AI and psycholinguistics educational app
Based on "Cognitive Mechanisms of Contrast Recognition and Interpretation and Their Functional Analogues in Artificial Intelligence Systems"
The article develops a seven-stage psycholinguistic model for explicit and implicit contrast interpretation, then compares human readers with ChatGPT, Gemini, and Claude under identical experimental conditions.
Article authors
The article is authored by scholars affiliated with Ukrainian universities and departments of linguistics, applied linguistics, journalism, and social-humanitarian sciences.
Liudmyla V. Shytyk
Ukrainian Linguistics and Applied Linguistics
Doctor of Science, Professor, and Head of the Ukrainian Linguistics and Applied Linguistics Department at Bohdan Khmelnytsky National University of Cherkasy.
Liudmyla P. Yuldasheva
Ukrainian Linguistics and Applied Linguistics
Doctoral Researcher at the Department of Ukrainian Linguistics and Applied Linguistics at Bohdan Khmelnytsky National University of Cherkasy.
Anzhelika K. Dosenko
Journalism
Associate Professor and Head of the Journalism Department at V. I. Vernadsky Taurida National University in Kyiv, Ukraine.
Oleksii V. Sytnyk
Institute of Journalism
Affiliated with the Educational and Scientific Institute of Journalism at Taras Shevchenko National University of Kyiv, Ukraine.
Myroslava M. Malysh
Journalism
Affiliated with the Journalism Department at V. I. Vernadsky Taurida National University in Kyiv, Ukraine.
Artem O. Halych
Social and Humanitarian Sciences
Doctor of Philological Sciences, Professor, and Acting Dean of the Faculty of Social and Humanitarian Sciences at Luhansk Taras Shevchenko National University.
Main distinction
Human recognition results
Each bar shows the number of participants out of 40 who correctly identified contrast in the representative examples discussed in the article.
Key article insights
These points summarize major findings reported in the article.
Validated framework
The model connects human psycholinguistic processing with functional computational analogues in LLMs, while preserving the distinction between functional similarity and cognitive equivalence.
Article logic
Explicit contrast is interpreted successfully because overt lexical or syntactic markers constrain recognition and interpretation.
Implicit contrast requires contextual information, conceptual knowledge, presuppositions, and inferential reasoning because the opposition is not directly encoded.
Human and LLM outputs can be compared as functional correspondences, without presenting LLMs as human-like cognitive agents.
Interactive case game
Read the poetic fragment, file a detective report, then unlock the article verdict: article authors' answer, human participant result, and ChatGPT/Gemini/Claude comparison.
Unlocked verdict
7-stage interpretation game
Move through the seven stages of interpreting a poetic fragment and compare how a human reader and an AI model construct meaning.
Stage 1
Final result
Interactive semantic graph
Build the hidden meaning map of a poetic fragment and discover the contrast between concepts.