Toward a Computational Theory of the Cognitive Relevance of Musical Similarities and Repetitions

In this study we inquire from a speculative but computationally informed perspective, whether repetition and similarity have an effect upon the calculation of musical relevance. Similarity in general is not a theoretical object easily tractable from a computational point of view, but in the case of the musical domain it seems possible to formalize it. We will proceed by combining Relevance Theory with Generative Theory of Tonal Music in the search for an algorithm for the calculation of Musical Relevance. This is the first version of a germinal study and is intended as a preliminary investigation of a new research area in cognitive and computational musicology.