Zentner, M., Didier, G., & Scherer K.R. Emotions Evoked by the Sound of Music: Characterization, Classification, and Measurement. Emotion, 8, 494-521.
By: Andrea Botticelli
The purpose of study 3 in this ongoing investigation was to extend the findings by using a larger and more representative sample of listeners. This round of tests examined emotion ratings provided when listeners were exposed to actual performances. Confirmatory factor analysis examined the structure of musical emotion ratings (496). Study 4 attempted to replicate the results with a different sample of listeners and musical excerpts. It also compared the differential validity of the framework in comparison with basic emotion and dimensional emotion models (497).
In the validity test, there was a good fit result for a more parsimonious 9-factor model. Subesequently, the model was reduced to include 9 factors. Also, the total number of emotion terms was reduced to 40. The 9 factors of musical emotion are: Wonder, Transcendence, Tenderness, Nostalgia, Peacefulness, Power, Joyful Activation, Tension, and Sadness (503). The results provide a domain-specific taxonomy of musically induced emotions. They make up the Geneva Emotional Music Scale (GEMS) (506).
In comparison with mainstream emotional models, such as the discrete emotion model, most musical emotions in the GEMS model are positive. Also, emotion categories such as wonder, nostalgia, and transcendence are prevalent in the model, whereas these emotions are not central in any other emotion model. There are some other more subtle differences in the emotion terms. For instance, joy in music implies a tendency to dance, which is unlike the common emotional meaning of joy. In music “joyful activation may be best seen as a form of joyful entrainment” (506). Also, musical sadness may not be like the basic emotion of sadness because respondents rarely reported feeling gloomy, depressed, or unhappy. The nine emotion factors can be classified with three larger terms: Sublimity, Vitality, and Unease. While the intercorrelations of these facets seem disturbingly high from the point of view of statistical procedure, they reveal the blended nature of musical emotions (506).
In study 4, the method to replicate the results in study 3 used only absolute music. There were some added variables gleaned from repeated suggestions by participants in free response categories of questionnaires (508). Therefore, the final taxonomy has 9 factors and 33 subfactors (509). Significantly, results using this model showed that listeners preferred to describe what they felt in these emotional terms. It also enhanced agreement across listeners in emotional ratings of music excerpts (511). Hence, the musical emotion model reflects more discrimination between musical excerpts (512). It is a step closer to developing a reliable method to evaluate musically induced emotions.
Reflection:
The final discussion in this seminal article makes the point that even with this more sharply discriminating model, there was still significant interindividual variability in emotional responses to a given excerpt (512). My question is what causes this variability? It is clear from this research (and my own experience as a music teacher) that not all individuals react emotionally to music. In this case, less than 50% of the sample reacted to the most common musically induced emotions. There has been research that illuminates other factors that moderate emotional reaction to music, such as performance variables, listener variables and contextual variables (516). In my opinion, there must be a domain-specific musical ability or awareness that accounts for this variability.
By: Andrea Botticelli
The purpose of study 3 in this ongoing investigation was to extend the findings by using a larger and more representative sample of listeners. This round of tests examined emotion ratings provided when listeners were exposed to actual performances. Confirmatory factor analysis examined the structure of musical emotion ratings (496). Study 4 attempted to replicate the results with a different sample of listeners and musical excerpts. It also compared the differential validity of the framework in comparison with basic emotion and dimensional emotion models (497).
In the validity test, there was a good fit result for a more parsimonious 9-factor model. Subesequently, the model was reduced to include 9 factors. Also, the total number of emotion terms was reduced to 40. The 9 factors of musical emotion are: Wonder, Transcendence, Tenderness, Nostalgia, Peacefulness, Power, Joyful Activation, Tension, and Sadness (503). The results provide a domain-specific taxonomy of musically induced emotions. They make up the Geneva Emotional Music Scale (GEMS) (506).
In comparison with mainstream emotional models, such as the discrete emotion model, most musical emotions in the GEMS model are positive. Also, emotion categories such as wonder, nostalgia, and transcendence are prevalent in the model, whereas these emotions are not central in any other emotion model. There are some other more subtle differences in the emotion terms. For instance, joy in music implies a tendency to dance, which is unlike the common emotional meaning of joy. In music “joyful activation may be best seen as a form of joyful entrainment” (506). Also, musical sadness may not be like the basic emotion of sadness because respondents rarely reported feeling gloomy, depressed, or unhappy. The nine emotion factors can be classified with three larger terms: Sublimity, Vitality, and Unease. While the intercorrelations of these facets seem disturbingly high from the point of view of statistical procedure, they reveal the blended nature of musical emotions (506).
In study 4, the method to replicate the results in study 3 used only absolute music. There were some added variables gleaned from repeated suggestions by participants in free response categories of questionnaires (508). Therefore, the final taxonomy has 9 factors and 33 subfactors (509). Significantly, results using this model showed that listeners preferred to describe what they felt in these emotional terms. It also enhanced agreement across listeners in emotional ratings of music excerpts (511). Hence, the musical emotion model reflects more discrimination between musical excerpts (512). It is a step closer to developing a reliable method to evaluate musically induced emotions.
Reflection:
The final discussion in this seminal article makes the point that even with this more sharply discriminating model, there was still significant interindividual variability in emotional responses to a given excerpt (512). My question is what causes this variability? It is clear from this research (and my own experience as a music teacher) that not all individuals react emotionally to music. In this case, less than 50% of the sample reacted to the most common musically induced emotions. There has been research that illuminates other factors that moderate emotional reaction to music, such as performance variables, listener variables and contextual variables (516). In my opinion, there must be a domain-specific musical ability or awareness that accounts for this variability.
No comments:
Post a Comment