Perplexity vs YesChat vs ChatGPT vs Human Literary and Linguistic Text Analysis on the Example of Conan Doyle's ‘White Company’

This study investigates the capabilities of three large language models - Perplexity, YesChat Text Analyzer, and ChatGPT - in performing literary and linguistic text analysis on Arthur Conan Doyle's “The White Company.” The research theoretical part highlights the intersection of artificial intelligence and linguistics, computational linguistics and education. The practical part of the study, by analyzing the first chapter of the novel, aims to identify which model demonstrates the most effective quantitative and qualitative analysis of stylistic devices. Each model was prompted to identify lexical, grammatical, and phonetic stylistic devices, with results compared to a human analysis based on I.R. Galperin's taxonomy. Findings reveal that Perplexity excels in providing a literary interpretation, while YesChat offers a more structured linguistic analysis, and ChatGPT presents a concise overview. The results also indicate human analysis considerably surpasses the performance of all three studied language models in terms of stylistic analysis quantity and quality.