THE ANALYSIS OF ‘OF COURSE’ USING GPT
THE ANALYSIS OF ‘OF COURSE’ USING GPT

Author(s): Péter Furkó, Ágoston Tóth
Subject(s): Translation Studies
Published by: Scientia Kiadó
Keywords: of course; persze; GPT; English–Hungarian contrastive data; AI-based and human expert analyses;
Summary/Abstract: Our paper explores the functional diversity of the English pragmatic marker ‘of course’ and compares it to its Hungarian near-equivalent ‘persze’ by contrasting AI-generated and human expert analyses. The findings highlight the limitations of relying on AI output and underline the importance of human supervision. With regard to the English–Hungarian contrastive data, we point out that pragmatic nuances are likely to make the successful application of the sense of enthusiastic agreement, whereas the English ‘of course’ does not have this component, but the latter can function effectively as a device of politeness. Differences like this are challenging for non-native speakers in the absence of targeted language-dependent contrastive information. Our demonstration GPT run failed to provide this specific contrastive information even when asked todo so, but it pinpointed other differences that were useful. As a result, the paper demonstrates the potential of using advancements in neural language modelling in the area of pragmatic research and language instruction, and it argues that a combination of AI-based and human expert analyses has the potential of discovering the intricacies of human communication precisely and effectively.

• Details
• Contents