You are on page 1of 32

Copyright © 2003. Oxford University Press. All rights reserved.

May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

26
EMOTIONAL EXPRESSION IN MUSIC

Alf Gabrielsson and Patrik N. Juslin

Pieces of music are dynamic events in time—not static, press and arouse emotions more readily than visual stim-
visible, or tangible as are pieces of pictorial and sculptural uli. People who have become deaf as adults experience “a
art. In the latter, perceived color, light, surface structure, consequent draining of feeling from the world about
and other properties cannot be separated from the objects them” (Brown, 1981, p. 240). Before the advent of sound
themselves. Sounds, tones, and music may likewise be film, a pianist was hired to strengthen the expression of
perceived as belonging to objects (e.g., instruments, per- what happened on the screen, and nowadays music is rou-
formers), but they may also be attended to as such—as tinely used to generate or amplify emotional expression in
“pure events”—without any consideration of their origin plays, films, videos, television programs, and so forth (Co-
or belonging (Scruton, 1997; Zuckerkandl, 1956). Music hen, 2001).
often seems abstract and ineffable, inaccessible to descrip- Before proceeding further, it is necessary to refer to a
tion by ordinary language, or only hinted at by means of distinction between, on the one hand, a listener’s percep-
similes or metaphors. tion of expression in the music—to perceive an expression
While most pictorial art depicts phenomena visible in of, say, sadness without being affected oneself—and, on
nature, music rarely depicts in that sense, except in a few the other hand, a listener’s response to the music—to feel
examples of program music (Davies, 1994, chap. 2). Nor sad. This distinction is not always observed, neither in
can music be regarded as a language because its elements theoretical treatises nor in empirical investigations (and in
(e.g., tones, chords) do not have fixed dictionary meaning, reality the border between the two alternatives is some-
as elements in language usually have (Davies, 1994, chap. what blurred). In the following, however, we focus mainly
1; Hermerén, 1986; Langer, 1957, p. 228; Scruton, 1997, on perceived emotional expression. (For discussions of
chap. 7). On the contrary, musical elements are ambigu- emotional reactions to music, see Gabrielsson, 2001;
ous; they mean different things in different contexts. Scherer & Zentner, 2001; Sloboda & Juslin, 2001.)
So what, then, can music do? Richard Wagner, who Expressive qualities of Western music have been dis-
wrote both libretto and music for his operas, declared that cussed by philosophers, music theorists, and others ever
music begins where language comes to an end, and that since antiquity, and many opinions have been issued con-
applicable copyright law.

music is the language of passion (Benestad, 1978). Langer cerning what can be expressed in music. We first take a
(1957) claimed that music “can reveal the nature of feel- look at some of these ideas as they relate to emotion. Then,
ings with a detail and truth that language cannot ap- we review empirical investigations on emotional expres-
proach” (p. 235), and Cooke (1959) maintained that music sion in music. Finally, we discuss methodological and the-
is not apt to expression of ideas or abstract concepts but oretical issues and consider implications for future re-
is the expression of emotion. Sounds, it seems, may ex- search.

503
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
504 PART IV. EXPRESSION OF EMOTION

and opera successively became the most large-scale at-


Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Historical and Theoretical Perspectives tempt to combine music and text (also dance, acting, scen-
ography) to describe and express a vast range of human
Music as Mirror of Human Character feelings, passions, and characters. Furthermore, it is about
this time that various expression marks began to appear
During antiquity, philosophers believed that music, like in composers’ scores, such as forte and piano for dynam-
other art forms, was a kind of mimesis—that is, imitation ics, allegro and adagio for tempo. The number of expres-
of nature—in this case an imitation of human character or sion marks successively increased during the following
states of mind.1 It was also assumed that what was ex- centuries to constitute hundreds. To mention but a few, in
pressed in music would be automatically imitated by the alphabetical order: affettuoso (with warmth), amoroso
listener (Allesch, 1987; Grout & Palisca, 1996; Scruton, (with love), burlesco (burlesque), capriccioso (capricious),
1997). Plato (427–347 B.C.) and Aristotle (384–322 B.C.) dolce (sweet), and espressivo (expressive).
therefore advised on which modes, instruments, and Music’s dependence on words became most obvious
rhythms ought to be used to educate young people to be- during the 17th and 18th centuries. In his influential trea-
come good citizens, as well as to bring balance and calm, tise Der vollkommene Capellmeister (1739; see also Lipp-
arouse enthusiasm or excitement, and so on. For instance, man, 1986), Johann Mattheson (1681–1764) presented a
Plato thought that the Dorian mode embodied male deter- plan for musical composition in direct connection to basic
mination, which he regarded as a political ideal.2 concepts in rhetorics. So-called rhetorical figures were
The Greeks also realized that the expressive qualities used in music to illustrate or emphasize words or ideas in
of music could be used for therapeutic purposes. In their the text. For instance, ascending stepwise movement
view, man’s natural and ideal state was one of perfect har- could be used to express raising in a literal sense or in a
mony, order, and health. Unfavorable states of the mind transferred sense (e.g., resurrection); sequences of short
could be relieved and harmony and order reestablished by notes could express rapidity such as in flight; chromatic
means of appropriate music, either by using music of a falling melody could express suffering and sadness; falling
different emotional character from that of the upset state melodies with inserted pauses, sighs, pain and exhaus-
of mind or by deliberately driving unwanted feelings into tion. There are long lists of musical-rhetoric figures (Be-
extreme states to be discharged, thus achieving catharsis— nestad, 1978; “Rhetoric and music” in Sadie, 1980; Unger,
the purification of the mind. Both principles can still be 1979).
found in different forms of music therapy (Pratt & Jones, The Doctrine of the Affections (German, Affektenlehre)
1987; Spintge & Droh, 1987). meant that music, predominantly vocal music, should ex-
press affects in the sense of idealized emotional states and
have listeners feel these states (Buelow, 1983). It was as-
Music in Support of Text
sumed that every affect had its particular character ex-
During most of the Middle Ages, music in the Roman pressible in music. Mattheson described some 30 affects
Catholic church served to express and strengthen Chris- (translated into English in Buelow, 1983, pp. 404–406),
tian faith as couched in the holy texts. Music had little but concrete musical means to represent them were sug-
aesthetic value in itself. Still, Saint Augustine (354–430) gested only for a few—for example, large intervals to rep-
was well aware of the expressive powers of music and resent joy, small intervals to represent sadness, ascending
described in his Confessiones (cited in Lippman, 1986, motion for pride but descending motion for humility, and
pp. 28–29) how he vacillated between the temptation of disordered sequences of notes for despair.
merely enjoying the beautiful melodies themselves, for-
getting the text, and his insight that faith was actually fa-
Music as Independent Expressive Art
vored by having the texts performed with such beautiful
melodies. Using music in the service of Christian religion During the latter half of the 18th century, the Doctrine of
has been, and still is, an important part of Western music the Affections was successively abandoned in favor of an
history. Empfindungsästhetik, the possibility and freedom to use
The idea of music as the “obedient daughter” of the text whatever means fantasy and intuition may suggest to ex-
was also characteristic of Italian monody around 1600 press subjective feelings in music—a parallel to Sturm
(i.e., monophonic singing with a simple chord accompa- und Drang in the literature of that time. Important pre-
applicable copyright law.

