Periodical peer-reviewed academic journal of INION RAS

Potential of the GPT-4 language model for linguistic research (based on the analysis of emotivity)

1) Imber Slavyana Yur’evna 2) Tomskaya Maria Viktorovna

1) Senior Lecturer at the Department Linguistics and Professional Communication in Law, Moscow State Linguistic University, Russia, Moscow, yanaimber@yandex.ru 2) PhD of Philology, Associate Professor, Head of the Gender Laboratory of the Center for Socio-Cognitive Discourse Studies, Associate Professor at the Department of Linguistics and Professional Communication in Law, Moscow State Linguistic University, Russia, Moscow, ma-ria.tomskaya@mail.ru

Abstract

The paper deals with the advantages and disadvantages of neural networks, in par-ticular the GPT-4 language model, as an analysis tool of emotivity in linguistic re-search based on posts on social networks. The authors attempt to determine the abil-ity of GPT-4 to recognize the emotiveness of text and the emotional state of its au-thor. The researchers conclude that the use of language models as the primary meth-od of analysis in linguistic studies should be limited; however, they may be em-ployed as an auxiliary tool, for instance, for the analysis of large text data, for pre-liminary text analysis, and for specific linguistic phenomena with clearly formulated prompts, as well as for verifying their hypotheses.

Keywords

emotivity; emotionality; the language model GPT-4; аrtificial intelligence.

Download text

For citing: Imber S.Yu., Tomskaya M.V. (2025) Potentialof the GPT-4 language model for linguistic research (based on the analysis of emotivity). Human being: Image and essence. Humanitarian aspects. Moscow. INION RAN.Vol. 2 (62). pp. 187-202. DOI: 10.31249/chel/2025.02.11


Order this article

Cash

Безналичный платеж по реквизитам

Печатный

Электронный (PDF по e-mail)