rience) x 2 (gender) x 2 (age) x 4 (music emotional quahty) with repeated mea- sures on the latter factor, the main effects of age (F,,,, = 5.00, p< .05) and.
Perceptual and Motor Skills, 1994, 79, 939-944.
O Perceptual and Motor Skills 1994
EMOTIONAL REACTIONS T O MUSIC BY GENDER, AGE, AND EXPERTISE ' CLAUD10 ROBAZZA, CRISTINA MACALUSO, VALENTINA D'URSO University of Padua Summary.-Fragments of classical music were submitted to 80 subjects, 40 children 9 to 10 years old and 40 adults 19 to 29 years old who were divided into eight groups of ten, to induce feelings of happiness, sadness, anger, and fear. The task required linking each piece of music to one emotion, identifying at the same time the intensity of the emotional response on a scale of 1 to 3. The goal was to study how gender, age, and exposure or expertise related to emotional perceptions of music. Analysis showed (a) experts in music and nonexperts ascribed similar emotions to pieces of music, (b) there was no difference in emotional response to music by gender, although women linked to music stronger emotions of anger than girls, (c) children perceived greater feeling of happiness in music and less feeling of anger than adults, and (d) emotions of anger and fear in music were often confused with one another.
Music has the power to create feelings and emotions in a quick and effective manner. What in a book would require many sentences for description or in a drama several lines, in music can often be conveyed by just one measure or one chord. Music is an eloquent means by which to express emotions (Ostwald, 1966); it can influence moods (Wheeler, 1985) and induce strong emotions (Kenealy, 1988; Rosenfeld, 1985). Davies (1978) and Radocy and Boyle (1979) stated that emotional reaction to music is based on social environment. According to Meyer (1956), instead, emotions come from the structure of music itself: delay in resolution of a musical phrase, without immediate satisfaction of expectations, raises tension while the subsequent resolution determines the emotional experience. Several authors have looked into how factors such as gender and age influence music perception. In one of the first studies children, high-school students, unjversity graduates, and expert musicians attributed to music virtually the same meaning (Watson, 1942). Learning, mood, musical style, and personal emotions influenced emotional responses (Sopchak, 1955). Emotional reactions were also linked to familiarity with music (Hart & Cogan, 1973). Through repeated listening comprehension of a piece of music increased but the music did not necessarily become more pleasant (Gaver & Mandler, 1987). Differences between gender and age were found by NielzCn and Cesarec (1981): women perceived high tension in musical fragments characterized as stressful and older adults found classical music more attractive than youngsters. In a survey concerning children's ability to recognize music ' ~ i r e c correspondence t to Claudio Robazza, Istituto Superiore Educazione Fisica, Via dei Colli 4, 35100 Padova, Italia.
940
C. ROBAZZA, ETAL.
of four epoques (baroque, classical, romantic, and modern), the a b h t y to identify music pieces increased with age (Gardner, 1973). Moreover, children showed less adherence to cultural rules in classification (Hargreaves, 1982) and were spontaneously inclined to define music with emotional terms. I n research on characteristics of music perception, Flowers (1983) discovered that children differed from adults in the description of musical pattern: children described tone, beat, and extra-music environment while adults referred also to height and melody. Dolgin and Adelson (1990) examined the ability to recognize melodies expressing "happy," "sad," "fearful," and "angry" feelings by age (children aged 4, 7, and 9 and adults) and gender. The melodies were presented by vocal and instrumental modalities, as a soprano voice singing nonsense syllables and a viola. The ability to discriminate emotions in music improved with age while partial differences were found between men and women. Some emotions were easier to interpret than others; happy and sad songs were more readily recognized while frightening and sad-sounding melodies tended to be confused. Similar results were reported by Terwogt and Van Grinsven (1988). Children 5 to 6 years old and 9 to 11 years old and adults linked musical fragments and emotions in a similar manner; however, adults identified emotions more accurately. Further, anger and fear were often confused with one another. The influence of gender and age on emotional perceptions in music was further investigated in this study as well as music experience. O n the basis of previous research, the following hypotheses can be stated: (1) experts and nonexperts claim similar emotions to musical fragments, (2) in children emotional response to music is not different by gender, (3) women attribute to music more stressful emotions of anger and fear, and (4) in adults and children there is disagreement when choosing the type of emotion related to music.
METHOD
Subjects Eighty subjects took part in this research, 40 children (20 boys and 20 girls) aged 9 and 10 years (M = 9.8, SD = .4) and 40 adults (20 men and 20 women) aged 19 to 29 years (M = 24.0, SD = 2.4). Children were divided according to their exposure to music, 20 children from an elementary school where music was taught and 20 from schools where music was not taught. Adults were divided according to musical expertise, 20 adults with musical instrument diplomas from a school of music and 20 with no music experience at all. Each group was made up of equal numbers of males and females, so eight groups of ten subjects based on gender, age, and musical expertise (or exposure) were formed.
Procedure Eight fragments of classical music of equal length of about 5 0 sec. each
EMOTIONAL REACTIONS T O MUSIC
94 1
were selected by music experts according to their emotional quality. This selection was based on tone and subject criteria, choosing exclusively piano execution pieces with titles referring to a particular emotional content. Two pieces for each emotion were chosen: "Happiness," measures 1 to 26 from Liadov's "The Musical Snuff-Box" and measures 1 to 30 of "The Tuileries" from Mussorgsky's "Pictures at an Exhibition"; "Sadness," measures 1 to 10 from the third tempo of Beethoven's "Sonata No. 29, Op. 106" and measures 1 to 31 of Chopin's "Good-bye Waltz" (No. 9); "Anger," measures l to 90 of "The H u t on Chicken Legs" (Baba-Iaga) from Mussorgsky's "Pictures at an Exhibition" and measures 1 to 25 of "Mazzeppa" (Studio No. 4) from Liszt's "Transcendental Studies"; and "Fear," measures 38 to 59 of "Gnomus" from Mussorgsky's "Pictures at an Exhibition" and measures 1 to 110 of Liszt's "Mefisto Waltz." Each subject was asked to associate a single emotion, "happiness," "sadness," "anger," or "fear," to each of the eight fragments and give a score of 1 to 3 for the intensity of emotional response. I n other words, subjects were invited to identify one of the moods elicited by each piece of music and to rate its intensity. Each subject participated in the research individually and listened to the music recorded on a tape while comfortably seated. The order of presentation of the fragments was counterbalanced.
RESULTS Scores given by subjects to each of the four emotions, that is, the over-all intensity of the mood elicited by the eight musical fragments, were considered as the dependent variable. Mean scores and standard deviations are reported in Table 1. From a multivariate analysis of variance, 2 (experience) x 2 (gender) x 2 (age) x 4 (music emotional quahty) with repeated measures on the latter factor, the main effects of age (F,,,, = 5.00, p < .05) and emotional quality (F,,,,, = 7.29, p < .01) were significant. Significant also were interactions of gender x age (F,,,, = 5.82, p < .05) and age x emotiond quality (F,,,,,= 4.69, p
