Predictive Processes Shape the Effects of Age and Parkinson’s on the Relation between Syncopation and the Urge to Move to Music

The pleasurable urge to move to music (PLUMM) shows an inverted U-shaped relation with syncopation. A predictive processing account hypothesizes that moderately syncopated rhythms maximize precision weighted prediction errors (PE) and thus elicit the strongest urge to minimize PEs via movement. We tested this hypothesis by combining computational modelling with a large sample of PLUMM ratings. We also investigated whether metrical model strength (MMS) drives the effects of age, musical training, and Parkinson’s disease (PD) on PLUMM. PLUMM ratings were collected from 73 PD patients and 515 healthy controls (HC) with large ranges in age (17-80) and years of musical training (0-46). As an index of precision weighted PE, surprisal values were calculated from a Bayesian model that inferred which of two underlying templates (metric, uniform) was active as the rhythms unfolded. PE minimization was modeled as the difference in surprisal between rhythms with and without a metronome, which served as a proxy for synchronized movement. Two parameters capturing distinct aspects of MMS were fit to the per-participant ratings. Rating results showed that both PD and age flattened the inverted U while musical training led to the opposite effect in HCs. Simulations showed that moderately syncopated rhythms maximize reducible PEs. Parameter fitting results showed that MMS accounted for the effects of age and PD but not musical training. These results suggest that PLUMM is driven by the affordance to reduce PEs and that age and PD weaken the representation of meter leading to an altered PLUMM response