niment). Vincenzo Galilei (1520–1591), father of Galileo, cursors were composers Johann Joachim Quantz (1752)
advised composers to study how prominent actors used and Carl Philipp Emanuel Bach (1753, 1762). The mimesis
their voices in order to express various affects. Singing, he principle, central in art ever since antiquity, was ques-
argued, should imitate the characteristics of emotional tioned (Lippman, 1986, parts 6–7), and music came to be
speech. ascribed an independent position as expressive art with
At this time, the earliest operas were composed in Italy, powers to affect listeners. From the late 18th century, in-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 505

strumental music came to dominate vocal music and was often mentioned as representative of absolute music. Stra-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

often considered as the only genuine type of music, in- vinsky declared that “music, by its very nature, is essen-
dependent of words. Nineteenth-century musical roman- tially powerless to express anything at all, whether a feel-
ticism emphasized the subjective and emotional in music. ing, an attitude of mind, a psychological mood, a
The composer was regarded as a divinely gifted genius, phenomenon of nature etc. . . . Expression has never been
who followed solely his own inspiration, capable of in- an inherent property of music. . . . If, as is nearly always
sights into areas inaccessible to the common human be- the case, music appears to express something, this is only
ing—the Infinite, the Eternal, the Transcendental. an illusion. . . . It is simply an additional attribute that . . .
we have lent it, thrust upon it, as a label, a convention”
(cited in Fisk, 1997, pp. 280–281). Later he modified this
Absolute Music and Program Music
statement but still held that “music expresses itself” (Fisk,
Music was also used to refer to various extra-musical phe- 1997, p. 281).
nomena in so-called program music. It could mean only In discussing these questions, Meyer (1956) made a dis-
occasional imitations of natural sounds—as bird song and tinction between “absolutists” and “referentialists.” The
a thunder storm in Beethoven’s Pastoral Symphony—or former claim that the meaning of music is intramusical
refer to a complete story as in Symphonie Fantastique by (embodied meaning) and music refers only to itself,
Hector Berlioz (1803–1869), in which he intended to de- whereas the latter claim that music receives meaning by
scribe an artist’s passionate and unhappy love of a woman. referring to extramusical phenomena (designative mean-
However, composers were aware of music’s limited pos- ing). Meyer emphasized, however, that these alternatives
sibilities to represent events or ideas, and rather referred are not mutually exclusive but may coexist in the same
to the feelings associated with the program. Berlioz de- piece of music.
clared that he did not try to depict abstract ideas or moral
qualities but only passions and impressions (Benestad,
Modern Theories of Emotional Expression
1978, p. 233).
in Music
The famous Viennese music critic Eduard Hanslick
considered program music and the claim that music
Cooke’s Theory
should describe emotions as a threat to music’s autonomy.
He instead argued for what is usually called absolute or The most pronounced spokesman for music as expression
pure music—that is, instrumental music that is self- of emotion is probably the musicologist Deryck Cooke. In
contained, not referring to anything outside itself. In his his book The Language of Music (1959), Cooke claimed
book Vom musikalisch Schönen (1854/1989), Hanslick that music considered as expression has three separate as-
claimed that “Der Inhalt der Musik sind tönend bewegte pects: (1) an architectural aspect typified by contrapuntal
Formen” (p. 59)—that is, the contents of music are works (e.g., a fugue) appealing to us by the beauty of pure
“sounding forms in motion” (Langer, 1953, p. 107), or form; (2) a pictorial aspect in a few works containing im-
“forms moved through sounding” or “forms moved itation of natural sounds (e.g., bird song); and a (3) literary
through tones” (Scruton, 1997, p. 353). With regard to feel- aspect “found . . . in most Western music written between
ings, Hanslick argued that music can never describe the 1400 and the present day, since music is, properly speak-
feelings per se, only their dynamic properties. It may re- ing, a language of emotions, akin to speech. The appeal of
produce the motion associated with physical events ac- this music is directly to the emotions and, to be fully ap-
cording to its momentum: rapid, slow, strong, soft, rising, preciated, should be responded to in this way” (pp. 32–
or falling. But, emphasized Hanslick, motion is only a con- 33).
comitant of feeling, not the feeling itself. That we perceive Following a discussion of the emotional expression of
motion qualities—walking, dancing, rocking, swinging, different musical elements, Cooke identified, using an im-
driving forward, accelerating, retarding, etc.—in music is pressive number of examples from the late Middle Ages
a generally accepted idea (for a review, see Shove & Repp, until our own time, 16 so-called basic terms of musical
1995), and motion qualities have emotional connotations: vocabulary, and he suggested the emotional expression
“motion is heard in music, and that motion presents emo- connected with each of them. They were described in
tion characteristics much as do the movements giving a terms of sequences as, for example, “Ascending 1-(2)-3-
person her bearing or gait” (Davies, 1994, p. 229; see also (4)-5 (Major)” designating an ascending major triad (1-3-5,
applicable copyright law.

Gabrielsson, 1988). Hanslick’s ideas still continue to elicit with possible insertions of the intervening notes, 2 or 4),
much discussion (Budd, 1985, chap. 2; Davies, 1994, said to express “an outgoing, active, assertion of joy”
p. 152, p. 202 onwards, p. 283; Kivy, 1990, pp. 184–189; (p. 115). Its counterpart in minor was characterized as “ex-
Scruton, 1997, p. 165, pp. 348–350, 353–354; Tarasti, pressive of an outgoing feeling of pain—an assertion of
1994, p. 30). sorrow, a complaint, a protest against misfortune” (p. 122).
The work of composer Igor Stravinsky (1882–1971) is Cooke went on to show how the short basic terms could

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
506 PART IV. EXPRESSION OF EMOTION

also be traced in extended musical themes and even in the a specification of the structure of feelings and how this
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

structure and emotional expression of separate move- should be measured, without which her hypothesis has
ments and complete works. (Empirical tests of Cooke’s little explanatory force. Therefore, there are no direct at-
theory are discussed in the next major section of this chap- tempts to test her theory empirically. However, Langer’s
ter.) theory is still often discussed and criticized (cf. Budd,
1985, chap. 6; Davies, 1994, p. 123 onward; Scruton, 1997,
p. 166; Tarasti, 1994, p. 12).
Langer’s Theory
Unlike Cooke, philosopher Susanne Langer claimed that
Clynes’s Theory
music has no literal meaning apart from some onomato-
poetic themes. The elements in music—tones, chords, Clynes (1973, 1977, 1980) has argued for the existence of
rhythms, and so forth—do not carry fixed lexical meaning biologically preprogrammed spatiotemporal patterns, or
as do words in language. In her books Philosophy in a New essentic forms, for the communication of emotions. Each
Key (1957) and Feeling and Form (1953), she developed a essentic form, Clynes claims, can be expressed by any of
theory of symbols in reason, rite, and art. Langer made a a number of output modalities (e.g., gesture, facial expres-
distinction between what she called discursive symbolism sion, tone of voice, a dance step, or a musical phrase), as
and presentational symbolism. Language is an example of long as the dynamic pattern is preserved. Clynes’s theory
discursive symbolism, in which the symbols have fixed is reminiscent of the seminal work by Tomkins (1962),
meanings. Music, in contrast, is an example of a presen- who postulated the existence of biologically based “affect
tational symbolism—that is, an open (“unconsummated”) programs,” stored in subcortical areas of the brain and as-
symbol, where the meaning of various symbolic elements sociated with specific nonverbal expressions of a set of
can be understood “only through the meaning of the “basic” emotions. Clynes went further in that he specified
whole, through their relations within the total structure” the exact spatiotemporal patterns used in such emotional
(1957, p. 97). communication. For the measurement of these patterns he
According to Langer, the composer’s general knowledge created a particular device—the sentograph—that allows
about feelings receives a symbolic expression in music. “A the recording of both vertical and horizontal pressure ex-
composer . . . articulates subtle complexes of feeling that erted by a finger upon a small disk placed on a small box.
language cannot even name; he knows the forms of emo- Participants are asked to express specific emotions by
tions and can handle them, ‘compose’ them” (1957, pressing this disk—literally ex-pressing the emotion.
p. 222), and “Because the forms of human feeling are Clynes claims that there is a high degree of concordance
much more congruent with musical forms than with forms of such expressive patterns for particular emotions (e.g.,
of language, music can reveal the nature of feelings with joy, grief, anger, hate, reverence, love, sex) both within and
a detail and truth that language cannot approach” (1957, across participants from different cultures; see figures and
p. 235). Music expresses the dynamic forms of feelings, discussion in Clynes (1977, 1980) and Clynes and Nett-
not their content; as earlier formulated by Pratt (1931, heim (1982).
p. 191), the auditory structures in music “sound the way Clynes suggested using the sentograph to study emo-
moods feel.” This comes close to Hanslick’s opinion, and tional expression in music by having listeners pressing the
Langer agreed with his description of music as sounding disk in accordance with the flow of musical events. This
forms in motion. “Such motion is the essence of music: a should serve to uncover the essentic forms hypothesized
motion of forms that are not visible, but are given to the for different emotions. He further proposed transforming
ear instead of the eye” (1953, p. 107). Langer did not go pressure patterns into melodic contours for different emo-
into much musical detail, but she approvingly referred to tions—for instance, a melody descending by semitone
an investigation by Huber (1923; see also “Emotional Ex- steps for expression of grief (Clynes & Nettheim, 1982,
pression and Musical Structure,” in the next section of p. 73); this reminds of an earlier rhetorical figure for sad-
this chapter) on expression in short pitch patterns as evi- ness (see earlier “Music in Support of Text”). (For attempts
dence for how many factors in musical structure can have to test Clynes’s theory, see “Clynes’s Theory” in next major
expressive functions. section of this chapter.)
The basic thrust of Langer’s theory is that there is an
isomorphism between the structure of feelings and the
applicable copyright law.

