نقش چندوجهی نغمه‌پردازی (پروسودی) در گفتمان فارسی

This study presents a systematic investigation into the role of Persian intonational features in discourse organization, with a specific focus on prosodic cues within simulated conversational speech. Employing acoustic analysis, the research examines fundamental frequency (F0) variations to elucidate their functions in discourse segmentation and the distribution of pitch accents (e.g., H, L, L+H*) within information structure. Furthermore, it explores the pragmatic determinants influencing pitch accent selection, particularly in relation to utterance predictability and the maintenance of discourse coherence. Methodologically, the investigation draws upon a corpus of elicited speech designed to encompass a range of diverse communicative contexts. Acoustic measurements, including pitch range, duration, and pause length, were extracted and analyzed using PRAAT software (Version 6.0(2020). Intonational patterns were annotated following the auto segmental-metrical framework of intonational phonology, with modifications for Persian-specific phenomena. Key findings demonstrate that Persian speakers employ systematic pitch range expansion as a prosodic marker for topic transitions, with F0 elevation observed at discourse boundaries. Furthermore, the distribution of pitch accents was found to be governed by information status, with new/unpredictable information preferentially receiving high-prominence accents (L+H*). The study contributes to the broader field of prosody-discourse interface research by (1) empirically validating Persian-specific intonational strategies and (2) illustrating how acoustic cues enhance discourse coherence in a typologically distinct language. Implications for intonational phonology models and speech technology applications are also discussed.