Muslcae Scientiae C 2002 by E5COM European Society

Special issue 2001 ·2002, 123-147 for the Cognitive Sciences of Music

Emotion perceived and emotion felt:

same or different?

ALF GABRIELSSON
Department of Psychology, Uppsala University, Sweden

• ABSTRACT
A distinction is made between emotion perception, that is, to perceive emotional
expression in music without necessarily being affected oneself, and emotion
induction, that is, listeners' emotional response to music. This distinction is not
always observed, neither in everyday conversation about emotions, nor in
scientific papers. Empirical studies of emotion perception are briefly reviewed with
regard to listener agreement concerning expressed emotions, followed by a
selective review of empirical studies on emotional response to music. Possible
relationships between emotion perception and emotional response are discussed
and exemplified: positive relationship, negative relationship, no systematic
relationship and no relationship. It is emphasised that both emotion perception
and, especially, emotional response are dependent on an interplay between
musical, personal, and situational factors. Some methodological questions and
suggestions for further research are discussed.

1
INTRODUCTION

The relationship between emorion perceprion and emorion induction in music -

in other words, how the emotional expression in a piece of music relates to the
listener's emotional response - has not received much attention in empirical music
psychology. I will first comment upon me distinction between emotion perception
and emotion induction, then summarise results from selected research in both areas
in order to subsequently analyse what types of relationship may exist between
perceived and induced emotion.
First a note on terminology. There is no unanimous agreement, neither in
everyday language nor in psychological terminology, on the definicion of terms like
feeling, emotion, mood, and affect, or their correspondences in other languages; see
Sloboda and Juslin (2001) and Scherer and Zentner (2001) for general discussions
of this question, Not to get trapped in this terminological confusion already from
the beginning, I will use "emotion" and "feeling" in a generic and broad-minded
123
sense, often involving cognitive components; "mood" and "affect" will be used when
employed by authors referred to in the text.

2
DISTINCTION BETWEEN EMOTION PERCEPTION AND EMOTION INDUCTION

Listeners' perception of emotional expression - to perceive an expression of, say,

sadness in the music without necessarily being affected oneself - is mainly a
perceptual-cognitive process. It should be distinguished from listeners' emotional
response to the music, for example, to feel sad. In the former case the music is an
object for perception and reflection, in the latter it has caused an emotional reaction.
This distinction is not always clearly observed, neither in everyday conversation
about emotions, nor in scientific papers.
In reality, the border between the two alternatives is somewhat blurred. We may
think of them as opposite extremes on a continuum from "pure", emotion-free
perception at the one end to intense emotional reaction at the other end. Rather
than being at any of these extremes, in most situations listeners are probably
somewhere along this continuum, depending on many circumstances. Adopting an
"emotion-free" attitude may be a relevant part of music critics' or music theorists'
working in general, including "objective" descriptions of the emotional character of
the music and the musical means used to achieve this character. On the other hand,
extremely intense emotional reactions to music in so-called "peak" experiences
(Maslow, 1976) or strong experiences related to music (Gabrielsson, 200 1;
Gabrielsson and Lindstrom Wik, 2000) occur relatively rarely according to the
respondents' own statements.
A related and frequently discussed question concerns so-called "listener types" or
rather "listening types", that is, what artirude to the music the listener adopts; see
reviewsin Farnsworth (1969, p. 131),]0rgensen (1988, p. 35, 64), or Mursell (1937,
p. 218). Commonly discussed alternatives are (a) objective or analytical listening
focusing on various "objective" properties of the music (e.g., instrumentation,
formal construction); (b) associative listening, relating the music to extra-musical
phenomena (events, persons, nature, etc); (c) listening for the emotional expression
in the music (happy, sad, excited, etc); (d) focusing on one's own emotional
response, or on (e) physical responses to the music. Several studies (Bartel, 1992;
Hargreaves, 1986, p. 126; Hargreaves and Colman, 1981; Hedden, 1973; Pike,
1972; Weld, 1912) indicate that emotional responses and intellectual (objective,
analytical) responses tend not to occur together, and that musically experienced
listeners are more likely to focus on intellectual features, whereas less musically
experienced listeners tend to focus on emotional properties. However, this may
also reflect the fact that musicians have acquired a terminology for "objective"
descriptions of music not availableto non-musicians and further that musicians may
be subject to demand characteristics vs- feeling that they are expected to describe
124
 Emotion perceived and emotion felt
Al.F GABRIELSSON

their response in music-technical terms). Furthermore, the above results mainly refer
ro musicians within classical music; the situation may be different considering other
musical genres.
Lehmann (1997), using five emotion-related items from a listening inventory,
found stronger self-reported affective reactions to music in music-majors than in
non-music majors and claimed that formal training in music may foster a music-
specific, cognitively mediated emotionality (training-mediation hypothesis).
Lehmann also suggested that personally salient features of the music may elicit
heightened emotional response; an example may be the present author's heightened
emotional response to pieces of music that he used in various studies (e.g.,
Gabrielsson, 1973). Moreover, musicians report as strong emotional reactions to
music as non-musicians when asked about the strongest experience of music that
they have ever had (Gabrielsson, 2001). There is thus no simple answer regarding
comparison between musicians' and non-musicians' emotional response to music.

3
EMOTION PERCEPTION

The earliest empirical studies on emotion perception in music appeared already in

the 1890s, but it was not until the 1930s that studies became more frequent. The
purpose was to investigate to what extent listeners agreed on perceived emotional
expression, and/or to investigate what factors in the musical structure influenced
perceived expression. Stimuli were usually selected pieces of recorded music (e.g.,
Watson, 1942; Wedin, 1969) or other tonal stimuli (e.g., Scherer and Oshinsky,
1977). A few studies have applied systematic experimental manipulation of various
structural properties, for instance, by presenting the "same" piece of music in
different tempi, different modes, or different pitch levels (e.g., Hevner, 1936; Juslin,
1997; Rigg, 1940). In still other, more recent studies the stimuli have been pieces of
music performed in order to express various emotions, for instance, one and the
same tune intentionally performed to sound happy, sad, angry, or fearful (e.g.,
Gabrielsson and Juslin, 1996; [uslin, 2000).
Listeners reported perceived emotional expression using one of the following
alternatives: (a) free phenomenological descriptions, (b) choice among descriptive
terms, adjectives or nouns, provided by the investigator, (c) ratings of how well such
descriptive terms applied to the music in question, or, recently, (d) continuous
judgement during the course of the music (see Schubert, this issue).
Some 75 empirical investigations on emotion perception in music were
extensively reviewed with regard to stimuli, subjects, response type, and descriptive
attributes in Gabrielsson and Juslin (in press) and Gabrielsson and Lindstrom
(2001). Briefly, already in studies 1935-1955 (Campbell, 1942; Capurso, 1952;
Gundlach, 1935; Hampton, 1945; Rigg, 1937; Sopchak, 1955; Watson, 1942)
good listener agreement was found for various expressions of positive emotions (e.g.,
125
gaiety, happiness, joyfulness, triumph) and negative emotions kg., grief,
melancholy, mournfulness, sadness, despair), furthermore for expressions usually
associated with high arousal (e.g., agitated, angry, restless, violent) and low arousal
(e.g., gentle, relaxing, soothing, tranquil). Results in later studies using multivariate
analysis techniques - factor analysis, cluster analysis, correspondence analysis,
multidimensional scaling (Gabrielsson, 1973; Imberry, 1979; Kleinen, 1968;
Nielzen and Cesarec, 1981; Wedin, 1969, 1972) - suggested that many, perhaps
most, emotions recognised in music may be represented in a two-dimensional space
with valence (positive Vi. negative feelings) and arousal (high - low) as principal axes,
a result in principal agreement with the "circurnplex" model of emotion (Russell,
1980). This model has also been used as basis for recent proposals of continuous
recording of emotional expression (Madsen, 1997; Schubert, 1999, 2001).
Wedin (1969, 1972) also found a dimension labelled solemnity Vi. triviality but
suggested that it mainly reflected a stylistic dimension: solemnity is apt to
characterise certain kinds of art music but is rarely applicable to popular music.
Other feelings recognised in various studies, such as love, yearning, nostalgia,
reverence, assertion, determination, weirdness, and grotesqueness may be represented
as various combinations of valence and arousal. Certain confusions among emotions
were observed, for instance, yearning may be confused with sorrow, calm, and
tenderness; calm with tenderness and sorrow; tenderness with sorrow; and peace
with sorrow. All of them reflect low arousal, and the musical means to achieve these
expressions seem similar, for instance, slow tempo, soft or moderate loudness, and
legato articulation. Among so-called "basic" emotions, happiness and sadness were
usually better recognised than anger and fear. Good agreement may exist regarding
the basic emotional quality (e.g., sadness), but less regarding nuances or variants of
this quality (e.g., funereal, romantic, poignant, or reflective sadness); the latter may
require good knowledge of the musical style (Brown, 1981). Finally it may be
noticed that listener agreement was worse regarding emotions such as cruelty,
devotion, disgust, hate, horror, jealousy, pity, and worship.