Emotional Expression in Performance


structure of music. Formal characteristics that may be sim-
ilar in music and human feeling include “patterns of mo- While most discussions of emotional expression in music
tion and rest, of tension and release, of agreement and dis- concern variables in the musical structure specified by the
agreement, preparation, fulfillment, excitation, sudden musical notation (e.g., pitch, mode, melody, harmony), the
change, etc.” (1957, p. 228). Langer’s ideas are intuitively influence of variables in the performance (e.g., articula-
appealing, but one problem with the theory is to provide tion, timing, intonation) has so far received less attention.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 507

But the same notated structure can be performed in a num- terms of a small number of “basic” emotion categories
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

ber of different ways, and the way it is performed may (e.g., Izard, 1977; Oatley, 1992), which provide decoders
influence the listener’s impression of the music in pro- with maximum information and discriminability (cf. Ross
found ways. Moreover, the mechanisms underlying emo- & Spalding, 1994). In this context, it is hypothesized that
tional expression via performance might be different from basic emotions (1) have distinct functions, (2) are found
those underlying emotional expression via composed in all cultures, (3) are experienced as unique feeling states,
structure. Therefore, it makes sense to consider perfor- (4) appear early in human development, (5) are associated
mance separately. with distinct neurological substrates, (6) can be inferred
The only elaborated theoretical framework focusing in other primates, and (7) have distinct expressions (see
specifically on performance aspects has been presented by Oatley, 1992; Panksepp, 1998; Plutchik, 1994).
Juslin (1997a, 2001a, 2001b). He recommended that re- The second factor governing emotional expression in
searchers adopt a functionalist perspective on communi- performance is social learning, or specific memories. This
cation of emotion via music performance. This involves is a life-long process that begins with the interaction be-
the integration of ideas from research on emotion and non- tween mother and infant. When mothers talk to their in-
verbal communication with Brunswik’s (1956) lens model. fants—for example, if they want to calm an infant—they
reduce the speed and intensity of their speech and talk
with slowly falling pitch contours. If mothers want to ex-
Origin of the Code
press disapproval toward some unfavorable activity, they
As we shall see later in this chapter, performers can com- use brief, sharp, and staccato-like contours (Papoušek,
municate specific emotions to their listeners. What is it 1996). Certain aspects of emotional expression in perfor-
that makes this possible? According to the functionalist mance (e.g., the cues not shared with vocal expression)
perspective, the answer to this question lies in considering may be completely determined by cultural influence. Oth-
the functions that communication of emotions has served, ers, although based on innate mechanisms, are later mod-
and continues to serve, in social interaction. Emotional ulated by experience. This factor can explain individual
communication via acoustic signals did not begin when differences among performers and listeners. However, Jus-
humans invented the first musical instrument. Humans lin assumes that the first factor—the innate vocal code—
have always communicated emotions via the voice (see has the greatest impact on performers’ expression of emo-
chapter 28, this volume) and many animals communicate tion.
emotions via acoustic signals as well (see chapter 24, this This assumption leads to a number of testable predic-
volume). Although music performance may be a relatively tions: for example, that (1) there are parallels between vo-
recent addition to the human behavior repertoire, it seems cal expression of emotion and expression of emotion via
plausible that how performers communicate emotions is performance, (2) basic emotions are easier to communicate
constrained by psychological mechanisms for acoustic than other emotions, (3) decoding of emotions from per-
communication shaped by evolution (Juslin, 1997a). formance is fast, (4) there is cross-cultural similarity and
Accordingly, Juslin (2001a) argues that cue utilization accuracy in emotional expression in performance, (5) de-
in emotional expression via music performance reflects a coding of emotions from performance develops early, and
compromise between two factors. The first factor reflects (6) decoding of emotions from performance is largely in-
innate “brain programs” for vocal expression of emotion dependent of musical training (see “Emotional Expression
(e.g., Jürgens & von Cramon, 1982; Ploog, 1986). It is hy- in Performance,” in the next major section of this chapter).
pothesized that performers can communicate emotions to
listeners by using the same code as is used in vocal ex-
Description of the Code: The Lens Model
pression of emotion. The notion that there is an intimate
relationship between music and the human voice has a Juslin (1995) suggested that we should use Brunswik’s
long history (von Helmholtz, 1863/1954; Rousseau, 1761/ (1956) lens model to illustrate how a performer encodes a
1986; cf. Scherer, 1995), but this link is not sufficient to certain emotion by means of a large number of probabilis-
explain all aspects of music’s expressiveness (Budd, 1985, tic (i.e., uncertain) albeit partly redundant cues in the per-
chap. 7). Thus, it would seem that the hypothesis needs formance (e.g., articulation). The emotional expression is
to be constrained in order to have any explanatory force decoded by the listener, who uses the same cues to judge
(e.g., Juslin, 1999). More specifically, the parallels between the intended emotional expression. Intercorrelations
applicable copyright law.

vocal expression and music may apply only to those as- among the cues reflect both how sounds are produced on
pects of the music that are typically under the performer’s musical instruments (e.g., a harder string attack may pro-
control (e.g., tempo, sound level, timbre), and not to the duce a tone that is both louder and has sharper timbre)
other aspects of the musical structure (e.g., harmonic pro- and how performers employ the cues to express emotions.
gression, tonality). Juslin also suggests that encoding and To explain the success (or failure) of the communicative
decoding of emotion from acoustic signals proceed in process, one needs to describe the performer’s cue utili-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
508 PART IV. EXPRESSION OF EMOTION

zation (relationships between performer intention and uous or nonverbal recording of perceived expression have
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

cues), as well as the listener’s cue utilization (relationships recently been developed as well. Characteristics of most
between cues and listener judgment). For reliable com- studies with regard to stimuli, subjects, response type, and
munication to occur, the performer’s cue utilization must descriptive attributes are given in Table 26.1.
be as similar as possible to the listeners’ cue utilization. The studies reviewed in this section feature a broad
However, because the cues are intercorrelated, many dif- range of affective terms. There is a growing consensus
ferent cue utilization strategies may lead to a similar level among emotion researchers that terms like affect, emotion,
of communication accuracy (Juslin, 2000). and mood should be differentiated from one another (e.g.,
Batson, Shaw, & Oleson, 1992; Ekman & Davidson, 1994,
chap. 2). However, music researchers have tended to use
Summary of Modern Theories these terms interchangeably, and in the following review
The theories just discussed show both similarities and dif- we simply use the same terms as were used by the authors
ferences along a number of dimensions. First, they all themselves.
agree in assuming a consistent relationship of some kind
between musical structure and emotional expression.
Listener Agreement
However, whereas Langer holds that music can symbolize
only the general form of feelings, Cooke, Clynes, and Jus- A crucial question concerns whether listeners agree on
lin assume that music can express particular emotions. which emotions are expressed in music. Results depend
The latter three theories agree in that they assume that to a large extent on the methods used. When listeners were
there are specific patterns of musical factors that corre- asked to provide their own free descriptions of what the
spond to specific emotions, an idea that was expressed music expressed, a variety of emotions were reported with
already in ancient Greece and later in the Doctrine of Af- considerable interindividual variability, as in the earliest
fections, discussed earlier. All four theories differ with re- reports by Gilman (1891, 1892) and Downey (1897). Dow-
spect to scope: Langer is concerned with music in general; ney noted great differences in the extent to which individ-
Cooke primarily with melodic factors; Clynes with uals received definite impressions, and that good agree-
patterns of pitch, loudness, and duration; and Juslin solely ment may exist regarding the broad emotional quality but
with factors in the performance (e.g., timing). The theories less regarding nuances or variants of this quality. Large
also differ regarding the specificity with which they de- interindividual variation in perceived expression was also
scribe the emotions associated with different musical fac- found by Huber (1923), who asked musically trained lis-
tors: Langer is fairly unspecific, Clynes and Juslin consider teners to freely describe their impressions of short pitch
broad emotion categories (e.g., joy, sadness), and Cooke patterns—only two or three tones. He classified their re-
provides elaborate descriptions (e.g., “an outgoing, active, ports into (1) mood impressions; (2) impressions of human
assertion of joy”). The nature of the relationships between character described in emotional terms; (3) emotionally
musical factors and specific emotions is most determin- colored announcements (e.g., request, reproach, com-
istic in Cooke’s theory, fairly deterministic in Clynes’s the- plaint); (4) impressions of movement (e.g., rising, falling)
ory, but probabilistic in Juslin’s theory. or of activity; and (5) various inner images (e.g., land-
scape).
Later researchers, instead of using free reports, com-
Review of Empirical Research piled a list of descriptive terms—adjectives or nouns—and
asked listeners to choose those terms that were appropri-
ate for the music in question. Rigg (1937a) used this
Emotional Expression and Musical Structure
method as well as free reports and found that in some
Empirical research on emotional expression in music cases the supposed emotion was not frequently in the free
started in the late 19th century, but it was not until the description but more frequently chosen when presented
1930s that studies became more frequent. Usually, the pur- among the emotion terms. Thus the two methods gave
pose was to investigate listener agreement on perceived somewhat different results. Free descriptions are rare in
emotional expression, or to investigate what factors in the later studies but may have a certain revival (e.g., Imberty,
musical structure, as provided in musical notation, influ- 1979; Juslin, 1997c; Osborne, 1989).
enced perceived expression. Subjects listened to selected Gundlach (1935), Hampton (1945), Capurso (1952), and
applicable copyright law.

