Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates conversational entrainment in code-switched speech across three language pairs: Mandarin-English, Hindi-English, and Spanish-English, revealing that while lexical entrainment is consistent, acoustic-prosodic and stylistic entrainment varies by context. The researchers also assess how well classification models, including classical and Transformer-based classifiers, capture these human behaviors, finding that these models often prioritize less relevant features compared to those that humans use. This work establishes a human-centered framework for evaluating model performance in multilingual contexts and highlights the challenges in creating conversational agents that can effectively engage in code-switched dialogue.
Classification models miss key features of human conversational entrainment, raising questions about their effectiveness in multilingual dialogue systems.
Conversational entrainment is well-studied in monolingual and written contexts, but remains underexplored in spoken code-switching (CSW). We present a novel cross-lingual analysis of entrainment in Mandarin-English, Hindi-English, and Spanish-English dialogue and show that, while lexical entrainment generalizes across language pairs, entrainment over acoustic-prosodic and CSW style aspects exhibits context-specific variation. We build on these findings by asking whether classification models capture these human behavioral patterns. Applying feature importance and ablation analyses, we find that classical and Transformer-based classifiers detect entrainment reasonably well but consistently prioritize features other than those most salient to human entraining behavior. Our approach introduces a human-grounded framework for evaluating model decision-making in multilingual stylistic contexts, and suggests future challenges for developing conversational agents capable of producing naturalistic code-switched speech.