Você está na página 1de 7

Music2703_02 1/11/10 3:38 PM Page 177

Intensity of Emotions Conveyed and Elicited 177

I NTENSITY OF E MOTIONS C ONVEYED


AND E LICITED BY FAMILIAR AND U NFAMILIAR M USIC

. .
S. O MAR A LI AND Z EHRA F. P EYNIRC I O G LU This study examined the role of familiarity in modulat-
American University ing emotional response to music. Experiment 1 replicated
the positive association between familiarity and liking
WE REPLICATED PREVIOUS FINDINGS AND DEMONSTRATED (Gaudreau & Peretz, 1999; Hargreaves, 1984; Peretz,
that familiarity with musical stimuli increased liking Gaudreau, & Bonnel, 1998; Sampson & Peretz, 2005;
or preference for the stimuli. We also demonstrated Stevens & Latimer, 1991; Szpunar, Schellenberg, & Pliner,
that familiarity increased the intensity of emotional 2004; Tan, Spackman, & Peaslee, 2006) and examined
responses to music, but only when the stimuli were whether a similar finding would emerge with emotions.
made highly familiar through en masse repetitions Because liking is a positive state or attitude, the liking
(Experiment 3) rather than through interspersed repe- expressed by participants for more familiar music could
titions (Experiment 1). In addition, intensity ratings imply an underlying emotion (happiness or pleasure).
were higher when participants were asked to judge the Experiment 2 examined whether any differences would
emotion conveyed by the music than when they were emerge when participants were asked to rate the intensity
asked to judge the emotion elicited by the same music of emotions conveyed by music compared to the intensity
(Experiments 2 and 3). Finally, positive emotions (i.e., of emotions elicited by the same music. The emotion that
happy and calm) were rated higher compared with neg- music intends to convey and the ability to judge its
ative emotions (i.e., sad and angry) for both types of intended intensity does not necessarily imply that the
ratings (i.e., conveyed or elicited). The findings suggest emotion is actually experienced in the same way (e.g.,
that familiarity plays a role in modulating a listeners Gabrielsson, 2001/2002). Experiment 3 combined the
emotional response to music. methods of Experiments 1 and 2 and examined whether
responses differed as a function of familiarity when par-
Received July 3, 2008, accepted September 16, 2009.
ticipants were asked to rate the intensity of emotions con-
Key words: music, emotion, familiarity, preference, veyed by the music as compared to that elicited by the
repetition same music.

Experiment 1

The purpose of this experiment was to replicate the

C
ONSIDERABLE RESEARCH HAS EXPLORED HOW MUSIC
positive association between familiarity and liking (or
influences emotions (e.g., Juslin & Sloboda,
preference) reported by others (e.g., Hargreaves, 1984)
2001); however, few studies have examined the
and to explore whether the liking expressed by partic-
role of familiarity in modulating emotional response to
ipants could be construed as an underlying emotion
music. Listeners often play music to enhance their current
(i.e., happiness or pleasure).
mood state (e.g., playing romantic music during a can-
dlelit dinner; Knobloch & Zillmann, 2003), or to alter
Method
their current mood state (e.g., playing relaxing music
during stressful situations; Stratton & Zalanowski, PARTICIPANTS
1997). Yet, as listeners become more familiar with music, it The participants were 64 American University students
is unclear whether the subjective emotional intensity of who received extra credit in psychology courses.
the music increases or diminishes. Meyer (1956) postu-
lates that expectations play a key role in the interpretation MATERIALS, DESIGN, AND PROCEDURE
of emotion in music. Thus, decreased novelty may dimin- The stimuli comprised 32 musical excerpts with no
ish the listeners emotional intensity for the music, but lyrics, all taken from instrumental classical and jazz
increased familiarity may enable listeners to predict what pieces as well as soundtracks (see Ali & Peynircioglu,
follows in the music, resulting in anticipatory arousal. 2006, for the list of musical excerpts). The excerpts

