In a debate in this journal concerning the importance of structure in the effects of musical compositions, Robert Batt suggested the first movement of Mozart's Symphony in G Minor K. 550 as the ideal example on which research could contravene the minimal effects of structural interventions that have been observed in prior research by Konečni. This article reports the results of putting Batt's suggestion to empirical test. As Batt pointed out, the first movement of K. 550 can be broken down into nine segments that can be broadly defined as expository (two themes), developmental, and recapitulative (two themes). The original and four differentially intrusive versions of the piece were presented to 42 college non—music-student subjects and 11 college music-student subjects. On the dimensions of pleasingness, interestingess, and a desire to own a copy of the piece, the original version consistently failed to elicit greater preference than the altered versions did in both subject populations, although all versions were much liked by these nonmusician and musician subjects.
Different recordings and arrangements of the Goldberg Variations by Johann Sebastian Bach were evaluated on cognitive, emotional, and perceptual dimensions. In Study 1, eight different renditions of the piece were classified as Classical or Romantic interpretations. Both harpsichord and piano versions of each style were included. Only minimal differences were found in subjects' appreciation for the harpsichord versus the piano recordings. Comparisons between the Classical and Romantic styles also revealed only slight differences in subjects' ratings. In addition, no differences were found in subjects' enjoyment of different recordings made by a single artist at different points in his career. In Study 2, the structure of the piece was modified by rearranging the order of the variations. A preference for the original version over the modified arrangements was indicated on only 1 of the 15 dimensions measured. In Study 3, specific triplets of variations were played to subjects in their original order and in a random sequence. No differences were found in subjects' appreciation for the original versus the modified versions.