pieces of music or other tonal stimuli and reported per- Sopchak (1955) had listeners choose among descriptive
ceived emotional expression by means of either (1) free terms (see Table 26.1) and determined which of them were
phenomenological descriptions; (2) choice among descrip- chosen by at least 50% (in Gundlach, 45%) of the listeners
tive terms, adjectives or nouns, provided by the investi- for one or more of the musical pieces used. Best agreement
gator; or (3) ratings of how well such descriptive terms was found for various expressions of positive emotions
applied to the music in question. Techniques for contin- (e.g., gay, glad, happy, joyful, brilliant, triumphant), neg-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.1 Music Stimuli, Subjects, Response Types, and Descriptive Attributes in Empirical Studies of Expression in
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Music (FA ⫽ factor analysis, MDS ⫽ multidimensional scaling, CA ⫽ cluster analysis, MRA ⫽ multiple regression
analysis)

Author Stimuli Subjects Response Type Descriptive Attributes


Gilman 1891, 1892 11 classical pieces 28 musical ama- Free reports (an- Various: happiness, sadness, peace, fear,
teurs swers to ques- pain, desire, religious sentiment, yearn-
tions) ing, active and passive moods
Downey 1897 6 classical pieces 22 non musicians Free reports Various: happiness, sadness, religious
devotion, grief, hope
Huber 1923 Pitch patterns, 2–3 6–12 musicians Free reports Moods, characters, announcements, mo-
tones tion, various inner images
Sherman 1928 Single tones per- 30 psychology Free reports 14–18 emotions: e.g., sorrow, anger-
formed by singer students hate, fear-pain, surprise
Heinlein 1928 Triads, major/minor 30 mixed Either of → Happy/joyful, melancholy/sad
Gundlach 1932 334 Indian songs in Analysis of scores Songs for love, healing, war
scores
Gundlach 1935 40 classical pieces, 112 students Choice, any num- Brilliant, animated, uneasy, tranquil,
phrases ber out of 17 dignified, triumphant, exalted, glad,
terms → somber, melancholy, mournful, delicate,
whimsical, flippant, sentimental, awk-
ward, and grotesque
Hevner 1935a 10 classical pieces, 205 students Choice, any num- 14 adjective clusters (later reduced to 8
manipulated in ber out of → clusters, see Figure 26.4)
mode
Hevner 1936 26 classical, pieces, 450 students, Choice, any num- 8 adjective clusters, see Figure 26.4
manipulated in most of them ber out of →
rhythm, melodic musically un-
line, harmony trained
Hevner 1937 21 pieces, most 222 students Choice, any num- 8 adjective clusters, see Figure 26.4
classical, some spe- ber out of →
cially composed,
manipulated in
tempo and pitch
level
Rigg 1937a 20 phrases, classical 100 students Free reports ⫹ (Mainly) joy, lamentation, hopeful long-
and composed choice, one of → ing, sorrowful longing, love
Rigg 1939 5 composed phrases 105 ⫹ 99 male Choice, one of → (Mainly) pleasant/happy (e.g., joy,
systematically ma- students hopeful longing, love), serious/sad (e.g.,
nipulated lamentation, sorrowful longing)
Rigg 1940a 5 phrases, same as 84 psychology Choice, one of → (Mainly) pleasant/happy (e.g., joy,
Rigg 1939, varied in students hopeful longing, love), serious/sad (e.g.,
pitch level and to- lamentation, sorrowful longing)
nality
Rigg 1940b 5 phrases, same as 88 psychology Choice, one of → (Mainly) pleasant/happy (e.g., joy,
Rigg 1939, varied in students hopeful longing, love), serious/sad (e.g.,
tempo lamentation, sorrowful longing)
Campbell 1942 7 classical pieces, 7 40 seniors and Choice, one of → Gaiety, joy, yearning, sorrow, calm, as-
folk songs freshmen sertion, tenderness
Watson 1942 30 classical pieces 100 listeners at First and second Happy, mischievous, amusing, very
applicable copyright law.

each of six age choice among → happy, exciting, very exciting, dignified,
levels ⫹ 20 mu- kingly, peaceful, serious, pleading, sad,
sical experts tragic, mysterious, unclassified

(continued)

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.1 (continued)
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Author Stimuli Subjects Response Type Descriptive Attributes

Hampton 1945 10 classical pieces 58 female stu- Choice, any num- Rage, pity, determination, hate, despair,
dents ber of 30 terms → joy, terror, love, disgust, fear, surprise,
grief, suspense, longing, resignation,
sadness, devotion, horror, contempt, tri-
umph, praise, defiance, agony, and
seven others
Capurso 1952 105 classical pieces 1075 students Choice, one of → 6 categories: (a) happy, gay, triumphant;
(b) agitated, restless; (c) nostalgic,
soothing, relaxing; (d) prayerful, rever-
ent; (e) sad, grieving, lonely; (f) eerie,
weird, grotesque
Farnsworth 1954 56 classical ex- 200 students Choice, any num- 8 adjective clusters, see Figure 26.4
cerpts ber of adjectives

Sopchak 1955 5 classical pieces, 7 553 sophomore Choice, any num- Sorrow, joy, calm, yearning, love, eroti-
popular with lyrics, students ber of 12 terms → cism, jealousy, wonder, solemnity, cru-
3 folk tunes with elty, rage/anger, assertion
Russian lyrics
Kleinen 1968 35 classical pieces, 114 with varying 55 bipolar scales 5 factors: Heiterkeit-Ernst; Robustheit-
5 popular pieces musical knowl- ⫹ FA → Zartheit; Kosmos-Chaos; Phantasie-
edge → Einfallslosigkeit; Gefuhi-Geist
7 music experts 21 musicotechni-
→ cal scales
Wedin 1969 18 classical and 2 100 psychology 40 adjective rat- 3 factors: tension/energy-relaxation;
jazz pieces students, 100 ing scales ⫹ fac- gaiety-gloom, solemnity
music students tor analysis (FA)
Wedin 1972a 18 classical, 2 jazz, 26 music stu- 40 adjective rat- 3 factors: tension/energy-relaxation,
and 2 popular dents ing scales ⫹ fac- gaiety-gloom, solemnity-triviality
pieces tor analysis (FA)
Wedin 1972b 18 classical and 2 49 students Choice of 40 ad- 3 dimensions: tension/energy-relaxation;
jazz pieces jectives ⫹ multi- gaiety-gloom; solemnity
dimensional scal-
ing (MDS)
Wedin 1972c 40 classical, jazz, 100 psychology Choice of adjec- 3 dimensions: intensity-softness;
popular students → tives ⫹ MDS → pleasantness-unpleasantness; solemnity-
15 music experts Ratings on 13 triviality
musicotechnical
scales
Gabrielsson 1973 20 pieces of dance 23 musicians ⫹ 80 adjective rat- 5 factors: vital/gay-dull; simplicity-
music 24 nonmusicians ing scales ⫹ FA complexity; excited/intense-calm/soft;
knocking/hard-rocking/soft; solemn-
swinging
Batel 1976 30 classical 52 music stu- 55 adjective rat- 7 clusters: Ordnung; Prägnanz; Temper-
dents ing scales ⫹ clus- ament; Erregung; Fremdartigkeit; Ruhe;
ter analysis Empfindung
Kotlyar & Morozov 50 renditions of a 10 opera singers, Forced choice → Anger, fear, joy, sorrow ⫹ neutral
1976 vocal phrase 10 musically ⫹ performance
trained listeners measurements
applicable copyright law.

