In other words, communicants regularly change their roles in speech activity, alternately occupying the position of subject and object.

Adding the principle of dialogism to the theory of speech activity allows us to construct a theoretical construct that takes into account the entire «line of communication» – from one consciousness to another.

From the theory of speech activity to the theory of speech communication

The theory of speech communication emerged from an attempt by a group of linguists led by E.F. Tarasov to resolve the problem of the reductionist approach of activity theory and speech activity theory. Specifically, theoretical propositions concerning the form and content of speech communication were developed.

It should be noted that the development of a theory of verbal communication largely reflected the need to overcome the crisis of the language model [4]. By this point, a significant volume of empirical materials had accumulated, the theoretical understanding of which within the framework of the language model no longer provided the necessary heuristics, and the emerging new ontology—for example, hypertext or a text corpus—required a fundamentally new approach.

The incorporation of a new ontology into theoretical frameworks facilitated the consideration of speech activity in general and verbal communication in particular as a complex object that cannot be described within the framework or using the scientific approaches of only one discipline. Therefore, it requires the use of triangulation [26]. In fact, the triangulation method within the theory of verbal communication has been significantly expanded compared to the original methods in sociology.

In the theory of verbal communication, the two main methodological foundations utilize the representation of both subject and speech activity as transformed forms possessing a certain isomorphism [17]. Moreover, the totality of transformed forms includes the cultural component of linguistic consciousness as an integral element of verbal communication. Psycholinguistics, which is essentially a cross-cutting discipline, has received the opportunity to directly integrate methods, approaches, and empirical research material from other scientific fields into the research process. In particular, the ontological contradiction between verbal text, which is considered in linguistics, and non-verbal text, which is fragmentarily considered in a wide range of scientific disciplines, from psychology and psychiatry to cultural studies and jurisprudence, has been overcome.

Within the framework of speech communication theory, a theoretical construct was proposed that linked the visual and verbal components of text included in speech communication [18]. The problem of incorporating visual images into the analysis of perception and presentation of texts in the process of organizing joint activities was solved.

The heuristic nature of creolized texts allowed for a theoretical understanding of how the "picture of the world" is reflected in verbal communication. A conceptual framework was developed that integrates the cultural (ethnic) component into the complex semantic structure of interaction when organizing joint activities. This approach allowed for the resolution of certain problems in the theory of translation and interethnic (international, intercultural) communication [16].

It should be noted that the development of intercultural communication theory is the result of the further development of the concept of linguistic consciousness as a transformed form of the worldview, which derives from the theoretical principles of activity theory. In fact, intercultural communication is viewed in speech communication theory as a special, "defective" form of communication. The "defectiveness" of intercultural communication is associated with the influence of differences in the mental models of representatives of different ethnic cultures [23].

The theory of verbal communication is open to the analysis of new ontologies emerging both at the level of the formation of the consciousness of a specific ethnic group and at the level of various phases of globalization processes. In particular, the rapidly developing communicative environment of the Internet has facilitated the identification and description of new linguistic phenomena that have demonstrated an unprecedented speed and depth of change [16]. In other words, within the framework of the theory of verbal communication, a set of scientific tools has been developed that allows for the long-term assessment of basic sets of consciousness, such as a value system, which is considered one of the fundamental elements of an ethnic group in its historical development.

The author in the text and the text in the author

Speech activity theory, like speech communication theory, considers the author of a text "after the fact": if a text exists, generated within the context of subject-matter or speech activity, it always has an author. The author may remain unknown, but within the framework of speech activity theory, a text cannot exist without an author.

Text is generated by communicants within the context of speech activity and, according to activity theory, serves as a transformed form of objective activity. Text, as a product of speech activity, reflects the specific characteristics of the author and contains a significant amount of information about the author. When engaging in verbal communication, the author, in addition to the text itself, presents a whole set of additional mental images, thereby ensuring the effectiveness of the communication process. For example, the author, especially if he or she is visibly participating in the communication, inevitably conveys information about their demographic characteristics—gender, age, social status, etc.

When the communication process is disrupted in time—for example, between a writer and a reader—or in space—for example, when working remotely—measures are taken to generate additional text used to convey the "image of the author." This may include various commentaries, critical analysis of the work, the author's biography, and various elements of the "textual apparatus.".

In the late 1960s, an "authorless" approach to analyzing literary works emerged in European philosophy and literary studies. In their works, representatives of this academic school, known as structuralism, argued that when analyzing a text, attention should be paid to the structure of the text, not the author or the plot of the work [24]. In other words, structuralists examined linguistic form without reference to a specific communicative situation. Essentially, this was the Shannon-Weaver communication model, from which the signal itself was removed, leaving only the communication line and the influence of interference.