4
EMOTION INDUCTION

There is no single comprehensive review of studies on emotional responses to music,

but different approaches to this subject are discussed in contributions to the recent
volume edited by Juslin and Sloboda (2001), such as music as a technology for
emotion construction (DeNora, 2001), production rules for emotional effects of
music (Scherer and Zentner, 2001), emotions in everyday listening to music
(Sloboda and O'Neill, 2001), and emotions in strong experiences with music
(Gabrielsson, 2001).
Investigating emotional responses to music meets with many methodological
problems (Scherer and Zen mer, 2001 j Sloboda and Juslin, 2001 j see also
126
 Emotion perceived and emotion felt
ALFGABRlEl.SSON

Gabrielsson and ]uslin, in press). Verbal report of felt emotion is the most common
approach but is sensitive to several biases such as inter-individual differences in
vocabulary and linguistic capacity, various demand characteristics, regard to social
desirability, and the like. Furthermore. listeners may confuse perceived emotion in
the music wirh whar they actually feel themselves. In fact, asking listeners to report
on their emotional reaction may severely interfere with the reaction itself, because
the listener must alternate between two attitudes: feeling and observing feeling; as
noticed earlier, emorional and intellectual responses tend not to occur together.
Finally, many listeners find words insufficient to describe their experience
(Gabrielsson, 2001; Pike, 1972).
Still, verbal report is the only way of accessing the subjective emotional
experience. It cannot be replaced by any other measures, but should preferably
be supplemented by other measures in a multi-method approach. Scherer and
Zentner (2001) regard emotion as a multi-componential phenomenon including
subjective feeling, physiological reactions, motor expressive behaviour, possibly also
motivational (action tendencies) and cognitive (appraisal) components, and suggest
that all these aspects should be investigated in combination. No investigation so far
has come close ro this requirement and "it is not to be denied that it is a formidable
undertaking" (Scherer and Zentner, 2001, p. 384). Of course, the idea of a multi-
method approach is not new, but so far the results of studies in which emotional
responses to music have been studied by verbal reporr in combinarion with some
other rneasurets) are still very limited, as described in the following.

PHYSIOLOGICAL AND BEHAVIOURAL MEASURES

Recordings of physiological responses to music, assumed to be related to experienced
emotions, started already in the late 19th century, but extensive reviews (Bartlett,
19%; Hodges, 1980; see also Radocy and Boyle, 1997, and Scherer and Zentner,
2001) show that the relationships between physiological variables - heart rate,
blood pressure, respiration rate, electrodermal response, muscular tension, etc -
and experienced emotion in musical contexts largely remain obscure. A few recent
studies provide more positive evidence. Krumhansl (1997) recorded physiological
variables covering cardiac, vascular, electrodermal, and respiratory functions on a
second-to-second basis while subjects were listening to music representing
happiness, sadness, and fear. Immediately afterwards subjects rated how they felt
while listening to the music using 13 emotion adjectives (e.g., happy, sad, afraid,
angry, surprised). Correlations between physiological measures (the average for
each subject on each measure) and self-reported emotion were very low, usually
r <.20 (positive or negative) and non-significant. The low correlations may be due
to the averaging of physiological measures across the whole musical excerpt.
Somewhat higher correlations, up to abour r = .40, were found when rhe
continuously recorded physiological measures were related ro continuous ratings of
felt emotion (however, the latter made by another group of lisreners).

127
Nyklicek, Thayer, and van Doornen (1997) recorded a broad range of
cardiorespiratory variables in subjects listening to music pieces representing
happiness, sadness, agitation, and serenity. Although individual physiological
variables usually did not differentiate between the reported emotions, discriminant
analysis using all of the cardiorespiratory variablesachieved substantial discrimination
between the four emotion conditions. The first discriminant function, accounting
for 62.5% of the variance, differentiated between high arousal emotions (happiness,
agitation) and low arousal emotions (sadness, serenity) and was largely associated
with respiratory parameters. The second function could not be clearly interpreted,
but the third function distinguished between positive (happiness, serenity) and
negative (sadness, agitation) emotions; however, it accounted for only 10 percent of
the variance and its physiological background was less distinct. The results of this
study showed similarities as well as dissimilarities with results using other induction
techniques (e.g., voluntary facial action, imagery), and it was also suggested
that individual differences in physiological patterning during emotions may be
considerable, which indicates the need for idiographic studies.
Beside autonomic reactions, changes in facial muscle activity during music
listening were reported by Lundqvist et al. (2000) and Witvliet and Vrana (1996;
cited in Scherer and Zentner, 2001, p. 376). Listening to music with negative
valence was associated with increased corrugator activity (as in frowning), whereas
music with positive valence elicited larger zygomatic activity (as in smiling). This
borders on measures of motor expressive behaviour. However, although the
connection between music and movement is generally acknowledged, there is yet
lirrle hard empirical evidence for rhe producrion of moror patterns and action
tendencies by different kinds of music as reviewed by Scherer and Zentner (2001).

VERBAL DESCRIPTIONS
Verbal descriptions, in real time or in retrospect, remain the best and most natural
method to study emotional responses ro music. A common procedure is to use some
kind of rating scales referring ro various emotions, for instance, the ones used by
Krurnhansl (1997): Afraid, Amused, Angry, Anxious, Contemptuous, Contented,
Disgusted, Embarrassed, Happy, Interested, Relieved, Sad, Surprised. The selection
of scales varies among different investigations depending on the purpose, any
underlying theory, and other circumstances. Ratings are easily subjected ro common
methods for statistical description and inference. However, the limitation of
experimenter-selected scales could mean that other important aspects of subjects'
emotional responses remain unknown.
The opposite alternative would be ro have the listeners give a completely free
description of the experience. This has been tried in some studies. Weld (1912)
asked his subjects to provide detailed introspections and analysed them with regard
ro imagery, actual or imaged moror reactions, and emotions and moods. Concerning
emotions/moods the most used descriptive terms were "happy, gay, lively, joyous,