Hare 1977 16 classical ex- Students and Similarity ratings 3 dimensions: tempo/playfulness; po-
cerpts music students ⫹ MDS, adjective tency; musical period
ratings ⫹ FA

510
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.1 (continued)
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Author Stimuli Subjects Response Type Descriptive Attributes


Scherer & Oshinsky Synthesized 8-tone 48 psychology Ratings on 3 bi- Pleasantness-unpleasantness; activity-
1977 sequences, syste- students polar scales → ⫹ passivity; potency-weakness
matically varied choice of any Anger, fear, boredom, surprise, happi-
number of 7 emo- ness, sadness, disgust
tions →
Imberty 1979 16 pieces by De- 80 ⫹ 80 non- Free choice of 3 dimensions: Schèmes de tension et de
bussy, 16 by musicians any suitable ad- détente; Schèmes des résonance émoti-
Brahms jectives ⫹ corre- onelles; Représentations iconiques et ci-
spondence analy- nétiques
sis
Brown 1981 12 ⫹ 12 classical Nonmusicians, Arrange 12 pieces (A) Elegiac, gentle sorrow; (B) raptuous,
pieces instrumentalists into 6 pairs ⫽ ex- luminous; (C) tender, hushed; (D) rus-
and 19th century pressive catego- tic, good-humor; (E) spring morning
knowledgeables ries → buoyancy; (F) boisterous, rude, heavy;
and into 6 vari- (A) funereal, strong but sorrowful; (B)
ants of sadness sadness tinged with romantic mystery;
expression → (C) wistful, delicately regretful; (D) de-
pression, the “pits”; (E) poignant, plain-
tive; (F) relaxed, somber, reflective
Nielzén & Cesarec 13 specially com- 50 nonmusicians 20 bipolar scales 3 factors: tension-relaxation; gaiety-
1981 posed pieces ⫹ FA → gloom; attraction-repulsion
Nielzén & Cesarec 13 specially com- 75 students → 12 bipolar scales 3 factors: tension-relaxation; gaiety-
1982 posed ⫹ 13 classi- 8 music experts ⫹ FA → gloom; attraction-repulsion
cal pieces → 9 musicotechni- 3 factors: simple-sophisticated; vivid-
cal scales ⫹ FA placid; dark-light

Nielsen 1983, 1987 Haydn: London 11 musicians, 12 Continuous re- Tension
Symphony, R. high school stu- cording of per-
Strauss: Also dents ceived tension in
sprach Zarathustra music
Asmus 1985 3 pieces, popular, 2,057 students 99 adjective rat- 9 factors: evil, sensual, potency, hu-
classical ing scales ⫹ FA mour, pastoral, longing, depression,
sedative, activity
Crowder 1985 Sinewave triads, 24 young adults Choice of major Major-happy, minor-sad
major, minor or minor, happy
or sad
Senju & Ohgushi 10 performances of 1 performer, 16 Ratings on se- Weak, powerful, bright, sad, sophisti-
1987 Mendelsohn’s Vio- musically trained mantic differen- cated, beautiful, dreamy, fashionable,
lin Concerto listeners tial scales → ⫹ simple, deep
MDS
Cunningham & 30 classical pieces 112 children, Forced choice → Happy, sad, angry, afraid
Sterling 1988 ages 4, 5, 6 (18,
24)
Dolgin & Adelson 16 specially com- 128 children, Forced choice of Happy, sad, angry, frightened
1990 posed melodies ages 4, 7, 9 ⫹ drawings of faces
100 students →
Kastner & Crowder 12 short passages, 38 children, 3–12 Choice, one of Happy, contented, sad, angry
1990 major/minor years four faces →
Asada & Ohgushi Ravel’s Bolero 13 music stu- Ratings on 25 bi- 2 dimensions: energetic/powerful-gentle/
applicable copyright law.

1991 dents polar scales ⫹ soft; comic/coquetttish-serious/reserved


MDS →
(continued)

511
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.1 (continued)
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Author Stimuli Subjects Response Type Descriptive Attributes


Namba, Kuwano, 18 performances of 44 female and Continuous Triumphant, magnificent, stirring, pow-
Hatoh, & Kato 1991 Mussorgsky’s Pic- 454 male stu- choice among 15 erful, mild, quiet, leisurely, pastoral,
tures at an Exhibi- dents adjectives during calm, sorrowful, lonely, depressed, bril-
tion music → liant, delicate, smooth
Crowder, Reznick, Major and minor 14 infants, 6 Head-turning, di- Happy, sad (inferred from responses)
& Rosenkrantz 1991 chord, consonant months rection of gaze
and dissonant
chord
Terwogt & van 8 short classical 64 children, ages Forced choice of Happy, angry, fearful, sad
Grinsven 1991 musical extracts 5, 10; 32 adults drawings of faces
Thompson & Robi- Specially composed 14 moderately Ratings on 6 Joy, sorrow, excitement, dullness, anger,
taille 1992 pieces trained in music emotion scales → peace
Behrens & Green 24 improvisations 8 performers, 58 Ratings on 3 Angry, sad, scared
1993 on violin, trumpet, students scales →
timpani ⫹ vocals
Giomo 1993 12 short classical 173 children, Ratings, semantic Softness/intensity, unpleasantness/
compositions ages 5 and 9 differential → pleasantness, triviality/solemnity
Kratus 1993 30 excerpts from 658 children, Forced choice of Happy vs. sad (1st measure) Excited vs.
Bach’s Goldberg ages 6–12 facial drawings → calm (2nd measure)
Variations on piano
Madsen & Fredrick- Haydn (same as 40 musicians, 32 Continuous re- Tension
son 1993 Nielsen 1983) nonmusicians cording of per-
ceived tension
Baroni & Finarelli 3 phrases from 3 3 singers ⫹ 3 ac- Performance mea- Serene and joyful, sad and depressed,
1994 operas tors surements aggressive
Robazza, Macaluso, 8 classical pieces 40 adults, 40 Forced choice → Anger, happiness, sadness, fear
& D’Urso 1994 (piano) children
Gabrielsson & Lind- 20 synthesizer and 4 performers, 110 Forced choice → Angry, happy, sad, soft/tender, solemn,
ström 1995 sentograph perfor- listeners: stu- ⫹ performance indifferent
mances of 2 popu- dents, teachers measurements
lar pieces and musicians
Siegwart & Scherer 10, from 5 official 5 opera singers, Ratings of → ⫹ Madness, sadness, tender passion, fear
1995 recordings of opera 11 musically performance mea- of death, overall preference
performances trained listeners surements
Sundberg, Iwarsson, 15 excerpts, folk 1 opera singer, 11 Forced choice → Angry, happy, loving, sad, scared, secure
& Hagegård 1995 tunes, classical, musically trained ⫹ performance
opera listeners measurements
Gabrielsson & Jus- Performances of 4 9 performers, 93 Ratings on 7 Angry, happy, sad, fearful, tender, sol-
lin 1996 pieces (popular, university stu- scales → emn, expressive
folk, composed) on dents ⫹ performance
flute, violin, guitar, measurements
& singing voice
Gregory, Worrall, & 8 tunes, major/mi- 40 3–4-year olds, Choice between Happiness, sadness
Sarge 1996 nor 28 7–8-year olds two faces →
Krumhansl 1996 Mozart Piano So- 15 ⫹ 24, varying Continuous re- Segmentation, tension
nata, original and musical experi- cording of →
manipulated ence
applicable copyright law.