Music Perception VOLUME 27, ISSUE 3, PP. 177182, ISSN 0730-7829, ELECTRONIC ISSN 1533-8312 2010 BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIA . ALL
RIGHTS RESERVED. PLEASE DIRECT ALL REQUESTS FOR PERMISSION TO PHOTOCOPY OR REPRODUCE ARTICLE CONTENT THROUGH THE UNIVERSITY OF CALIFORNIA PRESS S
RIGHTS AND PERMISSIONS WEBSITE , HTTP :// WWW. UCPRESSJOURNALS . COM / REPRINTINFO. ASP. DOI:10.1525/MP.2010.27.3.177
Music2703_02 1/11/10 3:38 PM Page 178

178 S. Omar Ali & Zehra F. Peynirciog lu

were chosen to express four different types of emo- much time as needed to complete them before listening
tions (happy, sad, calm, and angry) corresponding to to the next excerpt.
the four quadrants of the circumplex model of emotion
adapted by North and Hargreaves (1997). Eight musi- Results and Conclusions
cal excerpts of approximately 20 s in length were used
for each emotion. Extensive pilot testing ensured that Table 1 presents the mean ratings from the targeted
these stimuli were not highly familiar to the participant rating scale (i.e., happy, sad, calm, or angry) for the
population. intended emotion of the musical excerpts. The ratings
Participants were tested either individually or in small were evaluated using multivariate analysis of variance
groups. There was a familiarization phase followed by a (MANOVA) with repeated measures.
rating phase. For the familiarization phase, half of the 32 There was a significant effect for familiarity, Wilks
excerpts (four excerpts from each of the four emotions) lambda = .82, F(1, 62) = 13.45, p < .05; the participants
were randomly selected and presented three times, with gave higher ratings to familiar music compared to the
each presentation of a given excerpt separated by at least unfamiliar music. However, there was also an interaction
two other excerpts. The excerpts were interchanged across between familiarity and type of rating, Wilks lambda =
two groups of participants so that each excerpt served .92, F(1, 62) = 5.29, p < .05. Posthoc Bonferroni com-
in the familiar and unfamiliar conditions equally often. parisons showed that liking was influenced by familiarity,
To make the familiarization phase meaningful and engag- t(254) = 2.10, p < .05, but not by emotional intensity,
ing, the participants were told to learn each excerpt for t(254) = 0.63, p > .10. Even the happy emotion by itself
a memory test (which did not exist) and to rate its was unaffected, p > .05.
memorability on a nine-point scale (1 representing There was a significant effect for the intended emo-
not memorable and 9 very memorable). tion of the music, Wilks lambda = .60, F(3, 60) = 13.36,
During the rating phase, all of the excerpts were pre- p < .05, but this effect did not depend on familiarity,
sented randomly with respect to the type of emotion. Wilks lambda = 0.94, F(3, 60) = 1.27, p > .10. There was
Half of the participants rated how much they liked each also an interaction between the intended emotion of
musical excerpt on a nine-point scale (1 representing the music and type of rating, Wilks lambda = 0.73,
not much and 9very much). The other half rated the F(3, 60) = 7.57, p < .05. Posthoc Bonferroni compar-
intensity of the emotion in each of the four categories: isons showed that the participants rated the happy as
(1) happy, joyful, exciting, or festive, (2) sad, depress- well as calm music higher than the sad and angry music
ing, or melancholy, (3) calm, relaxing, or peaceful, and on emotional intensity, ts(126) = 6.69, 4.01, 8.77, 6.46,
(4) angry, unsettling, disconcerting, or stressful, always respectively, all ps < .05. Consistent with our previous
in that order, regardless of the intended emotion of the results (Ali & Peynircioglu, 2006), participants rated
excerpt. The emotion ratings also were given on nine- the intensities of the positive emotions higher than those
point scales, 1 representing not (e.g., not happy) of the negative emotions (see also Pfister & Bhm,
and 9 very (e.g., very happy). Participants made 1992). Interestingly, this phenomenon did not occur
their liking or emotion ratings at any time during the when participants were asked to rate their liking of the
20 s presentation of each excerpt and were given as excerpts, all ps > .10.

TABLE 1. Mean Intensity Ratings (SD) of Liking and Emotional Responses to Musical Excerpts as a Function of the Familiarity Condition and the
Intended Emotion.