1:18
 Emotion perceived and emotion felt
ALF GAIlRIELSSON

cheerful, exhilarating, playful, restful, soothing, tender, quiet, peaceful, coquettish,

triumphant, sad, sombre, yearning, serious, longing, disquieting, restless, unhappy,
weird, pensive, mournful, despairing, cheerless" (ibid., p. 283). There were thus
positive emotions (happy, gay etc until triumphant) as well as negative emotions
(sad, sombre etc until cheerless). Within the positive emotions one may also see a
division into those reflecting relativelyhigh arousal (happy, gay etc until playful) and
low arousal (restful, soothing etc until peaceful). He concluded that "the feelings
arrange themselves between two pairs of poles in two dimensions of space. One of
these pairs can best be described as pleasantness-unpleasantness [...J the most
appropriate terms for the other pair seem to be excitement-repose" (ibid., p. 282).
The similarity to the valence-arousal model is apparent, and Weld himself referred
to Wundt's three-dimensional model of emotion (Wundt, 1905, pp. 99-100), in
which two of the dimensions were, in Weld's translation, pleasantness-unpleasantness
(German: Lust-Unlust) and excitement-depression (German: Erregung-Beruhigung).
Pike (1972) asked up to 200 hundred students to listen to many pieces
of instrumental or symphonic music and to describe their emotional responses
immediately after each piece. All descriptions were listed and compared, and
descriptions with similar meaning were gathered into the same category, for
instance, "feeling of exuberance", "feeling of joyfulness", and "feeling of ecstasy"
were all categorised as "feelings of pleasure". The results were summarised as five
"constituents of common experiential factors in music experience": "Feeling of
pleasure" (expressed by 96 percent of the subjects), "Feeling of movement" (65%),
"Feeling of oneness with the music" (83%), "Perception of spontaneous and
transient emotional states" (72%), and "Perception of stabile moods" (86%).
Regarding spontaneous and transient emotional states he briefly commented that
"The music builds up expectations, satisfies them, inhibits them, creates suspense,
surprise, doubt, uncertainty, momentary unpleasantness, pleasantness, tension-
release, excitement and so forth" (ibid., p. 265-266), and stabile moods were such as
"joy, sadness, nostalgia, and the like" (ibid., p 266). The experience was, on the
whole, pleasurable. Unfortunately, Pike's reporr of his extensive material was very
short and condensed; it would certainly have deserved a much broader exposition.
Maslow (1976, pp. 169-170) found music, especially the great classics, to be
one of the easiest ways, alongside of sex, to have a peak experience. His material
consisted of about 320 persons' descriptions of "the most wonderful experience of
your life" (Maslow, 1968, p. 71). There is no data on how many of these that were
connected with music, but as music was a frequent trigger they should be many.
Emotional aspects of generalised peak experience included complete loss of fear,
anxiety, inhibition, defence, and control. Moreover, "the emotional reaction in the
peak experience has a special flavor of wonder, ofawe, of reverence, ofhumility and
surrmder before the experience as before something great" (ibid, pp. 87-88), even a
fear of being overwhelmed of more than what one can bear. The experience may
occasionally be described as sacred.

129
 Working along similar lines, we (Gabrielsson, 200 1; Gabrielsson and Lindstrom,
1993, 1995; Gabrielsson and Lindstrom Wik, 2000) have obtained and analysed
hundreds of reports on people's strongest experiences of and with music. The analysis
of these reports has resulted in a descriptive system with seven main categories -
General characteristics, Physical Reactions and Behaviours, Perception, Cognition,
Feelings/Emotions, Existential and Transcendental Aspects, and Personal and Social
Aspects - each of them with further sub-divisions. Feelings/Emotions are primarily
divided into positive emotions and negative emotions; a listing of them appeared in
Gabrielsson (2001). Positive emotions extend from low arousal instances like peace,
calm, safety, and warmth to high arousal in euphoria and ecstasy. In between these
extremes, there are many other feelings, such as feeling moved, delight, happiness,
and elation. Negative feelings, although on the whole rare in strong experiences with
music, likewise range from low or neutral arousal as in nervousness and frustration
to high arousal as in horror and panic.
As the intensity of the experience may vary during the course of the music,
procedures for continuous indication of the emotional response are commendable.
Lowis (1998) had listeners use a button to indicate "a moment of particularly
deep and profound pleasure or joy [...] the son that produces a tingle in the
spine" (p, 212). Significant correlations were found between frequency of button
pressings and ratings for the following feelings evoked by the music: joy, love/
tenderness, longing, sadness, reverence/spirituality, action, and memory/
thoughtfulness. Waterman (1996) and Sloboda (1999) used partly similar methods
for obtaining listeners' continuous and immediate indication of emotionally loaded
moments in the music, thereby also increasing the possibilities of relating them to
properties of the music in question.
In sum, verbal reports of musically elicited responses demonstrate a large variety
of feelings. Many or even most of them are also found in studies of emotion
perception (see Section 3), but there are also emotional responses which seem to
have no correspondence in emotion perception. This will be further discussed
(Section 5.4) when we now turn to the primary question of this article, that is, what
types of relationship there may exist between emotion perception and emotion
induction in musical contexts.

5
RELATIONSHIP BETWEEN EMOTION PERCEPTION AND EMOTION INDUCTION

From the preceding it should be clear that a study of the relationship between
emotion perception and emotion induction can presently be based only upon
verbal reports. The relationship may take various forms: positive, negative, no
systematic relationship, or no relationship at all (Figure 1).

130
 Emotion perceived and emotion felt
AU GABRIELSSON

EMOTION
PERCEIVED INDUCED

Positive
relation

Happy Happy
Sad Sad

Negative
relation

Happy Sad
Sad Happy

No systematic
relation
(two cases)

Happy
"Neutral"
Sad

Varies for dif-

Happy ferent listeners
and occasions

No relation

Not perceived (but) Induced

Figure 7. Schematic illustration of four types of relationship between perceived emotion

and induced emotion. For simplicity only happy and sad are used as examples;
they may be replaced by any other emotion(s). See further information in the text

POSITIVE RELATIONSHIP

• Some historical notes.The by far most common idea is that there reigns a positive
relationship between emotion perception and emotion induction, that is, that the
listener's emotional response is in agreement with the emotional expression in the
music (Figure 1). Some historical examples easily come to mind. In Ancient Greece,
music was considered a form of mimesis, that is, imitation of nature, in particular
imitation of human character or states of mind; it was further supposed that this
character would be automatically imitated by the listener (Scruton, 1997, P: 118).
By careful selection of appropriate music one could therefore bring about desired
virtues or states of mind. Different modes were ascribed different ethical characters.
131
For instance, the Dorian mode was considered manly, strong, moulding a firm
character, and Plato thought that it embodied male determination which he
considered as a political ideal.
During the Baroque era the Doctrine of the Affictions, mostly associated with
German music, meant that music should express affects in the sense of idealised
emotional states and have listeners feel these states (Buelow, 1983, p. 396). The
composer and the performer were supposed to have own experience and a precise
conception of the affects in order to be able to express and arouse them adequately.
However, advice concerning what musical means should be used for expression
and induction of different affects was often rather vague, for example, use of large
intervals to represent joy, small intervals to represent sadness, ascending motion for
pride bur descending motion for humility, or disordered sequences of notes for
despair.
Music has been and is still frequently used in religious or political contexts,
usually in combination with text, to express and induce certain wanted feelings:
religious feelings, love in God, patriotism, enthusiasm for socialism, etc (Conversely,
music may be forbidden because its expressive and inductive effects are feared, for
instance in certain religious sects or by the former Talibans in Afghanistan.) In war
time, music is often used to express and induce feelings of power, superiority, and
invincibility. An opposite peaceful practice is singing of lullabies to have children
sleep (see Trehub and Nakati, this issue).

• Empirical investigations. Empirical investigations on the relation between

emotion perception and emotion response are scarce. Sharin (1970) presented
listeners a number of 2 minutes musical excerpts considered to express a successive
change from one emotion to another: from sad to gay, from restless to serene, from
bored to active, and from active to majestic. For each example listeners were asked
to describe "the way you felt emotionally from the beginning until the end of the
excerpts" by choosing among these four alternatives plus an alternative that there
was no special effect. Of the 76 male listeners, a majority (61-87 percent) indicated
the "correct" alternative for three of them, most convincingly for restless-to-serene;
the sad-to-gay alternative was indicated by 47 percent, and the no-special-effect
option was chosen by 5-20 percent of the listeners. These results suggested that
listeners felt in accordance with the expression in the respective pieces of music, bur
a
a relevant criticism 0rgensen, 1988) is that the design of the study may have given
the subjects the impression that the "correct" answer should be sought among the
four alternatives; in other words, an example of demand characteristics (Neale and
Liebert, 1986; see also Viistfjall, this issue). A related risk is that subjects, despite the
wordings in the instructions, in fact reponed the recognised emotional expression
rather than how they felt themselves, or mixed these twO alternatives.
Swanwick 0973, 1975) used a number of bipolar scales (e.g., calm-excited,
active-passive) on which his listeners should indicate both the perceived emotion in