Ohgushi & Hattori Vocalise by Fauré 10 music stu- Ratings of → Anger, fear, joy, sorrow
1996a dents
Ohgushi & Hattori Vocalise by Fauré 3 female singers Performance mea- Anger, fear, joy, sorrow ⫹ neutral
1996b surements

512
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.1 (continued)
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Author Stimuli Subjects Response Type Descriptive Attributes

Rapoport 1996 60 opera arias and 7 professional Performance mea- Neutral-soft, calm, expressive, short,
lieder from official singers surements transitional-multistage, intermediate,
recordings excited, virtuoso
Krumhansl & Lynn Mozart minuet ⫹ 27 university stu- Continuous re- Segmentation, tension, new ideas, ex-
Schenk 1997 choreographed dents cording of → pressed emotion
dance
Krumhansl 1997 6 classical pieces 40 university stu- Continuous re- Sadness, happiness, fear, tension
dents cording of → ⫹
physiological
measurements
Fredrickson 1997 Haydn (same as 112 school-age Continuous re- Tension
Nielsen 1983) students cordings of →
Madsen 1997 2 classical pieces 50 ⫹ 48 students Continuous re- Valence (happiness-sadness) and
cording of → Arousal (activeness-passiveness)
Juslin 1997a 15 performances on 3 performers, 24 Ratings on 4 Anger, happiness, sadness, fear
electric guitar of a university stu- scales →
jazz piece dents ⫹ performance
measurements
Juslin 1997b 108 synthesized 54 university stu- Ratings on 6 Anger, happiness, sadness, fear, tender-
and systematically dents scales → ness, expressiveness
varied performances (MRA ⫹ CA) ⫹ Anger, happiness, sadness, fear, tender-
⫹ 10 performances forced choice → ness, expressiveness
of folk tune
Juslin 1997c 10 synthesized and 27 university stu- Free labeling ⫹ Anger, happiness, sadness, fear, tender-
live performances dents, musically forced choice → ness
of folk tune trained (first and second
choice)
Lindström 1997 “Frère Jacques” ma- 19 psychology Ratings on 6 bi- Happy-sad, angry-tender, expressive-
nipulated in students polar scales → unexpressive, stable-unstable, simple-
rhythm, melodic complex, tense-relaxed
contour, & direction
Adachi & Trehub Children’s songs 160 children, 4– Performance mea- Happy, sad
1998 12 years old (fi- surements ⫹ ob-
nal sample) server coding
Peretz, Gagnon, & 32 synthesized clas- Brain-injured Ratings of → Happiness, sadness
Bouchard 1998, sical pieces manip- subject and nor-
Exp 2 ulated in tempo mal subjects
and mode
Watt & Ash 1998 24 short excerpts, Undergraduate Forced choice in People traits, people states, movement,
classical, popular students 14 dichotomies terms rarely applied to people
referring to →
Balkwill & Thomp- 12 performances of 30 listeners Forced choice → Anger, joy, sadness, peace
son 1999 Hindustani ragas
Juslin & Madison 120 performances 20 university stu- Ratings on 4 Anger, happiness, sadness, fear
1999 (piano) of 2 pieces; dents scales →
folk tune, jazz (MRA) ⫹ perfor-
(elimination of mance measure-
cues) ments
applicable copyright law.

Schubert 1999a, b 4 classical pieces 25 musicians, 42 Continuous re- Valence (happiness-sadness) and
nonmusicians cording of → Arousal (activeness-sleepiness)

(continued)

513
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
514 PART IV. EXPRESSION OF EMOTION

Table 26.1 (continued)


Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Author Stimuli Subjects Response Type Descriptive Attributes


Juslin 2000 72 performances on 30 students, Ratings on 4 Anger, happiness, sadness, fear
electric guitar of 3 some musically scales →
pieces; folk, jazz trained (MRA) ⫹ perfor-
mance measure-
ments
Juslin & Laukka 240 performances of 8 performers, 50 Ratings on 4 Anger, happiness, sadness, fear
2000 3 pieces (jazz, folk, students scales →
popular) on electric (MRA) ⫹ perfor-
guitar mance measure-
ments
Laukka & Gabriels- 42 performances on 2 professional Ratings on 7 Angry, fearful, happy, sad, solemn,
son 2000 drums of swing, drummers, 13 scales → tender, no expression
beat, and waltz students ⫹ performance
measurements
Bresin & Friberg 14 synthesized per- 20 listeners, Forced choice → Anger, fear, happiness, sadness, solem-
(2000) formances of 2 some musically nity, tenderness, no expression
pieces; classical trained
and children’s song

ative emotions (e.g., grief, melancholy, mournful, sadness, joyful/sad, angry/pleased, gentle/violent—than for other
sorrow, despair), high arousal (e.g., agitated, angry, rest- terms (e.g., prickly). On the basis of this result they hy-
less, rage, uneasy, violent), low arousal (e.g., gentle, relax- pothesized that “the action of music is to mimic a person”
ing, soothing, tranquil), and various other feelings such as (p. 49). This is reminiscent of the notion that listeners per-
love, yearning, nostalgia, sentimentality, prayerfulness, ceive music as a “virtual persona” (e.g., Levinson, 1996).
reverence, delicateness, assertiveness, determination, dig- Studies in this section have provided examples of both
nity, solemnity, eeriness, weirdness, and grotesqueness. good and poor listener agreement. Good agreement may
Terms not reaching 50% agreement were (in alphabetical exist regarding a basic emotional quality but less regarding
order) cruelty, devotion, disgust, eroticism, flippancy, nuances or variants of this quality (Brown, 1981; Downey,
hate, horror, irritation, jealousy, pity, whimsy, and wor- 1897; cf. also Campbell, 1942); appreciation of nuances
ship. may require good knowledge of the musical style (Brown,
Brown (1981) studied listeners’ ability to recognize 1981). Of course, results may depend on the selection of
similar emotional expression in pieces from different music and on the response method (cf. Rigg, 1937a). Gen-
styles and genres in classical music. He chose 12 musical der of the listener, on the other hand, did not seem to be
excerpts and asked listeners to order them into six pairs important (e.g., Sopchak, 1955).
according to six different emotional categories (see Table
26.1). In another task, listeners instead had to identify six
Accuracy
pairs out of 12 other musical excerpts representing
“twelve variations on sadness”—that is, within the same Accuracy refers to listeners’ “correct” recognition of emo-
broad emotion category. While listeners were fairly suc- tional expression according to some independent criterion
cessful in the first task, they were not so in the second (e.g., composer’s intention, expert judgment). Rigg (1937a)
task, until Brown supplied his own descriptions of the six used short phrases supposed to express joy, lamentation,
sadness categories (see Table 26.1); however, nonmusi- longing, and love. The expression of joy was correctly rec-
cians were still unsuccessful. Brown thus concluded that ognized by the listeners, but the recognition of the other
if the different expressions are not too similar (as in the expressions was far less accurate. Campbell (1942) se-
first task), synonymous pairs in expression can be identi- lected pieces to represent seven emotions (see Table 26.1).
fied even by persons not highly knowledgeable about clas- The subjects were asked to associate each piece with one
sical music. However, with pieces as close in expression of the seven emotions. Agreement among listeners and
applicable copyright law.

as in the variations on sadness, “the agreement on syn- agreement with Campbell’s emotion designations were
onymous pairs can only be achieved by listeners highly very high for gaiety, joy, and assertion; lower but still well
conversant with the traditions involved” (p. 264). above chance level for calm, sorrow, tenderness, and
Watt and Ash (1998) found higher intersubject agree- yearning. Yearning could be confused with sorrow, calm,
ment in choice of descriptive terms that can refer to “peo- and tenderness; and calm with tenderness and sorrow. Dif-
ple traits” or states—male/female, young/old, good/evil, ferent amounts of musical training gave no difference in