The presence of a text's author is the cornerstone of speech activity theory. Within this theory, the generated text is the product of the activity of a subject interacting with an object, the specific characteristics of which prevent the subject from solving the tasks facing it without the assistance of others. To organize joint activity, the subject engages in verbal communication with the goal of modifying the behavior of the other subject in solving the problem.

The other participant in verbal communication—the communicant—seeks to determine their own motivations in interacting with the other communicant. This involves not only explicitly presented mental images but also attempts to determine the other communicant's implicit motivations. The need to determine the "hidden agenda" is particularly common when the goals pursued by the communicants (both declared and implicit) are inconsistent or, often, conflicting.

The identification and analysis of hidden motivations has a long history and is a goal of many scientific fields. Almost always, the subject of analysis is the author's identity, which can be identified in the text [8]. However, the development of computer technology has made it possible to develop techniques and procedures for generating text in which the authorial component is blurred to the point of being virtually impossible to identify and analyze.

Artificial Intelligence and Text Generation

In the 21st century, the development of computer technology has led to the emergence of so-called "artificial intelligence" (AI). Although the terms intelligence and artificial intelligence have a long history of use in both scientific and lay discourse, there is no established understanding of the term, and researchers involved in the development and application of AI use adhoc definitions [13].

AI is a technology that has been promoted for decades as a breakthrough technological project that will provide the country developing it with a competitive advantage in the creation of complex industrial or combat systems. However, due to the nature of research and development funding, AI is more of a marketing ploy, employing science fiction imagery to metaphorically describe computer technologies that utilize various algorithms to solve applied problems with great efficiency. A characteristic of computer technologies, like any other virtual (imaginary) technologies, is the widespread use of metaphors, driven by the inclusion of a significant number of profane individuals among the end users of computer systems. This leads to the creation of a sophisticated mythology associated with the need to integrate the AI terminology into relevant discourses.

AI has begun to be actively used to create various human-computer interfaces. Such interfaces effectively replace humans in performing routine, repetitive tasks, such as in information systems. Despite the high-sounding name, the AI function in such systems is limited to recognizing and interpreting templated user requests and providing templated responses. Furthermore, "speech synthesis" technologies have been developed, creating the illusion of "conversing with a robot" (virtual assistant).

Further advances in AI have led to the development of so-called "generative artificial intelligence" (GAI). GAI is a type of AI capable of generating text with varying degrees of creolization in response to a user's request. At the same time, characteristics such as coherence or consistency approach the level of text generated by a native speaker of natural language. The text generated by GAI goes beyond the templated dialogue of a human-machine system. When generating text, GAI is capable of adapting to the user's response and, up to a certain point, following the logic of the narrative or the nature of the human's delivery.

The GII text generation algorithm differs from the human text generation algorithm—in essence, the GII is a sophisticated database management program. For example, the most well-known program today, ChatGPT, is an internet information search program. Its algorithm uses texts found online, generated by native speakers of natural language, to generate quasi-texts that, at first glance, appear to be the product of human speech. Furthermore, the program is capable of machine learning, which involves adapting the algorithm to the task at hand and taking into account previous experience solving similar problems [1].

It should be noted that the aforementioned AI marketing campaign, and in particular, the GII, has fallen on fertile ground in social and political mythology—the promotion of GII programs relies on images of consciousness well-developed in works of science fiction, where authors describe human-machine interaction [6] as speech activity using natural language. Given that the emergence of such programs was expected and predictable, the promotion of GII has followed the path of realizing this "dream." Furthermore, the source code of most GII programs is closed, and from the perspective of the lay (end) user, the program is simply a "black box" whose operation can only be simulated.

The GII has the functionality to enable the creation of^[2] Texts that can be incorporated into verbal communication, for example, in mass communication systems. In particular, "narrative technologies" that emerged at the turn of the century have demonstrated significant effectiveness in shaping value orientations and managing mass behavior in various fields, such as politics and education [7]. The use of narrative technologies is seriously limited by the need to produce a significant volume of texts. The task of expanding coverage is usually achieved by repeatedly broadcasting the same text across various channels (social networks and mass media). However, such duplication of information flows leads to a significant reduction in the effectiveness of information impact. Moreover, such use of narrative technologies can be considered information aggression, which in many cases will inevitably provoke a response from the state [10].

The widespread introduction of texts created by the GII into various discourses requires theoretical understanding: what challenges are associated with the onset of the "information revolution," and what is the place of new technologies in the life of modern society? Specifically, the inability of people to recognize GII texts without specialized tools or training undermines the very foundation of verbal communication in society: the existing notion of the "meaningfulness" of a text and the author's verbal intent no longer reflects the surrounding reality, which is filled with images of uncertain truth or falsity.

