Linguistic analysis of structural and textual ambiguity in english
Ambiguity is widely recognized as one of the fundamental challenges in natural language processing. It significantly affects both human language comprehension and computational language interpretation. Among its various forms, structural and textual ambiguity present particular difficulties for linguistic analysis and artificial intelligence systems. This study investigates the linguistic characteristics of structural and textual ambiguity in English and examines how contemporary language models process and interpret ambiguous linguistic structures.The research is based on a corpus of selected a three authentic English sentences containing structural and textual ambiguity. An experimental study was conducted with 45 students at Azerbaijan University of Languages. In the first phase, structurally ambiguous sentences were analyzed using the PRAAT computer program to examine their linguistic and prosodic features. In the second phase, students were instructed to generate texts in ChatGPT using the selected structurally ambiguous sentences as prompts. The resulting texts were subsequently analyzed and compared with outputs generated by ChatGPT, Gemini and Claude to investigate how these artificial intelligence systems interpret, contextualize and resolve structural ambiguity at the textual level.The findings provide insights into the mechanisms employed by large language models in processing ambiguous language and demonstrate similarities and differences in their contextual interpretation of structurally ambiguous constructions. The study contributes to the theoretical understanding of structural and textual ambiguity highlighting their communicative functions in linguistics, natural language processing and artificial intelligence.The relevance of the study lies in its interdisciplinary approach, integrating theoretical linguistics, experimental research, and AI-based language analysis. The results contribute to the growing body of research on ambiguity resolution and offer practical implications for the development of more accurate and context-sensitive natural language processing systems.