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 515

results. Watson (1942), using musical experts’ judgments among ready-made clusters of descriptive terms, now
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

as criterion for the intended expression in selected pieces gathered a large number of descriptive terms, asked listen-
of music, found that even children in 6th and 8th grades ers to provide numerical ratings of them, and then let mul-
showed good (⬎50%) agreement with the criterion, best tivariate analysis of the ratings show what terms belonged
for expressions such as mischievous, happy, exciting, together. Rating scales were either bipolar—such as happy
kingly, dignified, and sad. Results were worse for expres- versus sad, excited versus calm—as in the semantic dif-
sions such as pleading, tragic, and mysterious, for which ferential technique, or unipolar in cases where exact op-
50% agreement was not reached until respondents were posites were hard to find.
the age of college students. The ability to discriminate mu- Despite large variation regarding musical stimuli, de-
sical expressions had no relation to general intelligence, scriptive terms, and number and type of subjects, several
only low correlations with scores on the Seashore Mea- studies using factor analysis (mostly), correspondence
sures of Musical Talents (Seashore, Lewis, & Saetveit, analysis (Imberty, 1979), or multidimensional scaling re-
1939/1960), and no relation to musical training except at sulted in two common fundamental dimensions. One re-
higher age levels. Robazza, Macaluso, and D’Urso (1994) flected positive versus negative emotions: “Heiterkeit–
asked adults and children (9–10 years) to decode the ex- Ernst” (cheerful–serious; Kleinen, 1968; see Figure 26.1);
pression of eight pieces of classical piano music (selected gaiety–gloom (Wedin, 1969, 1972b; see Figure 26.2; Niel-
by music experts to express anger, happiness, sadness, and zén & Cesarec, 1981); pleasantness–unpleasantness
fear) by means of forced choice. The results showed high (Wedin, 1972c); vital, gay, playful versus dull, restrained,
overall accuracy for happiness and sadness (about 75% and hesitating (Gabrielsson, 1973, experiments 4–5); and
correct), but lower accuracy for anger (28%) and fear “Les schèmes de résonances émotionelles” (positive vs.
(39%). There were no differences between listeners with negative emotions; Imberty, 1979).
regard to musical experience or gender. In a study by Another dimension mainly reflected arousal or po-
Thompson and Robitaille (1992), listeners moderately tency, or both: “Robustheit–Zartheit” (strong/powerful–
trained in music were able to recognize expressions of joy, soft/tender; Kleinen, 1968); tension/energy (e.g., tense vs.
sorrow, excitement, anger, and dullness in melodies spe- relaxed, vehement vs. mild, aggressive vs. gentle; Wedin,
cially composed to carry these expressions; partial con- 1969, 1972b); intensity–softness (Wedin, 1972c); aggres-
fusion occurred between expressions for peace and sor- sive, excited, intense vs. calm and soft (Gabrielsson, 1973);
row. tension–relaxation (Nielzén & Cesarec, 1981); and “Les
A special case of musical communication concerns pro- schèmes de tension et de détente” (tension vs. relaxation;
gram music (see earlier, “Absolute Music”), which aims at Imberty, 1979).
describing events or situations as well as associated emo- Thus, there is a good agreement on these emotion di-
tions. If listeners do not know the intended program, they mensions in several studies. The configuration of descrip-
are usually unable to provide a “correct” description of tive terms within the two-dimensional plane may be strik-
events or situations, but they are somewhat better at per- ingly similar, as seen in a comparison of Figures 26.1 and
ceiving the intended emotional expression (Batel, 1976; 26.2 (and slight rotation of the axes may increase the sim-
Brown, 1981; Downey, 1897; Foros, 1975; Hampton, 1945; ilarity further). Note that most expressions studied so far
Osborne, 1989; Rigg, 1937b; Weld, 1912); Osborne’s study belong to only two of the quadrants, which are opposite
provided the most positive outcome. each other. This might reflect the choice of musical stimuli
Taken together, these studies indicate good recognition and descriptive terms in the studies.
of intended emotional expressions for positive emotions Dimensions similar to those just mentioned also ap-
(gaiety, happiness, joy, mischievous), negative emotions peared in other studies: tempo/playfulness and potency
(sadness, sorrow; anger-hate; dullness), high arousal (ex- (Hare, 1977); humor (amused, playful), depression (sad,
citement), low arousal (calm), tenderness, and dignity. gloomy), evil (anger, hate), activity (determined, vigorous),
Confusions may occur—for instance, yearning may be pastoral (calm, relaxed), sedative (serene, tranquil), po-
confused with sorrow, calm, and tenderness; calm with tency (heroic, majestic), all in Asmus (1985); “tempera-
tenderness and sorrow; and peace with sorrow. ment” (merry, playful), “Erregung” (excited, dramatic),
“Ruhe” (calm, tired) in Batel (1976, cluster analysis); and
a dimension extending from gentle and soft to energetic
Fundamental Dimensions in Emotional Expression
and powerful, reflecting the successive crescendo in Ra-
applicable copyright law.

Following the introduction of the semantic differential vel’s Bolero (Asada & Ohgushi, 1991). Solemnity–triviality
technique (Osgood, Suci, & Tannenbaum, 1957) and the may be an additional, independent dimension (Wedin,
development of efficient computer programs for multivar- 1969, 1972a, 1972b, 1972c).
iate statistical techniques—factor analysis, cluster analy- Although not present among the descriptors used in
sis, correspondence analysis, multidimensional scaling— Figures 26.1 and 26.2, other feelings—love, tenderness,
investigators, rather than asking for listeners’ choices nostalgic, sentimental, prayerful, reverent, delicate, asser-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
516 PART IV. EXPRESSION OF EMOTION
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Figure 26.1 The first two dimensions obtained in Kleinen’s (1968) factor analysis. Re-
produced by permission of the author.

tion, determination, eerie, weird, and grotesque—could be waves of tension with periods of relaxation in between; in
placed within the two-dimensional plane. Some confu- the development there was a practically continuous in-
sions among expressions—for instance, yearning confused crease of tension up to its end, after which the pattern
with sorrow or calm, calm or peace with sorrow—seem from the exposition returned in the repetition. The average
understandable from their positions close to each other in curve of the inexperienced group looked similar to this
Figures 26.1 and 26.2. already in their first listening, and the similarity increased
in a second listening.
Nielsen’s results have been replicated by Madsen and
Continuous Recording
Fredrickson (1993), using another response device (CRDI,
As emotional expression may vary during a piece of mu- Continuous Response Digital Interface). Fredrickson
sic, researchers have recently devised procedures for con- (1997) found high correlation between young students’
tinuous recording of perceived expression. Nielsen (1983, (grades 2, 5, 8, and 11/12) continuous ratings of perceived
1987) pioneered this work having listeners—one group of tension in the Haydn movement. Krumhansl (1996) had
musically experienced listeners and another group of listeners adjust a slider at the center of the computer
young nonmusical students—continuously press a pair of screen to indicate variations in tension in Mozart’s Piano
tongs in proportion to perceived tension in the music (the Sonata K. 282, and to indicate experienced expression of
more tension, the harder press, and vice versa), when lis- happiness, sadness, fear, and tension in six excerpts of
tening to the first movement of Haydn’s London Sym- classical music (Krumhansl, 1997). Krumhansl and Lynn
phony and to the initial part of Richard Strauss’s Also Schenk (1997) studied relationships between dance and
applicable copyright law.

