earticle

영어

This study examines the phenomenon of backtracking in English sentence processing from syntactic, semantic, and cognitive perspectives. It explores how adverbial clause type, tense combination, information structure, and collocational relations affect the degree of backtracking, with special attention to adverbial clauses such as ‘after’, ‘before’, ‘when’, ‘while’, and the ‘not A until B’ construction. Experimental analysis demonstrates that structural complexity, information focus, and (non-)collocational relationships between adverbial clause transitive verbs and main clause subjects are key factors determining backtracking strength and frequency. Longer clause structures and repeated tenses intensify the need for information reanalysis, thereby increasing backtracking effects. L2 learners often fail to backtrack in temporarily ambiguous sentences, leading to lower comprehension than L1 speakers. L2 learners need focused instruction and practice regarding backtracking. By using generative AI tools like ChatGPT 4.1, this study opens new possibilities for research in language processing and pedagogy.