Statistical Learning in the Music Theory Classroom

Lessons from Teaching Classical Form and Galant Schemas

Abstract

This article argues that statistical learning can and should be harnessed as a core pedagogical strategy in the music theory classroom. Statistical learning describes the largely implicit process by which learners extract patterns and regularities from repeated exposure to structured stimuli. This mechanism is widely acknowledged as central to how listeners acquire musical style outside the classroom, but it has rarely been treated as a deliberate pedagogical resource within formal music theory instruction. By explicitly centering statistical learning, instructors can design pedagogies that are more effective and grounded in lived musical experience. The article begins by distilling pedagogically salient findings from the statistical learning literature. I show that a pedagogy centered on statistical learning involves not only immersive exposure, but also the use of naturalistic, cue-rich materials calibrated to balance novelty and familiarity, as well as assessments that emphasize prediction and judgment over knowledge reproduction. I then demonstrate the efficacy of these principles through two example lessons: one in which students sketch a two-part exposition in response to brief symphonic excerpts, and another in which students develop fluency with the Do-Re-Mi schema through a sequence of listening, composition, and comparative judgement tasks. Through these examples, I show how statistical learning supports the development of intuitive musical understanding while also furnishing a robust substrate for explicit analysis, evaluation, and creative practice.