Literature

1. Ayvazyan, S.A., Enyukov, I.S., Meshalkin, L.D. Applied statistics: fundamentals of modeling and primary data processing. Moscow: Finance and Statistics, 1983.
2. Averintsev, S.S. Archetypes // Myths of the peoples of the world: Encycl. in 2 volumes / ch. ed. S. A. Tokarev. — 2nd ed. - M.: Soviet Encyclopedia, 1987. - T. 1: A-K. — 671 p.
3. Bakhtin, M.M. Problems of Dostoevsky's poetics. - M.: Prosveshchenie, 1972. - 346 p.
4. Vdovichenko, A.V. Development of a communicative model of the verbal process in the context of a crisis of the language model: project, meaning formation, communication // Questions of Psycholinguistics, 2 (32), 2017. Pp. 50–61.
5. Dorofeeva, I.V. The Shannon-Weaver Model and Its Significance for the Development of Communication Theory // Language Discourse in Social Practice. Collection of Scientific Papers of the International Scientific and Practical Conference. — Tver: Tver State University, 2013. — P. 49–53.
6. Clark, A. 2001: A Space Odyssey. — Moscow: Eksmo, 2018. — 320 p.
7. Comuzzi, L.V. The Place and Role of Didactic Narrative in Multicultural Education // Language and Cultural Interaction in the Arctic Region: Proceedings of the International Conference. — Murmansk: Moscow State University for the Humanities, 2014. — P. 218–226.
8. Leontyev, A.A., Shakhnarovich, A.M., Batov, V.I. Speech in forensic science and forensic psychology. Moscow: Nauka, 1977, 62 p.
9. Leontiev, A.N. Activity, consciousness, personality. - M.: Smysl, Publishing Center "Academy", 2004. - 352 p.
10. Manoilo, A.V., Petrenko, A.I., Strigunov, K.S., Neshkov, S.V., Sinitsyna, T.Yu., Zaitsev, A.Ya. Information and Hybrid Warfare. A Textbook for Universities. Moscow: Goryachaya Liniya-Telecom, 2024. 594 p.
11. Maslow, A. Motivation and Personality. 3rd ed. / Translated from English. — St. Petersburg: Piter, 2019. — 400 p. — (Series "Masters of Psychology").
12. Fundamentals of the Theory of Speech Activity / Ed. Dr. of Philological Sciences A.A. Leontiev; USSR Academy of Sciences. Institute of Linguistics. Moscow: Nauka, 1974. 367 p.; Rubinstein, S. L. Fundamentals of Psychology. Moscow, 1935.
13. Russell, S., Norvig, P. Artificial Intelligence: A Contemporary Approach. 4th ed. — Hoboken: Pearson, 2021.
14. Rubinstein, S.L. Fundamentals of Psychology / S.L. Rubinstein. - M.: [b. i.], 1935.
15. Sorokin Yu.A., Markovina I.Yu., Kryukov A.N. and others.Ethnopsycholinguistics / resp. ed. Sorokin Yu. A. - M.: Nauka, 1988. - 192 p.
16. Tarasov, E.F. Towards the construction of a theory of intercultural communication // Language consciousness. Formation and functioning. - M.: Institute of Linguistics of the Russian Academy of Sciences, 1998. - P. 30-35.
17. Tarasov, E.F. Problems of the Theory of Speech Communication: Scientific Report on Published Works. Moscow: Institute of Linguistics, Russian Academy of Sciences, 1992, 56 p.
18. Tarasov, E.F., Nistratov, A.A., Matveev, M.O. Creolized Text: Semantic Perception. Collective Monograph. Moscow: Institute of Linguistics, Russian Academy of Sciences, 2020, 206 p.
19. Federal Law of July 25, 2002 N 114-FZ "On Counteracting Extremist Activity" URL: https://base.garant.ru/12127578/
20. Foucault, М. What is an Author? // In Harari, J. V. (Ed.), Textual Strategies: Perspectives in Post-Structuralist Criticism. Ithaca, NY: Cornell University Press, 1979.
21. Hamas, A. What is an Author? // Philosophical Journal, 83 (11), Eighty-third Annual Meeting of the American Philosophical Association, Eastern Division: pp. 685-691.
22. Shchedrovitsky, G. P. Initial concepts and categorical means of activity theory // Shchedrovitsky G. P. Selected works. - M.: School of cultural policy, 1995. - P. 233.
23. Linguistic Consciousness and the Image of the World. Collection of Articles / Ed. N. V. Ufimtseva. — Moscow, 2000. — 320 p.
24. Barthes, R. The Death of the Author. Fontana Press, 1997.
25. Rosanoff, AJ. Free Association Test (Kent-Rosanoff). Wiley & Sons, Incorporated, 1927.
26. Rothbauer, P. Triangulation // In Given, L. (Ed.), The SAGE Encyclopedia of Qualitative Research Methods. Sage Publications, 2008. pp. 892–894.