U2
 Emotion perceived and emotion felt
ALF GABRIELSSON

the music and how they felt themselves, an interesting proposal rarely used in music
psychology research. However, Swanwick's musical stimuli were extremely simple,
such as repeated two tone patterns, and therefore of limited ecological validity.
A similar approach was taken by Gembris (1982), using eight pieces of music as
stimuli. His subjects were asked to rate, using a four-step scale (0 to 3), on the one
hand "properties of the music" (e.g., sadness) and on the other hand "how this music
affects myself" (e.g., makes me feel sad). A difference score between these two ratings
was calculated for each case. Zero difference meant that there was a complete
agreement between the musical expression (e.g., sadness) and the experienced
emotion (e.g., feeling sad), whereas the maximum difference score (3) would mean
complete disagreement. Data also allowed calculation of correlations between the
two aspects. However, results were given only for a few illustrative cases.
In a recent study by Zentner, Meylan, and Scherer (2000; see also Scherer and
Zentner, 200 1, p. 380) participants were asked to indicate, referring to their
preferred musical genre (classical, jazz, pop/rock, latin, or techno), which affecrive
rerms our of 150 such terms would apply to, on the one hand, emotions expressed
in the music, and on the other hand emotions aroused by the music. Elimination of
rarely used terms and facror analysis on the remaining terms resulted in 11 reliable
emotion scales common for both conditions. Across genres, negative emotions like
sadness, anger, and anxiety were considered more frequently expressed rhan aroused,
whereas peacefulness seemed more frequently aroused than expressed. The pattern
differed between different genres, for instance. while thrills were more induced than
expressed for classical music and jazz, no such difference appeared for latin music
or techno. It was emphasised, as in the previous studies, that it is necessary to have
subjects carefully observe the distinction between expression and induction when
conducting experiments on emotions in music.

• Musical mood induction. Studies on mood induction by means of music generally

rely on an assumed positive relationship between the emotional expression in the
music and the listener's emotional response. As mood induction is treated by
Vastfjall (this issue), only a few comments are given here with special regard to
people's private use of music in order to affect their feelings. In a study by
Sloboda (1992) music listeners frequently reponed that they used music to change,
intensify, or release existing emotions. This was as well amply demonstrated, in
particular for women, in later studies by Sloboda (1999), Sloboda and O'Neill (2001),
and especially by DeNora (2000, 2001). In our studies of strong experiences related
to music, there are many examples to the same end, such as this one by a young
woman:

Sometimes 1 want to confirm or strengthen a feeling. For instance, if I fed a little sad or
melancholic - then I wanr to play such music. I know which tunes contain this, touch that
feeling. Or if I come back home and am super-happy, I choose an inspiring tune that

133
 makes everything even happier. I can also do the reverse - change my mood wirh the
help of music. I may be becoming a little sad - then I can decide that: No, now I must
put on "this and that" ro make me happier and get some confidence again. And it can
really work as a flip of the finger. I get quite incredibly affected by music [...] it is good,
it is like therapy, It is a good trick when you learn how it works. It is nice, It is a resource.

Furthermore, studies of young peoples' motives for using music (Roe, 1985;
Zillman and Gan, 1997, p. 167) show that the reasons adolescents give for listening
to music are, among others, to relieve tension, set the right mood, make one feel less
alone, etc. Car drivers may compile cassettes with selected pieces of music to affect
their feelings during driving (Oblad, 2000, pp. 141-142) .

• Practices assuming positive relationship. There are practices based upon the
(often implicit) assumption that there is a positive relationship between the
emotional expression in the music and the listener's emotional response. For
instance, in music therapy the so-called "iso principle" (Altschuler, 1948; see also
Shatin, 1970) suggests that one should start with music of the same (iso) emotional
expression as the client's present mood and then successively change the expression
towards the mood that one wants to attain. In several music therapy practices, there
are lists of pieces of music assumed suitable for affecting mood in desired ways, for
instance, in Capurso (1952), and Bonny and Savary (1990). The wide-spread use
of background music in shops, restaurants, waiting rooms, aeroplanes, buses,
and telephone queues is another example of the assumption that the emotional
expression of the music will affect listeners' feelings and behaviours in the same
direction. Results sometimes seem to be reflect such straightforward effects, for
instance, that slow music, in comparison with fast music, gave slower in-store
traffic flow and higher sales volume (Milliman, 1982) and longer time spent eating
in restaurants (Milliman, 1986).
However, both regarding music therapy and background music many studies
demonstrate that the assumed positive relationship between expression and response
is simplistic and that several other factors must be taken into account, such as
clients' or customers' age, gender, familiarity with and attitude to music, present
physical and psychological state, degree of involvement (high or low), and their
conception of how. well music "fits" various situations; see reviews in Nonh and
Hargreaves (1997), Radocy and Boyle (1997), and Wigram, Sapersron, and
West (1995).

NEGATIVE RELATIONSHIP
This means that the listener reacts with an emotion "opposite" to that expressed in
the music: positive emotion in the music elicits negative emotion in the response or
negative emotion in the music elicits positive emotion in the response (see Figure 1).
A common and easily understood situation is that the music in the past has been

1::14
 Emotion perceived and emotion felt
Al.F GABRlELSSON

associated with an event of the opposite emotional quality. This is an authentic

example told by a lady referring to an experience during her childhood:

I was playing one of my favourite tunes on our piano - sweet melody and lyrics about a
tiny flower. Then it suddenly knocked on the door and my aunt came in, dressed in black,
and told that my granduncle passed away last night. The lovely atmosphere was abruptly
broken, it became cold and dark, even gruesome and unreal. Arrer this event the tune was
no longer a sunshine rune for me, it became associated with grief and tears.

The happy piece became associated with a sad event. The reverse case - a sad
piece associated with a happy situation - may occur as well. Of course, such
associations are usually wholly idiosyncratic. The principle of association is of course
neutral in relation to what emotions happen to be involved. While the example
above is of a negative relationship between expression and reaction, an association
may as well reflect a positive relationship - happy music associated with a happy
event, sad music associated with a sad event - or any other possible relationship,
for instance, happy music associated with a fearful event.
A related kind of association is when exciting or aggressive, "high arousal" music
makes the listener feel calm and safe. Sometimes aggressive hard rock, which
perhaps drives parents crazy, can make their rebellious youngster feel calm and
relaxed. To him or her this music is familiar and associated with his or her identity
and belonging to youth culture in general, or to any of its sub-cultures with attitudes
and feelings often in opposition ro parents' or other authorities' views.
A sad piece of music can make the listener happy because the sadness is so
convincingly and beautifully expressed in the composition or in the performance or
in both, for instance, in listening to Dido's Eareuell in Purcell's opera Dido and
Aen~aJ or to Kathleen Ferrier singing Mahler's Kindertotenlieder - a kind of
aesthetic-intellectual enjoyment mentioned already by Weld (1912). On the other
hand, a pleasant piece of music may be so poorly performed that it makes the
listener sad or even angry.
Background music may call forth negative reactions because the listener has
a generally negative attitude to the use of music as "sound tapestry". And to hear
one's favourite piece of music in a commercial for a soap may evoke disgust. If the
expression in a piece of music is nor considered typical or appropriate for the actual
situation - for example, a vulgar rune at a funeral - it may cause negative emo-
tion. For further examples, see North and Hargreaves (l997a, 1997b).
Finally another authentic example from a student demonstrates how attempts to
manipulate one's mood can make things worse:

If I am in a depressed mood, I may try to encourage myself by listening to happy music.

I really want to feel as happy as the music tells. But if I can't do it, it makes me even more
depressed than before.

