A Functional Trade-off between Prosodic and Semantic Cues in Conveying Sarcasm
Abstract
This study investigates the functional trade-off between prosodic and semantic cues in conveying sarcasm. We analyze how speakers modulate their vocal expressions when semantic content alone is insufficient to signal sarcastic intent. Our findings reveal that prosodic features become significantly more pronounced when semantic markers of sarcasm are absent or ambiguous, suggesting a compensatory mechanism that maintains communicative effectiveness. This trade-off has important implications for multimodal sarcasm detection systems and speech synthesis applications.
Type
Publication
In Proceedings of Interspeech 2024
This paper explores the functional trade-off between prosodic and semantic cues in expressing sarcasm. We demonstrate that when semantic content alone doesn’t clearly signal sarcastic intent, speakers compensate by using more pronounced prosodic features. This finding has significant implications for speech technology, particularly for sarcasm detection systems and natural-sounding speech synthesis.