References

1. Aivazyan, SA, Enyukov, IS, Meshalkin, LD. (1983). Applied statistics: Basics of modeling and initial data processing. Moscow: Finance and Statistics.
2. Averintsev, SS. (1987). Archetypes. In A. Tokarev (Ed.), Myths of the peoples of the world: Encyclopedia in 2 volumes (2nd ed., Vol. 1, p. 671). Moscow: Soviet Encyclopedia.
3. Bakhtin, MM. (1972). Problems of Dostoevsky's poetics. Moscow: Prosveshchenie, 346 p.
4. Vdovichenko, AV. (2017). Development of a communicative model of the verbal process in the context of the crisis of the language model: Project, meaning-making, communication. Questions of Psycholinguistics, 2(32), 50–61.
5. Dorofeeva, IV. (2013). The Shannon-Weaver model and its significance for the development of communication theory. In Linguistic discourse in social practice: Proceedings of the International Scientific and Practical Conference (pp. 49–53). Tver: Tver State University.
6. Clarke, A. (2018). 2001: A Space Odyssey. Moscow: Eksmo, 320 p.
7. Komuzzi, LV. (2014). The place and role of didactic narrative in multicultural education. In Linguistic and cultural interaction in the Arctic region: Proceedings of the International Conference (pp. 218–226). Murmansk: MGGU.
8. Leontiev, A.A., Shakhnarovich, A.M., & Batov, VI. (1977). Speech in forensic science and legal psychology. Moscow: Nauka, 62 p.
9. Leontiev, AN. (2004). Activity, consciousness, personality. Moscow: Smysl, Academy Publishing Center, 352 p.
10. Manoilo, A.V., Petrenko, A.I., Strigunov, K.S., Neshkov, S.V., Sinitsyna, T.Yu., & Zaitsev, A.Ya. (2024). Information and hybrid wars: A textbook for universities. Moscow: Hot Line – Telecom, 594 p.
11. Maslow, A. (2019). Motivation and personality (3rd ed., Trans. from English). Petersburg: Piter, 400 p.
12. Fundamentals of the theory of speech activity. (1974). In A. A. Leontiev (Ed.), Academy of Sciences of the USSR, Institute of Linguistics (p. 367). Moscow: Nauka.
13. Russell, S., & Norvig, P. (2021). Artificial intelligence: A modern approach (4th ed.). Hoboken: Pearson.
14. Rubinstein, S.L. (1935). Fundamentals of psychology. Moscow: [no publisher].
15. Sorokin, Yu.A., Markovina, I.Yu., Kryukov, AN, et al. (1988). Ethnopsycholinguistics (A. Sorokin, Ed.). Moscow: Nauka, 192 p.
16. Tarasov, EF. (1998). Towards the construction of the theory of intercultural communication. In Linguistic consciousness: Formation and functioning (pp. 30–35). Moscow: Institute of Linguistics, RAS.
17. Tarasov, EF. (1992). Problems of the theory of speech communication: A scientific report based on published works. Moscow: Institute of Linguistics, RAS, 56 p.
18. Tarasov, EF, Nistratov, AA, & Matveev, MO. (2020). Creolized text: Semantic perception. A collective monograph. Moscow: Institute of Linguistics, RAS, 206 p.
19. Federal Law No. 114-FZ of July 25, 2002, On counteracting extremist activities. URL: https://base.garant.ru/12127578/
20. Foucault, M. (1979). What is an author? In V. Harari (Ed.), Textual strategies: Perspectives in post-structuralist criticism (pp. 685–691). Ithaca, NY: Cornell University Press.
21. Hamas, A. What is an author? Philosophical Journal, 83(11), 685–691.
22. Shchedrovitsky, GP. (1995). Basic concepts and categorical means of activity theory. In P. Shchedrovitsky: Selected works (p. 233). Moscow: School of Cultural Policy.
23. Linguistic consciousness and the image of the world. (2000). In N. V. Ufimtseva (Ed.), Collection of articles (p. 320).
24. Barthes, R. (1997). The death of the author. Fontana Press.
25. Rosanoff, AJ. (1927). Free association test (Kent-Rosanoff). Wiley & Sons.
26. Rothbauer, P. (2008). Triangulation. In L. Given (Ed.), The SAGE Encyclopedia of Qualitative Research Methods (pp. 892–894). Sage Publications.

[1] For obvious reasons, the authors could not, and perhaps did not set themselves the task of expanding the theory of activity to take into account the realities of the 1960s.

[2] Unlike humans, the GII does not generate text, but compiles it from existing texts using an algorithm that does not follow the algorithm for generating human speech.