Emotion of Melodies

Happy Sad Calm Angry Overall

Liking
Familiar music 5.10 (1.50) 5.40 (1.94) 4.98 (1.97) 4.64 (1.78) 5.03 (1.81)
Unfamiliar music 4.72 (1.41) 4.81 (1.49) 4.70 (1.90) 4.11 (1.37) 4.59 (1.57)
Overall 4.91 (1.45) 5.11 (1.74) 4.84 (1.93) 4.38 (1.60) 4.81 (1.70)
Emotion
Familiar music 7.70 (0.90) 6.65 (1.19) 7.55 (0.98) 6.15 (1.31) 7.01 (1.27)
Unfamiliar music 7.91 (0.93) 6.66 (0.86) 7.20 (1.01) 5.87 (1.41) 6.91 (1.30)
Overall 7.80 (0.92) 6.65 (1.03) 7.37 (1.00) 6.01 (1.36) 6.96 (1.28)
Music2703_02 1/11/10 3:38 PM Page 179

Intensity of Emotions Conveyed and Elicited 179

TABLE 2. Mean Intensity Ratings (SD) of Emotional Responses to Musical Excerpts as a


Function of the Type of Emotion and the Type of Rating.

Type of Rating

Intended Emotion Conveyed Elicited Overall

Happy 3.63 (0.64) 3.41 (0.74) 3.53 (0.70)


Sad 3.15 (0.82) 2.81 (0.63) 2.98 (0.75)
Calm 3.69 (0.59) 3.29 (0.90) 3.49 (0.78)
Angry 3.19 (0.87) 2.56 (0.77) 2.87 (0.87)
Overall 3.41 (0.77) 3.02 (0.83) 3.21 (0.82)

Experiment 2 start making their ratings during the third presentation


of each excerpt and were given as much time as needed
The purpose of this experiment was to examine to complete their ratings before listening to the next
whether responses differed when participants were excerpt.
asked to rate the intensity of emotions conveyed by the
music as compared to when they were asked to rate the Results and Conclusions
intensity of emotions elicited by the same music.
Table 2 presents the mean ratings from the targeted
Method rating scale (i.e., happy, sad, calm, angry) for the
intended emotion of the musical excerpts. A MANOVA
PARTICIPANTS showed a significant effect for type of rating, Wilks
The participants were 44 American University students lambda = .16, F(3, 40) = 70.87, p < .05. Participants who
who received extra credit in psychology courses. None were asked What emotion does this music convey?
had participated in Experiment 1. gave higher ratings than those who were asked How
does the music make you feel? Thus, the participants
MATERIALS, DESIGN, AND PROCEDURE
were able to distinguish between the two questions.
The stimuli included a randomly selected 24 of the 32
As in Experiment 1, there was a significant effect for
excerpts from Experiment 1, with six of each emotion,
the intended emotion of the music, Wilks lambda =
and were presented five times in succession with no
.13, F(3, 40) = 88.47, p < .05; participants gave higher
pauses between the repetitions in order to immerse the
ratings to music with positive emotions than music
participants in the music and give an opportunity for
with negative emotions, Wilks lambda = .26, F(1, 42) =
the intended emotion to be evoked. Half of the partici-
117.66, p < .05. More pertinent to the present experiment
pants were asked to judge the emotion that was con-
was the lack of an interaction between the intended
veyed by the excerpts (What emotion does this music
emotion of the music and the type of rating made by
convey?) and the other half were asked to judge the
the participants, Wilks lambda = .97, F(3, 40) = 0.45,
emotion that was elicited by the same excerpts (How
p > .05. Even though the overall intensities were judged
does the music make you feel?). To this end, partici-
to be less when the task was to judge the emotion
pants rated each excerpt across each of four emotional
elicited by the melody, the difference between the pos-
domains (happy, sad, calm, and angry) using the fol-
itive and negative emotions was similar in both types
lowing 12 items from the Positive and Negative Affect
of rating tasks.
ScheduleExpanded Form (PANAS-X; Watson &
Clark, 1994): Happy, Joyful, Cheerful, Calm, Relaxed,
At-ease, Sad, Lonely, Blue, Angry, Hostile, and Scornful. Experiment 3
Thus, there were three PANAS-X items to gauge each
emotion. To conform to the standard PANAS ratings, The purpose of this experiment was to combine the
the ratings were given on five-point scales, where 1 methods of Experiments 1 and 2 and explore whether
represented very slightly or not at all (e.g., very slightly familiarity affected the intensity of emotional responses
or not at all happy) and 5 represented extremely as a function of the type of emotional rating (i.e.,
(e.g., extremely happy). Participants were asked to conveyed versus elicited emotion).
Music2703_02 1/11/10 3:38 PM Page 180