135
No SYSTEMATIC RELATIONSHIP
There are occasions in which a potential relationship, positive or negative, between
expression and response does not occur. In statistical terminology, this alternative
may be seen as an example of "zero correlation" between expression and response
(see Figure 1). Two cases may be distinguished:
(a) Whatever emotional expression there is in the music, the listener Stays in the
same mood (whatever it is) or stays "emotionally neutral". This may be typified by
deliberate analytical listening, in this context focusing on what emotion is expressed
in the music and by what musical means this is achieved. Analytical listening is
fairly common among musically experienced listeners and among music theorists
and music critics as part of their work. An analytical attitude may, at leasr in
principle, also be requested from listeners in experiments where they are asked to
judge the emotional expression of different pieces of music.
(b) "Zero correlation" between expression and response may also occur in the
opposite case, that is, one and the same emotional expression in the music evokes
different emotional responses in different listeners, or in the same listener, or on
different occasions. Presumably any listener can remember that a piece of music has
been experienced differently on different occasions, depending on various personal
or situational factors. Several respondents in our study of strong experiences
(Cabrielsson, 2001) remarked that the strong experience did not at all appear at later
listening to the same piece, for some not even at an almost immediately repeated
listening.

No RELATIONSHIP
This alternative means, strictly interpreted, that there is not even a potential
relationship between variables in the expression and variables in the response
(Figure 1). Though it seems that most emotions discussed in research on emotion
perception (Section 3) also may appear in studies of emotion induction (Section 4),
some emotional responses have no obvious counterparts in musical expression. For
instance, a common response to music is to feel "moved" - but one can hardly
imagine how an expression of "feeling moved" would sound in music (see also
Scherer and Zentner, 200 I, p. 384, and Scherer et al., this issue, for some comments
on "being moved"). Likewise many other induced feelings - feel safety, warmth,
consolation, gratitude, bliss, fascination, wonder, awe, reverence, humility,
sacredness, or feel overwhelmed - are experiences which hardly can be expressed in
music, at least not with high degree of listener agreement. We may also note that
listener agreement was poor for emotions such as devotion, disgust, jealousy, pity,
and worship (Section 3).
The reason for this is fairly obvious. How we experience/respond to music is
not only a function of musical properties bur also of listener properties (earlier
experiences, expectations, attention, preferences, anitudes, personality, present
physical and psychological state, etc) and of the situation (environment, acoustics,

136
 Emotion perceived and emotion felt
ALPGABRlELSSON

time of day, being alone, in company, etc). There is always an interplay between
factors related to the music, the listener, and the situation in ever changing and often
unpredictable combinations. In many cases, personal and situational factors
may very well override musical factors, for instance, memories associated with
an emotion; see examples under Negative relationship (Section 5.2), further in
DeNora (2000, 2001), Gabrielsson (2001), and Sloboda and O'Neill (2001).
Emotion induction in musical contexts cannot be limited to musical induction,
induction results from personal and situational factors as well.
Of course, the interplay between music, person, and situation is present also in
listener perception of emotional expression. However, the relative importance
of musical, personal, and situational factors may be different in comparison with
emotion induction. In experiments where people are instructed to judge emotional
expression in music, they (should) focus on the music as an object and are expected
to suppress the influence of personal and situational factors. This is also typical for
music analysis and, at least partly, for music criticism. When it comes to one's own
experiences and reactions, however, personal and situational factors have free access
and may have as much, or even more, influence on our response than the music
itself.
Another, supplementary approach is to consider ideas about what emotions can
be expressed, and not expressed, in music. Hanslick in his classic l10m musikalisch
Schiinen (1854/1989) claimed that "Der Inhalt der Musik sind ronend bewegre
Formen" (p, 59), that is, the contents of music are "sounding forms in motion".
Music can never describe the feelings per se, only reproduce the motion associated
with physical events such as rapid, slow, strong, soft, rising, or falling. But,
emphasised Hanslick, motion is only a concomitant of feeling, not the feeling itself
It is generally accepted that we perceive motion qualities in music (see also Juslin,
Friberg and Bresin, this issue), and that motion qualities have emotional connotations:
"motion is heard in music, and that motion presents emotion characteristics much
as do the movements giving a person her bearing or gait" (Davies, 1994, p. 229).
Davies continues: "the expressiveness of music depends mainly on a resemblance
we perceive between the dynamic character of music and human movement, gait,
. bearing or carriage (ibid., P: 229), and "the range of emotions music is heard as
presenting [...] is restricted [...] to those emotions or moods having characteristic
behavioural expressions" (ibid., p.239). It is, indeed, hard to find dear behavioural
expressions of such emotions that were given as examples of emotions not expressed
in music - safety, warmth, consolation, gratitude, bliss, fascination, wonder, awe,
reverence, humility, devotion, disgust, jealousy, pity, worship, sacredness, feeling
overwhelmed - but which may very well appear in listener response due to personal
and situational factors.
A slight reservation could be made. It can be argued that even if feelings like
those just mentioned are not specifically perceivable in music, their general valence
and/or arousal may still be perceived and possibly felt and in that limited sense

137
represent a positive relationship between perceived and felt emotion. For instance,
perceived positive valence and low arousal might induce a feeling of safety. However,
it might as well induce feelings of warmth, consolation, gratitude, wonder,
reverence, humility, devotion, worship, sacredness or still other alternatives sharing
positive valence and low arousal. Which one of these would be felt, if any, would
depend primarily on personal and/or situational factors.

6
EXPRESSION AND INDUCTION IN PERFORMERS

All examples earlier refer to listeners' perception or response. Music performers are
listeners, too, and run the risk of becoming so affected by the expression in the
music they play that the performance breaks down. A violinist for the first time
rehearsing the Adagietto in Mahler's Fifth Symphony told:

What absolutely should not happen during a rehearsal still happened. This [music]
made be begin crying. Tears and snot ran down my face and there was nothing to be
done about it [...] Before the next rehearsal I came with ice-cold determination [...]
not a single tear should drop. When we came to the Adagimo I bit into the inside of
my lips so that the blood poured. My pulse went up [0 200 and I was dripping with
sweat. But when I had to play, my bow-hand began to tremble so that I no longer
reached down to the string [...] I simulated playing some centimetres above the
string [...] (Cited in Wallner, 1988, p. 183; Swedish original text translated into
English by the present author.)

Performing artists - actors, musicians, dancers - have to face the question how
to best achieve a convincing expression. Can it be achieved by "cold" technical use
of appropriate musical means (tempo, loudness, articulation, timing etc) without
any personal engagement - that is, no systematic relationship between expression
and response - or should one try to push oneself into the same feeling as in the
music - that is, positive relationship between expression and response? If the
latter alternative is driven very far, there is a risk of being overwhelmed by feelings
and losing control of the performance. Opinions differ and so do performers among
themselves as well as on different occasions. Maybe most perform-ers adopt an
intermediate posirion: they strive for emotional identification but srill have some
conscious control of performance. A further factor to take into account is the
audience's reactions to the performer's overt behaviour. Listeners usually prefer
performers who show strong personal engagement in facial expression and body
language to performers who show less or nothing - the latter said, for example,
about the violinist jascha Heifetz (Farnsworth, 1969, p. 57).

