A Swan Song for the Swan-Song Phenomenon: Multi-Level Evidence against Robust End-of-Life Effects for Classical Composers

Abstract

Simonton (1989) examined late-life creativity, analyzing 1,919 works by 172 classical composers, and found that their late compositions were less melodically original (and sometimes shorter in duration) but more critically acclaimed. He suggested that composers approaching death may show a last burst of creativity—a “swan-song” effect. Here we refined that analysis, examining 2,270 works by 57 composers, employing variables isomorphic to those of Simonton but re-analyzed using hierarchical linear modeling. Five characteristics of each work (both with and without overall age controls) were used to predict linear or exponential time until death. Results showed few reliable effects, sometimes in the opposite direction than found earlier. Isomorphic regression analyses were also performed on each composer individually; not one case of the swan-song effect as described by Simonton was found. These results suggest that the swan-song phenomenon—at least as presently characterized—is not reliable.