Biases in the perception of dynamics in harpsichord performance

It is widely claimed that it is not possible to vary dynamics on theharpsichord through touch; however, recent studies on single tonesshow that small dynamic differentiation can be obtained depending onthe type of keypress. These differences are perceived accurately incomparisons of single tones; within a musical excerpt, they may beharder to detect, especially if biases exist. This study explores whetherpreconceptions regarding the ability of the harpsichord to producedynamics influences perception of such differences. Two experimentsare detailed: Experiment 1 uses two participant groups to test theeffect of a bias (factor “cover story”) in the perception of two types ofmusical excerpt (factor “stimulus set”: dynamic variationperformed/no dynamic variation performed). Experiment 2 adds a setof artificially manipulated excerpts to the stimuli, and presents alltrials to participants in a fully within-groups design. Results ofexperiment 1 show a main effect of stimulus set (F (1, 30) = 24.01, p< .001). Experiment 2 results show a main effect of cover story (F(1,81) = 80.67, p < .001), and of stimulus set (F (2.30, 186.41) = 24.60,p < .001), with no interaction. The main effect of the stimulus set inboth experiments demonstrates that it is possible to effect theperception of dynamics in musical excerpts. These results haveimplications for the understanding of historical performance practices.