138
 Emotion pl!fceived and emotion felt
ALFGABRIEl.SSON

7
DISCUSSION

Albeit there are lots of examples of positive relationship between emotion perception
and emotion induction, this relationship is far from general. Examples of negative
or no systematic relationship are easily found. The common belief in a strong
positive relationship reflects another belief, namely that listeners' responses are
exclusively, or at least, predominantly determined by musical factors. As we have
seen, this is a simplistic point of view which suggests neglect of personal and
situational factors. The music-person-situation interplay can never be disregarded. It
certainly complicates questions, but human response to music is complicated; that
is why we Still understand so little about it.
As noted in Section 5.4, the relative impact of musical factors may be stronger
regarding perception of emotion in music than regarding induction of emotion in
listeners. However, personal and situational factors affect emotion perception as
well: all reports on emotion perception (Section 3) invariably demonstrated
large inter-individual variation, and listener agreement practically never reached
100 percent. In fact, a common criterion for regarding a certain expression as
recognised was that at least half of the listeners (50 percent) agreed on the respective
emotion. When the influence of personal and situational factors is free to show itself
as in listeners' emotional response, it is no wonder that the inter-individual variation
may look infinite.
A pertinent methodological problem is that neither all researchers, nor all subjects
clearly observe the distinction between emotion perception and emotion induction.
In reading certain reports, one feels uncertain concerning what the subjects reported:
their perception of emotion in the music, or their own emotional response, or a
mixture of both. From informal listening experiments in which I have asked listeners
to describe "how you experience this music", it is obvious that some listeners mostly
describe the qualities of the music, other their own reactions, and still others mix
these alternatives in different ways. As pointed OUt by Gembris (1982), Ray (2000),
Swanwick (1975), and Zenrner et aL (2000), it is necessary to make the distinction
clear to the listeners in order to possibly separate the two kinds of responses
and thereby be able to study what relationship there is between them. However,
applying this distinction may not be that easy; about 40 percent of Gembris's (1982)
listeners found it rather difficult. A further problem is that several verbal descriptors
may refer either to perceived "objective" properties of the music or to the listener's
own response, or to both (Imberty, 1976, pp. 37-38; Ray, 2000), and the researcher
has no reliable criterion for how to interpret the descriptors. For a simple example,
if the listener tells you that "This music is gruesome", is this a description of
expression in the music, or a description of the listener's reaction, or perhaps both?
The treatment in this paper is based on empirical research as far as possible.
Questions about relation between emotion perception and emotion induction have
 been much discussed by music aestheticians, For instance, Kivy (1990) distinguished
between "cognirivists" and "ernotivists", the former meaning that music can "only"
be expressive of emotions, the latter that music also elicits emotions in the listeners
(see London, this issue). Kivy belongs to the "cognitivists", contending that music
never arouses the emotions of which it is expressive. This standpoint is not shared
by others, for instance, Davies (l994), and certainly not by me. Krumhansl's (l997)
listeners reported emotional responses and showed physiological reactions indicative
of experienced emotions; this was also evident in the study by Nyklicek et al. (1997).
In our studies of strong experiences related to music (Gabrielsson, 2001) listeners
describe both strong emotions and physiological reactions - tears, shivers/chills,
gooseflesh, changed breathing, heart race, and several others - as well as overt
behaviours such as moving to the music, shouting, smiling or, in contrast, dosing
one's eyes, sitting absolutely still hardly breathing, moreover action tendencies such
as wanting to move, wanting to stay in the situation forever, wanting not to leave
although forced to do so, etc. All together these reactions indicate a differentiated
emotional response.
The relationship between emotion perception and emotion induction in
musical contexts requires much further research. Of course, as long as basic facts
in both these areas remain uncertain, we cannot expect to find fully reliable
relationships between them. Because there are only a handful studies directly aimed
at studying the relationship, it is first of all important to carefully plan and conduct
such studies in a systematic way. The distinction between emotion perception and
emotion induction should be dearly explained to the subjects. whether in within-
subjects or between-subjects design, and they should also be instructed to indicate,
as far as possible, which alternative (perceived or induced emotion) is meant when
ambiguous wordings are used (cf. the example with "gruesome" above). Subjects in
within-subjects design may find it difficult to make this distinction consistently (if
Gembris, 1982) or to complete both tasks at the same time. Between-subjects design
may have an advantage in that subjects can completely focus on either perceived
or felt emotion; of course, subject groups shouid then be equivalent in all other
relevant aspects. Multi-method approaches in which verbal reports are supplemented
by physiological and behavioural measures are generally commendable, preferably
using techniques for continuous recording (see Schubert, this issue). Recent
techniques for brain-imaging provide interesting new possibilities (Peretz, 2001).·Of
course, music researchers should generally pay increased attention to theoretical
developments in emotion psychology, as emphasised by Sloboda and ]uslin (2001)
and Scherer and Zentner (2001); this link has been almost totally missing until
recently.
How to come to grips with the complex interplay between musical, personal,
and situational factors is difficult to answer. Systematic selection of many different
types of music, of subjects, and of situations in order to possibly study their effects,
separately and in combinations (seemingly innumerable) would be a standard

140
 Emotion perceived and emotion tell
AU GABRIEl.SSON

answer, but it also represents a "formidable undertaking", like the one suggested by
Scherer and Zentner (2001, p. 384). Awaiting results of continued research and
discussion, we have to be patient and try to successively establish at least some hard
facts and fruitful hypotheses for how to proceed. For some consolation I find it
appropriate to quote Carl Stump's wonderful (underjstaternent concerning the
complexities of mind and music as he put it in the preface to the second volume of
his Tonpsychologit: "Sind doch auch sowol die Musik als die Seele nicht gam. einfache
Dinge" (Stumpf, 1890, p. VUI)l.2.

(1) Approximate translation: "After all, music as well as the mind are not quite simple matters."
(2) The author's research mentioned in this article was supported by The Bank of Sweden
Tercentenary Foundation. The author is grateful to three anonymous reviewers fortheir constructive
comments.

Address for correspondence:

Alf Gabrielsson
Department of Psychology
Uppsala University
Box 1225
SE-751 42 Uppsala
Sweden
e-mail: [email protected]

141
• REFERENCES

Altschuler, 1. M. (1948). A psychiatrist's experience with music as a therapeutic agent. In

D. M. SchulJian and M. Schoen (eds), Music and medicine (pp. 266-281). New York
Henry Schumann.
Bartel, L. R (1992). The: developmenr of the cognitive-affective response test - Music.
Psychomusicology, 11, 15-26.
Bartlett, D. L. (1996). Physiological responses to music and sound stimuli. In D. A. Hodges (ed.),
Handbook of music psychology (2nd ed., pp. 343-85). San Antonio, TX: Institute for
Music Research Press.
Bonny, H. L., & Savary, L. M. (1990). Music and your mind (rev. ed.), Barrytown, NY: Station
Hill Press.
Brown, R. (1981). Music and language. In Docummtllry rqJorr of the Ann Arbor symposium.
National symposium on the application! ofpsychology to tb« uaching and learning of
music (pp. 233-65). Reston, VA: Music Educators National Conference.
Buelow, G.]. (1983). Johann Marrheson and the invention of the Affektenleme. In G.]. Buc:low
and H.]. Marx (eds) , New Manheson studin (pp, 393-407). Cambridge: Cambridge
University Press.
Campbell, 1. G. (1942). Basal emotional patterns expressible in music. American journal of
Psychology, 55, 1-17.
Capurso, A. (1952). The Capurso study. In A Capurso, V. R Fisichelli, L Gilman, E. A Gutheil,
]. T. Wright and F. Paperte (eds), Music and your emotions (pp. 56-86). New York:
Liveright.
Davies, S. (1994). Musical meaning and expression. Ithaca, NY: Cornell University Press.
DeNora, T. (2000). Music in rotryday lift. Cambridge: Cambridge University Press.
DeNora, T. (2001). Aesthetic agency and musical practice: new directions in the sociology of
music and emotion. In P. N. ]uslin and ]. A. Sloboda (eds), Musicand emotion: Thtory
and rmarch (pp. 161-80). New York Oxford University Press.
Farnsworth, P. R (1969). The social psychology of music (2nd ed.), Ames, Iowa: The Iowa State
University Press.
Gabrielsson, A. (1973). Adjective ratings and dimension analysis of auditory rhythm patterns.
Scandinavian journal ofPsychology, 14, 244-60.
Gabric:lsson, A. (2001). Emotions in strong experiences of music. In P. N. [uslin and
]. A. Sloboda (eds), Musicand emotion. Theory and researcb (pp. 431-49). New York
Oxford University Press.
Gabric:lsson, A., & [uslin, P. N. (1996). Emotional expression in music performance: Between the
performer's intention and the listener's experience, Psychology ofMusic, 24, 68-91.
Gabrielsson, A., & ]uslin, P. N. (in press). Emotional expression in music. In R. ]. Davidson,
H. H. Goldsmith and K. R. Scherer (eds), Handbook of affective sciences. New York
Oxford University Press.
Gabric:lsson, A., & Lindstrom, E. (2001). The influence of musical structure on emotional
expression. In P. N. ]uslin and ]. A Sloboda (eds), Music and emotion. Theory and
Rrrtarch (pp. 223-48). New York: Oxford University Press.
Gabrielsson, A., & Lindstrom, S. (1993). On strong experiences of music. Musikpsychologie.
jahrbuch do D~tschm &sclLrchaftfir Musikprychologie, 10, 118-39.

