
Tipping the Iceberg: Estimating Lost Fourteenth-Century Music
Michael Scott Asato Cuthbert, teaching MIT's 21M.383 Computational Music Theory and Analysis, presents statistical methods for estimating how many musical works from the fourteenth century have been lost entirely. Rather than studying surviving manuscripts directly, Cuthbert treats the known corpus as a sample and applies techniques borrowed from ecology and statistics, the kind used to estimate undiscovered species from observed ones, to infer the size of the original repertoire before losses from war, decay, and neglect. The talk walks through how survivorship bias distorts music history, why counting what remains understates what existed, and how a model can put rough numbers on the gap. It is a short, focused segment within the course rather than a full lecture on medieval music itself, aimed at students already following the course's computational methods. The emphasis stays on the statistical reasoning and its implications for how confidently historians can describe a musical era from fragmentary evidence.