180 S. Omar Ali & Zehra F. Peynirciog lu

Method were asked How does the music make you feel?
Participants also gave higher ratings to music with pos-
PARTICIPANTS itive emotions than music with negative emotions,
The participants were 64 American University stu- Wilks lambda = .66, F(1, 62) = 31.83, p < .05. There was
dents who participated for extra credit in psychology also an interaction between type of rating and the mag-
courses. None had participated in either of the previ- nitude of the differences between positive and negative
ous experiments. emotions, Wilks lambda = .92, F(1, 62) = 5.33, p < .05.
Thus, unlike in Experiment 2, although the intensity
MATERIALS, DESIGN, AND PROCEDURE
differences between positive and negative emotions
The stimuli were the same 24 excerpts from Experiment 2,
emerged for both types of ratings, these differences were
but some were made more familiar than others. Half of
more pronounced when participants judged the emo-
the 24 excerpts (three from each emotion) were ran-
tion elicited by the music than that conveyed by the
domly selected and were presented five times in succes-
music, indicating a more heightened sensitivity to the
sion with no pauses between the repetitions (as in
emotions in the former condition.
Experiment 2); the other half were presented only once.
There was also a main effect for the familiarity con-
For the excerpts presented five times, participants
dition, Wilks lambda = .47, F(1, 62) = 69.18, p < .05.
began to make their ratings during the third presenta-
This time, unlike in Experiment 1, even though the rat-
tion of the excerpt (again, as in Experiment 2). For the
ings were entirely on emotion, familiar music was rated
excerpts presented once, participants made their rat-
higher on emotional intensity than unfamiliar music.
ings immediately after the presentation of the excerpt.
This effect did not vary as a function of type of rating,
The excerpts were interchanged so that, across two groups
Wilks lambda = .99, F(1, 62) = .68, p > .10.
of participants, each excerpt served in the familiar and
unfamiliar conditions equally often.
Discussion
Results and Conclusions
This study examined whether familiarity with musical
Table 3 presents the mean ratings from the targeted stimuli intensified or otherwise modified emotional
rating scale (i.e., happy, sad, calm, angry) for the responses, as it had been previously reported to do so
intended emotion of the musical excerpts. MANOVA with liking responses (e.g., Peretz et al., 1998).
showed significant effects for the type of rating, Wilks Experiment 1 suggested that it did not. The circumplex
lambda = .24, F(3, 60) = 63.33, p < .05, and the partic- model (North & Hargreaves, 1997) characterizes emo-
ular intended emotion of the excerpt, Wilks lambda = tions expressed in music by both their levels of like-
.25, F (3, 60) = 61.07, p < .05. As in Experiment 2, par- ability and arousal. Ritossa and Rickard (2004)
ticipants who were asked What emotion does the music investigated the relative utility of pleasantness and lik-
convey? gave higher ratings than the participants who ing in predicting the emotions expressed in familiar

TABLE 3. Mean Intensity Ratings (SD) of Emotional Responses to Musical Excerpts as a Function of Type of Emotion, Type of Rating, and the
Familiarity Condition in Experiment 3.

Type of Rating

Emotion Conveyed Emotion Elicited

Unfamiliar Unfamiliar
Emotion Familiar Music Music Overall Familiar Music Music Overall

Happy 3.84 (0.65) 3.72 (0.65) 3.78 (0.65) 3.52 (0.88) 3.19 (0.84) 3.36 (0.87)
Sad 3.53 (0.71) 3.18 (0.86) 3.36 (0.80) 2.74 (0.70) 2.41 (0.75) 2.57 (0.74)
Calm 3.86 (0.69) 3.41 (0.90) 3.63 (0.83) 3.77 (0.73) 3.15 (0.78) 3.45 (0.81)
Angry 3.50 (0.91) 3.10 (0.78) 3.30 (0.86) 2.36 (0.86) 2.03 (0.65) 2.19 (0.77)
Overall 3.68 (0.76) 3.35 (0.83) 3.52 (0.81) 3.10 (0.97) 2.69 (0.90) 2.90 (0.96)
Music2703_02 1/11/10 3:38 PM Page 181

