Two studies examined the influence of mental imagery type and instruments’ tone-decay times on tempo fluctuations between repeated rehearsals and performances of musical works. The first experimental study tested the predictions that 1) motor imagery – in contrast to non-motor imagery – would lead to smaller discrepancies between imagined and performed tempos; and 2) musical works having high note density would tend to be slowed down when imagined, whereas works having low note density would tend to be sped up when imagined. The second correlational study tested the prediction that the same work would exhibit significant changes of average tempo when performed on instruments having different tone-decay times.
In the first study, musicians performed slow and fast musical excerpts either vocally or on their major instrument. These excerpts were recorded. Then, each participant repeatedly attempted to mentally replicate their excerpts’ tempos by using motor or non-motor imagery. Excerpt beginnings were signaled by three-second prompts from the recordings of each excerpt, and participants indicated excerpt endings by ringing a call bell. Excerpt duration discrepancies were calculated by subtracting the performed excerpt length from the imagined excerpt length. The results did not yield support for either of our hypotheses. There was no significant difference of tempo discrepancies when using motor versus non-motor imagery; and there was no significant main effect of variable note density on tempo discrepancies. Post hoc analyses suggested that successive mental rehearsals might lead to smaller and less variable tempo discrepancies across musicians of diverse skill levels. Other post hoc analyses suggested that repeated practice attempts might improve or at least maintain tempo accuracy, except in music exhibiting low note density – below approximately 1.5 notes-per-second. However, musicians do not seem to have immediate conscious access to the gains or losses of tempo accuracy in mental practice.
The second part of the study analyzed the effect of instrument tone-decay time on average-tempo differences among repeated recorded performances of North-American Folk music and Western-Classical music. Recordings were paired in such a way that each piece was represented by performances on two (banjo and guitar) or three similar instruments (harpsichord, fortepiano, and piano). Varying tone-decay time comprised one salient difference within each instrument group. The results failed to support our prediction regarding tone-decay length and tempo differences. There was no significant difference of a work’s average tempo when it was performed on different instruments. Post hoc analyses of the folk songs showed that rhythmic speed was higher in performances on banjo than on guitar. This suggests that works not having a definitive version, as in a musical score, might exhibit faster rhythms when performed on instruments with shorter tone-decay times.