Words, music, and other measures: Predicting the repertoire popularity of 597 Schubert lieder.

Abstract

To address the venerable aesthetic question of the relative importance of music and words, musical and textual characteristics were used to predict recording counts of 597 Franz Schubert lieder. Predictors included career variables (age and concurrent productivity), “whole song” variables (e.g., performance duration, structural complexity, major vs. minor key), thematic measures of the first six notes of the vocal incipit (e.g., melodic and rhythmic originality), poet variables (Goethe, Schiller, etc.), and text variables (computer content analysis measures of primordial and conceptual thought and positive and negative emotions). Many correlations between predictors were small but reliable; career variables accounted for most strong correlations. A 45-predictor regression model was highly significant (R .63), with 13 reliable ( p .05) predictors, mostly career and whole song variables plus two text measures. Finally, “local” measures of the vocal incipits revealed that emotion words are related to rhythmic, rather than tonal, aspects of the melody.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 9 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.