Abstract: To mark the 250th anniversary of Beethoven’s birth, Deutsche Telekom initiated and funded "Beethoven X", a project to create some kind of realization of Beethoven’s fragmentary sketches for a 10th symphony using machine learning. This experimental undertaking brought together a team of music and computer science experts and shone a light on the many musicological and computational decisions that go into such a task. This talk uses our experience with the Beethoven X Project to discuss some of those general considerations, including the nature of possible human-computer interactions, (dis-)similarities between computational and traditional (non-computational) methods of composing, and the probably decisive role the next generation of this work will reserve for computational analysis.
Bio: Mark Gotham specialises in computational methods for music theory, analysis, and composition. He holds the rare distinction of having been appointed to faculty positions in both STEM (Assistant Professor of Computer Science at Durham University, from 2023) and in the humanities (previously Professor of Music Theory at T.U. Dortmund). Previous, non-faculty positions include those at Cornell university (US), Universität des Saarlandes (DE), the University of Cambridge (UK) and the Royal Academy of Music (UK).
Mark holds a PhD from Cambridge, MMus from the RNCM, and a BA from Oxford (where he graduated at the top of his cohort). He runs FourScoreAndMore.org to widen access to music theory and the debut CD of Mark’s compositions – ‘Utrumne est Ornatum’ – attracted 5-star reviews.