Do We Feel What We Think We Feel? Mental Imagery and Interoception in Music-Induced Emotion

Introduction: Emotional responses to music vary substantially across listeners and are not solely determined by acoustic features. Music-induced emotions were found to emerge from interactions between the stimulus, the individual, and internal processes such as memory, imagery, and bodily states. While both cognitive and physiological dimensions of such emotional experiences have been investigated, the degree to which they correspond remains underexplored. This study explores the role of mental imagery in shaping interoceptive experiences during music listening and the extent to which subjective reports align with physiological responses. Methods: Two classical music excerpts were used for emotion induction: Johann Sebastian Bach’s Brandenburg Concerto No. 2 (positive) and Edward Elgar’s Cello Concerto in E minor, Op. 85 (negative), each presented for 3 minutes and repeated once. Twenty participants were randomly assigned to positive (n = 9) and negative (n = 11) conditions. Skin conductance responses were recorded using Mindfield eSense. Self-reported emotional valence and arousal were assessed using the Self-Assessment Manikin. Follow-up interviews were conducted to explore participants' emotional experiences during listening, and data were analysed thematically. Results: The positive group predominantly reported personal memories, often involving social or recreational experiences, whereas the negative group reported more non-personal or less self-referential content (e.g., historical or imagined scenarios). The personal imagery group reported higher self-reported arousal than the non-personal group (M = 1.31 vs. M = −0.25; Welch's t = 2.09, df = 5.81, p = .083). No difference in skin conductance was found between imagery types. 8 of 16 participants in the personal imagery group showed discordance between self-reported and physiological arousal, compared with 0 of 4 in the non-personal imagery group. Within the personal imagery group, negative and positive music conditions did not differ significantly in self-reported arousal or skin conductance. Discussion: These findings highlight the importance of accounting for both subjective self-reports and objective physiological measurements when designing and interpreting emotion research. The results also suggest that mental imagery may play a role in how individuals experience and interpret their bodily signals, consistent with constructionist accounts that emphasise the role of past experience and contextual meaning-making in shaping emotion. unknown unknown