Method

Participants. Thirty-three participants (out of 44) completed the task successfully.

Apparatus. Prime contexts were played with a flute timbre (XG074) and target tones with either an oboe (XG069, Timbre A) or a clarinet timbre (XG072, Timbre B). These timbres were chosen with the aim to have relatively similar timbres and minimize possible deviant effects (see Tillmann et al., 2006).

Procedure.Because discriminating between the two target timbres was expected to be more difficult than in Experiment 1a, the training phase was increased: Participants were trained on 28 single tones, 28 tone pairs and 6 example melodies. The experimental phase consisted of 44 melodies (i.e., 11 melody pairs with the target being played once by Timbre A and once by Timbre B).