Musicology meets MIR
The methodologies of musicology and Music Information Retrieval (MIR) are close to being opposites of each other, while both are concerned with addressing the same underlying questions. In this overview, I will consider developments in MIR as representative of trends in machine learning and artificial intelligence more generally, and how these can impact the practice of musicology (broadly speaking), in particular with reference to non-Western music. We could consider the use of MIR tools as an adjunct to music-analytic practices (computer-in-the-loop), or the use of musicological expertise in the training of automatic learners (human-in-the-loop). What perspectives are opened onto music analysis and artificial intelligence by looking at these practices in both directions?