Method

Participants. Twenty-two students from the University of Lyon participated in Experiment 1. Musical expertise, as measured by instrumental instruction, varied from 0 to 13 years (mean = 4.1  5.0 years; median = 0.5).

Material. Twelve pairs of melodies were composed to cover the 12 major keys (see Figure 1 for an example). All melodies were two bars of four beats long. Melodies within a pair were based on the same rhythmic pattern for the two bars: the first three beats consisted of eighth and/or sixteenth notes and the fourth beat was a quarter note. The specific rhythm of the first three beats was varied across melodic pairs. The second quarter note (the last note of the second bar) was the target tone. The melodies of a pair differed only by one (possibly repeated) tone in the first bar. This tone was a natural note in one of the two melodies and a sharp or flat note in the other. This difference changed the tonal function of the target to be either the tonic (related target) or the subdominant (less-related target). For example, in Figure 1, the target tone G-flat functions as a tonic in the related condition because the used tone set defines the tonality of G-flat major. The same target tone G-flat functions as a subdominant in the less-related condition because the altered C-natural in the first bar put the melody in the tonality of D-flat major. Eight example melodies were added as training examples for the task.

Apparatus. Melodies were created in MIDI with Cubase SX2 software (Steinberg) and the maxim digital audio (mda) piano soundfont (free software by Paul Kellett, Mac conversions by Glenn Olander and Brian Chrisman, 1998-2000). Prime contexts were played with the default settings of the mdaPiano. For the target tones, the Muffle (low pass filter) and Hardness (control of brightness) settings were changed to obtain either a dull or a bright piano timbre. Melodies were played at a tempo of 789.5 ms per beat (i.e., 76 beats per minute). The experiment was run on Psyscope software (Cohen, MacWhinney, Flatt, & Provost, 1993) with the Psyscope ButtonBox (New Micros, Inc., Dallas). For the pretest, the entire melodies were played with the default settings of the mdaPiano.

Procedure. The experiment contained a training phase and an experimental phase. In the training phase, participants had to identify the two timbres with 24 single tones and 8 example melodies. In this timbre identification task, they were asked to judge as quickly and as accurately as possible whether the single tone (or the last tone of the melodies) was played by Timbre A or Timbre B, by pressing one of two buttons on the ButtonBox. Timbre A referred to the dull piano timbre and Timbre B to the bright piano timbre. Example tones for each timbre were presented at the beginning. In the experimental phase, participants made the timbre identification task on the target tones of the 48 melodies (i.e., the twelve melody pairs presented with both timbres). The melodies were presented in pseudo-random orders for each participant. Pseudo-random orders were random orders with two constraints: the two melodies of a pair were separated by at least four other melodies, and a given target timbre (Timbre A, Timbre B) was not repeated more than five times in succession. In both training and experimental phases, an incorrect response was followed by an alerting feedback and a correct response stopped the target to encourage participants to respond as quickly as possible. The target was followed by a 250 ms noise mask and participants had to press the third (middle) button on the ButtonBox to start the next trial.

Figure 1.One example of the 12 pairs of melodies. The note differing between related and less-related melodies can be visually identified by the alteration marks. Examples of sound material are available at: http://olfac.univ-lyon1.fr/bt-sound.html.