The recognition of emotional qualities expressed in music. Proceedings of the 9th International Conference on Music Perception & Cognition (ICMPC9), Bologna/Italy, August 22-26 2006. ISBN 88-7395-155-4 c 2006 ICMPC-

The study tested the hypothesis that people’s emotion knowledge enables them to recognize subtle differences in the emotional quality expressed by a musical piece, and thus to distinguish Tenderness (T), Sadness (S), Anger (A) , and Fear (F) one from another, and each from Other emotions (O), such as Joy (J). To test the hypothesis, we designed three experiments in which subjects (half of whom were musicians in each experiment) expressed their perception while listening to music pieces. In the first experiment (1E), that tested whether subjects distinguish Tenderness from Sadness, subjects evaluated stimuli at two points in time for each piece, i.e., at 40sec and at 80 sec. In the second experiment (2E), that tested whether subjects distinguish Anger from Fear and Joy, subjects evaluated stimuli only once at 40 sec. In the last study, (3E), that tested whether subjects distinguish Anger from Fear and Sadness, subjects evaluated stimuli only once at 40 sec. Their latency in response time (reaction time, RT), as well as their verbal rating were recorded in every experiment. In 1E subjects expressed their rating on a 7-point Likert-type scale, with extreme points labelled T and S, and the intermediate labelled O (Other emotion). For 2E ad 3E, subjects simply choose a verbal label. Experimental stimuli were classical music pieces, for orchestra or solo instruments (e.g., piano; violin and piano), each lasting 85 seconds in 1E, and 40 seconds in 2E and 3E, selected from XVII-XX century repertoire, by several authors (Fauré, Grieg, Beethoven, Chopin, etc.) that contained the structures characterizing (Sloboda & Juslin 2001) such as S or T, A or F and J (e.g., Slow tempo for both; Soft timbre for T; Dull timbre for S). The results overall confirmed the hypothesis, but also showed that perception is influenced by other variables. In particular, in 1E both musical expertise, and 'location' of an emotional quality within a musical piece influenced the perceptions (e.g., in Tenderness was most easily perceived in single-emotion pieces, and least distinguished from Sadness when it followed it rather than preceding it). In 2E we found that Anger and Fear are often confused with each other, whereas Joy is correctly recognized, as found in previous studies. In 3E Anger and Fear were correctly distinguished one from the other more often than in study 2E; Sadness was more often recognized correctly than Anger and Fear; RT were shortest when pieces expressed Fear rather than Anger and Sadness, RT were longest for Sadness.