Thinking in Music: An Objective Measure of Notation-Evoked Sound Imagery in Musicians

Being able to imagine the sound of music from notation as a result of so-called notational audiation, without the physical presence of sound, is an indispensable skill for professional musicians. However, up until now there has been no assessment for the evaluation of the skill responsible for the reading-imagery task. The development of an assessment for the skill of using notation-evoked sound imagery (henceforth called NESI) described in this study was based on the embedded-melody paradigm. Participants first had to read a figural variation from notation and imagine the sound without singing or humming. After the notation disappeared from the screen, the sound of a theme was played, which either matched the original melody harmonically or contained small but significant deviations from the original harmonic progression (the so-called “lure” variation). Participants had to decide whether the variation and theme matched in terms of their harmonic structure. We analyzed the item characteristics in three pilot studies to select a reduced and validated number of item triads. Internal validity and unidimensionality were tested by use of item response theory. In the main study, this selection of item triads was validated by a sample of N = 55 music students in a comprehensive test setting to determine correlations with other subskills such as ear training skills, absolute pitch, working memory capacity, or melodic memory. The final path model showed that NESI has a strong relationship with general ear training skills but is only slightly correlated with working memory, spatial orientation and melodic memory.