142
 Emotion perceived and emotion felt
Al.F GAllRIELSSON

Gabrielsson, A., & Lindstrom, S. (1995). Can strong experiences of music have therapeutic
implications? In R. Steinberg (ed.), Music and th~ mind machine. Thepsychophysiology
and psychopathology of th~ sense ofmusic(pp. 195-202). New York: Springer.
Gabrielsson, A., & Lindstrom Wik, S. (2000). Srrong experiences of and with music. In
D. Greer (ed.), Musicology and sister discipli~s: P4St. pr~smt and funm (pp. 100-8).
Oxford: Oxford University Press.
Gembris, H. (1982). Zum Verhaltnis Musikuncil - Musikerleben. In K-E. Behne (ed.),
Musikpiidagogisch~ Forschung. Band 3: Gifuhl als Erlebnis - Ausdruckals Sinn (pp.146-
6 I ). Laaber: Laaber Verlag.
Gundlach, R. H. (1935). Factors determining the characterization of musical phrases. American
journal ofPrychology, 47, 624-44.
Hampton, P. J. (1945). The emotional element in music. journal of Gmtral Prychology, 33,
237-50.
Hanslick, E. (1854/1989). \10m musikalisch Schonm. Wiesbaden: Breitkopf & Hartel.
Hargreaves, D. J. (1986). Tb« dnJ~lopmmtal prychology of music. Cambridge: Cambridge
University Press.
Hargreaves, D. J., & Colman, A M. (1981). The: dimensions of aesthetic reactions to music.
Prychology ofMusic, 9, 15-20.
Hedden, S. K (1973). Listeners' responses to music in relation to autochthonous and experiential
factors. journal of&s~arch in Music Education, 21, 225-38.
Hevner, K (936). Experimenral studies of roe elements of expression in music. Amtrican journal
ofPrychology, 48, 246-68.
Hodges, D. A (l980). Physiological responses to music. In D. A Hodges (ed.), Hand/;ook ofmusic
prychology (pp, 393-400). Lawrence, Kansas: National Association for Music Therapy.
Imberty, M. (1976). Signification and meaning in music [on Debussy's preludes pour le piano].
Monographi~s tk sbniologi~ rt d'a7Jlllym musicaks, III Montreal: Faculte de Musique,
Universire de Montreal.
]uslin, P. N. (l997). Perceived emotional expression in synthesized performances of a short
melody: Capturing the listener's judgment policy. Music« Scientie; 1, 225-56.
Juslin, P. N. (2000). Cue utilization in communication of emotion in music performance: Relating
performance to perception. journal of Experimental Psychology: Human Perception and
Pnfiil77J1lnc~, 26, 1797-813.
]0rgensen, H. (l988). Musikkopp/n;~lsmsprykologi [The psychology of music experiena]. Oslo:
Norway: Norsk Musikforlag.
Kivy, P. (l990). Music alone. Philosophical rtflmio1'lJ on tht purtly musicalexperimce. Ithaca, NY:
Cornell University Press.
Kleinen, G. (1968). Experimentelle Studim zum musikalischm Ausdruck. Hamburg: Universitat
Hamburg.
Krumhansl, C. L. (l997). An exploratory study of musical emotions and psychophysiology.
Canadumjournal ofExptrimmtal Psychology, 51, 336-52.
Lehmann, A C. (1997). Research note: Affective responses to everyday life events and music
listening. Prychology ofMusic, 25, 84-90.
Lewis, M. J. (1998). Music and peak experiences: An empirical srudy. Tbe Mankind Q!lartn-ly, 39, 203-24.
Lundqvist, L. O. G., Carlsson, E, & Hilrnersson, P. (2000). Facial tkctromyography. autonomic
activity. and emotiona! experience to happy ll1'ld sad music. Paper given at the
XXVII International Congress of Psychology, Stockholm, Sweden, 23-28 July 2000.
143
Madsen, C K (1997). Emotional responses [0 music. Psychomusicology, 16, 59-67.
Maslow, A H. (1968). Toward apsychology of being (2nd ed.). New York: Van Nostrand Reinhold.
Maslow, A H. (1976). Thefartht:r rt:achf:! of hU7'nall nature. New York: Penguin Books.
Milliman, R. E. (1982). Using background music [0 affect the behavior of supermarket shoppers.
JournalofMarketing, 46, 86-9 J.
Milliman, R. E. (1986). The influence of background music on rhe behavior of restaurant patrons.
Journal ofConsU7'nt:r Rrsearch, 13, 286-9.
Mursell, J. L. (1937). Thepsychology ofmusit:. Wesrport, CT: Greenwood Press.
Neale, ]. M., & Liebert, R. M. (1986). Science and behavior: An introduction to methods of
research (3rd. ed.). Engelwood Cliffs, N]: Prentice Hall.
Nielzen, S., & Cesarec, Z. (1981). On the perception of emotional meaning in music. Psychology
ofMusic, 9, 17-3 J.
North, A. C, & Hargreaves, D.]. (1997a). Music and consumer behaviour. In D.]. Hargreaves
and A. C North (eds), The social psychology of music (pp, 268-89). Oxford: Oxford
University Press.
North, A. C, & Hargreaves, D.]. (1997b). Experimental aesthetics and everyday music listening.
In D. J. Hargreaves and A. C North (eds), Thesocial psychology ofmusic(pp. 84-103).
Oxford: Oxford University Press.
Nyklicek, I.,Thayer, ]. E, & Van Doornen, L ]. P. (1997). Cardiorespiratory differentiation of
musically-induced emotions. Journal of Psychophysiology, II, 304-21.
Oblad, C (2000). Art anviinda musi]: - Om bib som konsmlokal [Using music - Thecar asconcert
hall). Gothenburg University, Sweden: Department of Musicology. ISBN 91 8597459-5.
Peretz, I. (2001). Listen to the brain: A biological perspective on musical emotions. In P. N. ]uslin
and ]. A. Sloboda (eds), Music and emotion: Theory and research (pp. 105-34). New
York: Oxford University Press.
Pike, A. (1972). A phenomenological analysis of emotional experience in music. Journal of
Research in Music Education, 20, 262-7.
Radocy, R E., & Boyle,]. D. (1997). Psychological foundations of musical behavior (3rd. ed).
Springfield, Illinois: Charles C Thomas.
Ray, ]. (2000). Musikuppkveue - En kvalitativ studi« fA qualitative study of music experience],
Unpublished thesis, Department of Psychology, Uppsala University.
Rigg, M. G. (1937). Musical expression: An investigation of the theories of Erich Sorantin. Journal
ofExperimmtal Psychology, 21, 442- 55.
Rigg, M. G. (1940). Speed as a determiner of musical mood. JournalofExperimental Psychology,
27, 566-71.
Roe, K. (1985). Swedish youth and music: Listening patterns and motivations. Communication
Rmarch, 12, 353-62.
Russell, ]. A. (1980). A circumplex model of affect. JournalofPmonalityand Social Psychology. 39,
1161-78.
Scherer, K. R., & Oshinsky, ]. S. (1977). Cue utilization in emotion anriburion from auditory
stimuli. Motivation and Emotion, 1,331-46.
Scherer. K. R., & Zentner, M. R. (2001). Emotional effects of music: production rules. In
P. N. ]uslin and J. A Sloboda (eds), Music and emotion. Theory and research (pp. 361-
92). New York: Oxford University Press.
Schubert, E. (1999). Measurement and time series analysis ofemotion in music. Unpublished doc-
toral dissertation, University of South Wales, Sydney, Australia.

