This experiment was attempted to investigate the strategies of pitch coding in melody processing. Twenty-six musically trained and twenty-six musically naive subjects were instructed to make recognition for melodies after a 12-s retention interval, during which four conditions of interference (i.e. pause, interfering melody, nonsense syllables, musical note names) were interpolated. Both the standard and comparison melodies were 6-tone sequences which composed of high-tonality structure or low-tonality structure. It was found that the recognition of musicians was severely distrupted by the “note names” under a tonal melody, while it was disrupted by “interfering melody” under an atonal melody. On the other hand, the recognition of non-musicians was significantly worse than musicians, and there were no significant differences in distruptive effects between the interferences. These findings suggest that musicians could used verbal (note names) coding strategy for tonal melody and sensory pitch coding strategies (ex. humming, whistling) for atonal melody, but that non-musicians could not use any effective strategies for melody coding. Results were also discussed in relation to contour and pitch of melodies.