Intensity of Emotions Conveyed and Elicited 181

music and found that pleasantness was a more useful liking depend on contributions from only one of the
predictor of emotions expressed by familiar music than axes and thus possibly require less effort or information
liking. Thus, Ritossa and Rickards results suggest that for discrimination, whereas judgments of emotion also
preference for familiar music may have both intellectual involve a contribution from the arousal axis, and thus
(i.e., liking) and emotional (i.e., pleasantness) com- make discrimination more complicated. Thus, it is pos-
ponents. In short, likeability is only one component of sible that the task in Experiment 1 might have enabled
emotion because it can indeed be devoid of arousal that limited amount of familiarity to affect liking but
and does not share all of the properties associated with not emotional intensity.
happiness. Experiments 2 and 3 demonstrated that emotions
Experiment 3, however, demonstrated that familiarity conveyed by music were judged to be more intense than
enhanced emotional intensity. Of the several procedural emotions elicited by the same music. However, further
differences between Experiments 1 and 3 (for example, evaluation of these findings is warranted given the
the PANAS-X scales were used in Experiment 3, but not methodological limitations of this study. For example,
in Experiment 1), two differences are most likely to the participants may not have been comfortable experi-
have contributed to the emergence of the effect in one encing strong emotions in a laboratory setting. The dif-
experiment but not the other. First, in Experiment 1, ferences in the intensity judgments may have been a
participants were given an extra task of trying to remem- function of the exposure time, such that the partici-
ber some of the music. This may have led to a divided- pants were able to classify the emotions conveyed, but
attention situation, thus reducing the processing of were unable to actually process the emotions elicited
other elements at the expense of a possibly more salient during the relatively short presentation time (five repe-
intellectual element (i.e., trying to remember dur- titions of each musical excerpt and 20 s per repetition,
ing the familiarization process). Similarly, although the for a total of two minutes for each excerpt). The pre-
experiment did not actually test for memory during the dictability achieved through the en masse repetitions in
main task of emotion ratings, the imposed memory Experiment 3 may have enhanced the participants
load may have led to the unintended byproduct of lim- intellectual judgment of what the music intended to
iting cognitive resources and thereby not allowing the convey but blunted the participants actual experience
full processing of the emotional intensity. Second, in of the emotion because of habituation. Thus, a single,
Experiment 3, familiar excerpts were played more times but longer presentation of the music may diminish the
and en masse and unfamiliar excerpts were played observed differences.
only once, and thus the familiarity level reached in Finally, in all of the experiments, the participants
Experiment 1 (in which the familiar excerpts were rated the music that conveyed positive emotions
played only three times in an interspersed fashion) may higher in intensity than the music that conveyed neg-
have been enough to cause liking but not enough to ative emotions (cf. Ali & Peynirciog lu, 2006); howev-
reach a threshold to increase emotional intensity. er, this difference was not observed in liking ratings.
Similarly, in Experiment 1, because each familiar It may be that, by chance, the music selected to depict
excerpt was interspersed with unfamiliar excerpts, the the positive emotions were intrinsically more effec-
participants may have built up a different type of rep- tive at evoking them than those selected to depict the
resentation of familiar excerpts in Experiment 1 com- negative emotions. However, these results are consis-
pared to Experiment 3. Although we did not test for tent with those of Balkwill and Thompson (1999),
liking in Experiment 3, it is possible, but unlikely, that who also found that a tendency for participants to
the liking responses of Experiment 1 might not have rate their intensity of joy and peace to be higher
been replicated, given the procedural differences in than that of anger.
manipulating familiarity of the musical excerpts. If that In summary, this study showed that with more expo-
were the case, one could even argue that emotional sure and a greater focus on emotion ratings, familiarity
intensity ratings were actually more independent of lik- enhances emotional intensity ratings, independent of
ing ratings and were influenced not by familiarity, but the valence of the emotion. This is consistent with the
by more of an in-depth experience as a result of famil- idea that knowledge of (or familiarity with) the correct
iarity. In any case, both of these factors (i.e., reduced expectations set up by the organization of the melody
processing and not enough familiarity) may have can lead to more anticipatory arousal, which can be
worked in a similar manner. According to the circum- interpreted as greater emotional intensity (Meyer, 1956).
plex model (North & Hargreaves, 1997), judgments of Lastly, this study suggested that emotions conveyed by
Music2703_02 1/11/10 3:38 PM Page 182