144
 Emotion perceived and emotion felt
ALFGAIIRlELSSON

Schubert, E. (2001). Continuous measurement of self-report emotional response to music. In

P. N. ]uslin and ]. A. Sloboda (eds), Music and emotion. T~ory and mf:arch (pp, 393-
414). New York Oxford University Press.
Scruton, R. (1997). The aesthetics ofmusic. Oxford: Oxford University Press.
Shatin, L. (1970). Alteration of mood via music: A study of the vectoring effect. Journal of
Psychology, 75, 81-6.
Sloboda,]. A. (1992). Empirical studies of musical affect. In M. Riess-lones and S. Hellerman (eds),
Cognitivf: bam of musical communication (pp. 33-45). Washington: American
Psychological Association.
Sloboda,]. A. (1999). Musial performance: and emotion: Issues and developments. In S. W Yi (ed.),
Music, mind, and science (pp, 220-38). Seoul: Seoul National University Press.
Sloboda, ]. A., & ]uslin, P. N. (2001). Psychological perspectives on music and emotion. In
P. N. [uslin and ]. A. Sloboda (eds), Music and emotion. Thf:ory and research (pp, 71-
104). New York: Oxford University Press.
Sloboda,]. A., & O'Neill, S. A. (2001). Emotions in everyday listening to music. In P. N. ]uslin
and J. A. Sloboda (eds), Music and emotion. Theory and wf:arch (pp, 415-29). New
York: Oxford University Press.
Sopchak, A. L. (1955). Individual differences in responses to different types of music in relation to
sex, mood, and other variables. Psychological Monographs: Gmf:7al andApplif:d, 69, 1-20.
Stumpf C. (1890/1965). Tonpsychologif: (Zwf:itf:7 Band). Hilversurn: Frits A. M. Knuf and
Amsterdam: E. J. Bonser,
Swanwick, K. (1973). Musial cognition and aesthetic response. Psychology of Music, 1,7-13.
Swanwick, K. (1975). Can there be objectivity in listening to music. Psychology ofMusic, 3, 17-23.
Wallner, B.(1988). Uppleva - Iaktta - Beskriva - Formedla [Experience - Observe - Describe -
Convey]. In F. V. Nielsen and O. Winther (eds), Musik - opkvf:lsf:, analySt: og
formidling [Music - Experimce, analysis and mf:diation} (pp, 173-91). Egrved,
Denmark Egtved Edition.
Warerman, M. (19%). Emotional responses to music: Implicit and explicit effects in listeners and
performers. Psychology ofMusic, 24, 53-67.
WaOion, K. B. (1942). The nature and measurement of musical meanings. Psychological
Monographs, 54, 1-43.
Wedin, L (1969). Dimension analysis of emotional expression in music. SWf:dish Journal of
Musicology, 51, 119-40.
Wedin, L. (1972). Multidimensional study of perceptual-emotional qualities in music.
Scandinavian Journal ofPsychology, 13,241-57.
Weld, H. P. (1912), An experimental srudy of musical enjoyment. American journal ofPsychology,
23, 245-308.
Wigram. T., Sapersron, B., & Wesr. R. (eds) (1995). Tb« art and science of music thf:rapy: A
handhook. Chur, Switzerland: Harwood Academic Publishers.
Wundt, W (1905). Grundriss dn Psychologif: (7te Auflage). Lc:ipzig: Wilhelm Engc:lmann.
Zentner, M. R., Meylan, S., & Scherer, K. R. (2000). Exploring "musical emotions" across jiVf:
gmm ofmusic. Paper given at the Sixth International Conference:for Music Perception
and Cognition, Keele, UK, 5-10 August 2000.
Zillman, D., & Gan, SoL. (1997). Musical taste in adolescence. In D. J. Hargreaves and
A. C. North (eds), The social psychology of music (pp. 161-87). Oxford: Oxford
University Press.
145
 • Emoci6n percibida y emoci6n sentida: llgual 0 diferente?

EI trabajo propene una distind6n entre emoci6n y percepdon. esto es percibir

expresi6n emocional en rnuslca sin ser necesariamente afectado uno mismo, y la
mducclon ala emotion. esto es, la respuesta emocional del oyente a la musica. Esta
distinci6n no se observa siernpre, ni en las conversactones sobre emociones, ni
en los trabajos cientlficos. Se revisan brevemente los estudios emplricos sobre
percepcien de la emoclon reladonados con el juicio del oyente respecto a las
ernoclones expresadas, seguido de una revision selectiva de los estudios ernplrlcos
sobre la respuesta emocional a la musica. 5e discute una posible relaci6n entre
percepcion de la ernocicn y respuesta emocional y se ejemplifica como relacion
posltiva, relad6n negativa, relacion no sistematica y no relacion. 5e destaca que
arnbas, percepcion de la emoci6n y, especialmente, respuesta ernocional,
depend en de una intersecdon de factores musicales, personales y de contexte,
Finalmente se discuten algunas cuestiones rnetodologicas y sugerendas para
futuras investigaciones.

• Emozione percepita ed emozione sentita: uguali 0 differenti?

5i distingue fra percezlone emotiva, ossia perceplre un'espressione emotiva in

musica senza venirne necessartarnente coinvolti in prima persona, e induzione di
un'emozione, ossia la risposta emotiva alia muslca da parte degli ascoltatorl, Tale
distinzione non viene sernpre osservata nelle conversazioni quotidiane sulle
emozioni, e neppure negli artkoli scientifid. 5i passano brevemente in rassegna
studi empirici sulla percezione dell'emozione, con partlcolare riguardo alia
concordanza degli ascoltatori sulle emozioni espresse; segue una rassegna
selettiva di studi ernplrid sulla risposta emotiva alia musica. Vengono discusse ed
esemplificate possibili relazioni fra percezlone dell'emozione e risposta emotiva:
relazione positiva, negativa, relazione non sistematica e assenza di relazioni. 51
sottolinea come sla la percezlone dell'emozione sia, in particolar modo, la risposta
emotiva dipendano da un'interazione tra fattori musicalr, personali e contestuall, 5i
dlscutono a1cune questioni metodologiche e alcune proposte per la ricerca futura.

• Emotion per~ue et emotion ressentie : y a-t-i1 identite au difference?

On etabllt une distinction entre l'ernotlon percue (percevolr l'expression

emotionnelle en musique sans en Hre necessalrement affecte) et I'induction
emotionnelle (Ia reponse emotlonnelle a la musique). Cette distinction n'est pas
toujours observee, ni dans la conversation courante ni dans lestextes sctentifiques.
Les etudes empiriques de la perception ernotionnelle sont brievernent passees en
revue du point de vue de la concordance entre la reponse de I'auditeur et les
emotions expnmees, On se livre ensuite a une critique selective des etudes
ernpiriques de la reponse ernotlonnelle a la musique. Les relations eventuelles entre
146
la perception de i'ernotion et la reponse a i'emoticn sont discutees et iIIustrees :
relation positive, relation negative, absence de relation systematique, absence de
toute relation. II ressort de cette etude que tant la perception emotionnelle que la
reponse ernotionnelle sont fonction d'une interaction entre les facteurs rnusicaux,
personnels et situationnels. Certaines questions rnethodologiques sont discutees et
de nouvelles pistes de recherche sont avancees,

• Wahrgenommene und gefuhlte Emotion: gleich oder verschieden?

Es wird unterschieden zwischen Emotionsperzeption (Wahrnehmung des

emotionalen Ausdruckes in Musik ohne notwendigerweise selbst betroffen zu sein)
und Emotionsinduktion (die emotionelle Reaktion des Musikhorers). Diese
Unterscheidung wird nicht immer beachtet, weder in der Alltagssprache noch im
wissenschaftlichen Diskurs. Wissenschaftliche Studien zur Emotionsperzeption
werden in Bezug auf die Obereinstimmung der Harer hinsichtlich der zum
Ausdruck gebrachten Emotionen kurz besprochen. Es folgt eine selektive
Besprechung empirischer Studien zur emotionalen Reaktion auf Musik. Magliche
Beziehungen zwischen Emotionswahrnehmung und Emotionsreaktion werden
diskutiert und ellemplifiziert, namlich: positive Beziehung, negative Beziehung,
keine systematische Beziehung, keine Beziehung. Es wird betont, daB sowohl
Emotionsperzeption als auch insbesondere Emotionsreaktion vom Zusammenspiel
von musikalischen, person lichen und Situationsfaktoren abhangen, Einige
methodologische Fragen und Anregungen tur weitere Untersuchungen werden
diskutiert.

147