Affording Danceability: First- and Third-person Accounts of Wanting to Move to Participant-Selected Music

Danceability is often treated as a fixed musical property, yet listeners experience it as emerging from the interaction between musical structure and embodied, affective engagement. This study examined danceability as a relational affordance by combining first-person reports with computational analysis. Ninety participants selected one song they enjoyed dancing to and one they enjoyed but did not feel like dancing to, and provided open-ended explanations for both choices. Thematic analysis showed that dance songs were described as rhythmically driven, percussive, arousing, and positively valenced, whereas non-dance songs were portrayed as calm, contemplative, emotionally nuanced, and suited to attentive listening rather than movement. Acoustic and higher-level features were extracted for all tracks, and a mixed-effects model was fitted to predict Essentia’s danceability index. Most features differed between playlists, and engagement, arousal, low-frequency spectral flux, and tempo positively predicted modeled danceability. Together, qualitative and computational findings indicate two modes of embodied musical experience: an action-oriented, movement-focused mode associated with dance songs, and a more reflective, being-moved mode associated with non-dance songs. This suggests that danceability reflects dynamic organism–environment coupling rather than a purely intrinsic musical attribute.