Beyond semantics: Prosody can shape and modulate listeners' emotional experience of poetry - A cross-linguistic investigation.

Expressive prosody plays an important role in conveying emotion during oral poetry performance. This study investigated whether emotional prosody can serve as a perceptually robust cue for emotion judgment when semantic comprehension is unavailable. Experiment 1 confirmed systematic acoustic differences between joyful and sad poem recitations: joyful recitations exhibited higher normalized pitch, greater vocal intensity, and faster articulation rate, whereas sad recitations exhibited lower normalized pitch, reduced intensity, slower articulation rate, and reduced harmonicity. Experiment 2 employed a 3 (prosody-semantics alignment: congruent, incongruent, neutral) × 2 (valence: joy vs. sadness) × 2 (listener group: native Mandarin vs. Mandarin-naive Japanese listeners) design. When prosody and semantics were aligned, both listener groups produced the strongest emotion ratings consistent with the poems' intended valence. Critically, Mandarin-naive Japanese listeners, who had no access to Chinese semantic content, showed strong modulation of affective judgments by available prosodic cues when vocal expression conflicted with the poems' textual valence. Native Mandarin listeners integrated semantic and expressive prosodic information, but prosodic manipulations substantially weakened semantic biases and, in some cases, shifted semantically driven judgments. These findings suggest that prosodic configuration is an important component of expressive vocal realization in poetic recitation, shaping listeners' affective responses alongside semantic content. Arousal-related acoustic variation may further contribute to these prosodic effects.