182 S. Omar Ali & Zehra F. Peynirciog lu

musical stimuli are judged to be more intense than The experiments in this study were presented at the
emotions elicited by the same stimuli. 46th Annual Meeting of the Psychonomic Society in
Toronto, Canada.
Author Note
Correspondence concerning this article should be
This study was supported in part by a dissertation research addressed to Zehra F. Peynirciog lu, Department of
award from the American Psychological Association and Psychology, American University, Asbury Building 321,
by a non-competitive research grant from the Graduate 4400 Massachusetts Ave., N.W., Washington, D.C.,
Leadership Council of American University. 20016. E-MAIL: peynir@american.edu

References
. .
A LI , S. O., & P EYN IRC IOG LU, Z. F. (2006). Songs and emotions: P FISTER , H. R., & B HM , G. (1992). The function of concrete
Are lyrics and melodies equal partners? Psychology of Music, emotions in rational decision making. Acta Psychologica, 80,
34, 511-534. 199-211.
B ALKWILL , L.-L., & T HOMPSON , W. F. (1999). A cross-cultural R ITOSSA , D. A., & R ICKARD, N. S. (2004). The relative utility
investigation of the perception of emotion in music: Psy- of pleasantness and liking dimensions in predicting the
chophysical and cultural cues. Music Perception, 17, 43-64. emotions expressed by music. Psychology of Music, 32, 5-22.
G ABRIELSSON , A. (2001/2002). Emotion perceived and emotion S AMSON , S., & P ERETZ , I. (2005). Effects of prior exposure on
felt: Same or different? Musicae Scientiae, Special Issue, 123-147. music liking and recognition in patients with temporal lobe
GAUDREAU, D., & PERETZ, I. (1999). Implicit and explicit memory lesions. Annals of the New York Academy of Sciences, 1060,
for music in old and young adults. Brain and Cognition, 40, 419-428.
126-129. S TEVENS , C., & L ATIMER , C. (1991). Judgments of complexity
H ARGREAVES , D. J. (1984). The effects of repetition on liking and pleasingness in music: The effects of structure, repetition,
for music. Journal of Research in Music Education, 32, 35-47. and training. Australian Journal of Psychology, 43, 17-22.
J USLIN , P. N., & S LOBODA , J. A. (2001). Music and emotion: STRATTON, V. N., & ZALANOWSKI, A. H. (1997). The relationship
Introduction. In P. N. Juslin & J. A. Sloboda (Eds.). Music between characteristic moods and most commonly listened
and emotion: Theory and research (pp. 3-19). Oxford: to types of music. Journal of Music Therapy, 34, 129-140.
Oxford University Press. S ZPUNAR , K. K., S CHELLENBERG , E. G., & P LINER , P. (2004).
K NOBLOCH , S., & Z ILLMANN , D. (2003). Appeal of love themes Liking and memory for musical stimuli as a function of
in popular music. Psychological Reports, 93, 653-658. exposure. Journal of Experimental Psychology: Learning,
M EYER , L. B. (1956). Emotion and meaning in music. Chicago, IL: Memory, and Cognition, 30, 370-381.
University of Chicago Press. TAN , S., S PACKMAN , M. P., & P EASLEE , C. L. (2006). The
N ORTH , A. C., & H ARGREAVES , D. J. (1997). Liking, arousal effects of repeated exposure on liking and judgements of
potential, and the emotions expressed by music. Scandinavian musical unity of intact and patchwork compositions. Music
Journal of Psychology, 38, 45-53. Perception, 23, 407-421.
P ERETZ , I., G AUDREAU, D., & B ONNEL , A. M. (1998). Exposure WATSON , D., & C LARK , L.A. (1994). The PANAS-X. Manual
effects on music preference and recognition. Memory and for the positive and negative affect scheduleExpanded form.
Cognition, 26, 884-902. Iowa City, IA: University of Iowa.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Você também pode gostar