Sprach Zarathustra. The pressures were transformed to music by asking students to use a foot-switch to indicate
electric voltage and registered on a polygraph. The visible the amount of tension in dance performance to a Mozart
result was thus waves of tension and relaxation, such as minuet.
in Figure 26.3, which shows these waves for the Haydn Namba, Kuwano, Hatoh, and Kato (1991) assigned a
movement, averaged across the musically experienced lis- separate key on a computer keyboard to each of 15 adjec-
teners. In the exposition of the movement, there were tives (see Table 26.1). While listening to Mussorgsky’s Pic-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 517
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Figure 26.2 Wedin’s (1972a) three-dimensional model of emotional expression in mu-


sic. Reproduced by permission of the author.

tures at an Exhibition, the subjects successively pressed tive two versions of a piece to the manipulated variable.
the different keys according to how well the respective Response variables were eight clusters of descriptive terms
adjectives described the perceived expression. Schubert in a circular configuration, an “adjective circle” (see Fig-
(1999a, 1999b), in connection to results from multivariate ure 26.4). This has an obvious, although not perfect, sim-
analyses, had his subjects successively move the computer ilarity to the results from later multivariate analyses (cf.
cursor to various positions in a two-dimensional emotion Figures 26.1 and 26.2): the vertical axis (cluster 6 vs. clus-
space with two main axes—valence (horizontal) and ter 2) corresponds to gaiety versus gloom. The horizontal
arousal (vertical)—according to perceived expression. The axis (cluster 8 to cluster 4) has a certain similarity with
subjects’ responses to four pieces of classical instrumental tension/energy versus relaxation, and the similarity seems
music were studied in relation to variations of various fac- even greater if one considers clusters 7 and 8 together (this
tors in the musical structure using time series and regres- would correspond to tension/energy) versus clusters 3 and
sion analyses. A similar device for continuous recording 4 together (relaxation). Although Farnsworth (1954)
of emotional responses to music was developed by Mad- slightly modified Hevner’s clusters and abandoned the cir-
sen (1997). For a more extensive review of continuous re- cular representation, Hevner’s adjective circle has ap-
sponse methodology, see Schubert (2001). peared again in later investigations (e.g., Batel, 1976; Greg-
ory & Varney, 1996).
In Hevner’s studies, listeners (usually hundreds of stu-
Determinants of Perceived Expression
dents) were instructed to mark as many of the terms in the
Which factors in the musical structure influence the per- adjective circle as they found appropriate for the music in
ceived emotional expression? This question has been in- question. Variables with largest effects on their choice
vestigated using both experimental and post hoc analyses. turned out to be tempo and mode, followed by pitch level,
harmony, and rhythm, whereas melodic direction had lit-
Experimental approaches. Hevner introduced experimental tle, if any, effect. Fast tempos elicited judgments in the
manipulation using short pieces of tonal music. Besides happy-gay and exciting-restless clusters (clusters 6–7), and
applicable copyright law.

the original version of a piece, she created a variant dif- to a lesser extent in vigorous and graceful clusters (clusters
fering in mode (Hevner 1935a), melodic direction, har- 8 and 5), slow tempos in the dignified, sad, sentimental-
mony, and rhythm (Hevner, 1936), tempo, and pitch level tender, and calm-serene clusters (clusters 1–4). Major
(Hevner, 1937). By manipulating one variable at a time mode was associated with happiness, gaiety, playfulness,
while keeping all others constant, researchers could as- and sprightliness (clusters 5–6), minor mode with sadness,
cribe any difference in listeners’ judgments of the respec- sentimental, yearning, and tender effects (clusters 2–3).

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or
applicable copyright law.

Figure 26.3 Tension curves from listeners’ continuous rating of tension in Haydn’s Lon-
don Symphony, first movement, from Nielsen (1983) (spænding ⫽ tension, Exp ⫽ ex-
position, Gf ⫽ development, Rp ⫽ recapitulation, horizontal axis shows bar numbers,
vertical axis shows exerted hand pressure, bold curve ⫽ first listening, dotted curve ⫽
second listening). Reproduced by permission of the author and Akademisk Forlag, Co-
penhagen.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 519

VI
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

bright
cheerful
gay
happy
joyous
merry
VII V
agitated delicate
dramatic fanciful
exciting graceful
exhilarated humorous
impetuous light
passionate playful
restless quaint
sensational sprightly
soaring whimsical
triumphant
VIII IV
emphatic calm
exalting leisurely
majestic lyrical
martial quiet
ponderous satisfying
robust serene
vigorous soothing
tranquil
I III
awe-inspiring dreamy
dignified longing
lofty plaintive
sacred pleading
serious sentimental
sober tender
solemn yearning
spiritual yielding
II
dark
depressing
doleful
frustrated
gloomy
heavy
melancholy
mournful
pathetic
sad
tragic

Figure 26.4 Hevner’s adjective circle. Adjectives within each cluster appear in alpha-
betical order. Adjectives used by Hevner (1936, 1937) to represent the respective clus-
ters are underlined. Adapted from Farnsworth (1954).

Higher pitch level had greatest effect on the sprightly- on the Seashore Measures of Musical Talents (Seashore et
humorous cluster (cluster 5), lower pitch level on the sad, al., 1939/1960).
vigorous-majestic, and dignified-serious qualities (clusters Hevner (1935b) emphasized the necessity of using
2, 8, and 1). Simple harmonies were characterized as pieces with similar emotional expression throughout the
applicable copyright law.

happy, graceful, serene, and sentimental (clusters 6, 5, 4, piece. If the piece contains different expressions in differ-
3), complex dissonant harmonies as exciting and to some ent parts, the resulting judgments may look very peculiar,
extent both sad and vigorous (clusters 7, 2, 8). Flowing even contradictory. She further warned against drawing
rhythm was light, happy, graceful, and sentimental (clus- too far-reaching conclusions from these data, pointing out
ters 5, 6, 3), firm rhythm dignified (cluster 1). These results their possible dependency on the selected pieces, diffi-
were largely independent of musical training and of scores culties in constructing adequate examples, and, above all,

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
520 PART IV. EXPRESSION OF EMOTION

the fact that emotional expression usually results from hansl, 1996, 1997; Nielsen, 1983), using the score to con-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

several musical factors in complex interplay. struct indices for formal complexity and dynamism (Im-
Experimental manipulation was also applied by Rigg berty, 1979), or using musical experts as judges (Kleinen,
(1939), who composed and manipulated five four-bar 1968; Watson, 1942; Wedin, 1972c). For instance, Wedin
phrases regarding tempo, mode, articulation (legato, stac- (1972c) had musical experts rate various musico-technical
cato), pitch level, loudness, rhythm (iambic, trochaic), and aspects of the pieces used, and their relations to the emo-
certain intervals; further transposed the same phrases up tion dimensions were investigated by multiple regression
or down (Rigg, 1940a); and had the same five phrases analysis. Intensity–softness was most affected by articu-
played by a pianist at six different tempos (Rigg, 1940b; lation (staccato–legato) and intensity (forte–piano); the
Rigg, 1964, provided a useful review of his work as well combination of staccato and forte expressed activity and
as that of Gundlach, Hevner, and Watson). Scherer and energy, whereas legato and piano made the character soft.
Oshinsky (1977) synthesized eight-tone sequences that In the pleasantness–unpleasantness or gaiety–gloom di-
were manipulated in two levels concerning amplitude var- mension, the pleasant–glad pole was determined by con-
iation, pitch level, pitch contour, pitch variation, tempo, sonance, fluent and fast rhythm, major mode, and high
envelope, and filtration cut-off level. Multiple regression pitch; the opposite pole was determined by dissonance,
analysis was used to estimate the predictive strength of firm and slow rhythm, minor mode, and low pitch. Solem-
each acoustic parameter on each response dimension (see nity was achieved by low pitch, slow tempo, and forte.
Table 26.1). Juslin (1997b, experiment 2) used synthesized
versions of “Nobody Knows,” manipulating tempo (slow, Summary of Results. Results from these studies are sum-
medium, fast), sound level (low, medium, high), frequency marized in Table 26.2 in terms of how the most common
spectrum (soft, bright, sharp), articulation (legato, stac- emotions are related to variables in musical structure as
cato), and tone attack (slow, fast). Students rated these ver- studied by different investigators. Every expressed emo-
sions on six adjective scales: happy, sad, angry, fearful, tion is related to a large number of structural variables,
tender, and expressive. Multiple regression analyses whose relative importance may be roughly proportional to
showed that the variance accounted for (R2) by the manip- the number of investigators cited for each structural factor;
ulated factors was typically 0.78–0.88. Lindström (1997) for some emotions (boredom, disgust, surprise) studied by
similarly used synthesis to manipulate rhythm, melodic a single investigator, factors are listed in importance ac-
contour, and melodic direction in different versions of cording to the respective author. Tempo, mode,
“Frère Jacques” and asked listeners to rate them on six consonance-dissonance, loudness, and pitch recur for
bipolar scales (e.g., happy-sad, angry-tender, tense- most emotions, whereas factors related to melody, rhythm,
relaxed). The results revealed complex interactions among timbre, articulation, and envelope appear less regularly.
the manipulated factors in their effects on listeners’ rat- Various examples of how emotions have been expressed
ings. Peretz, Gagnon, and Bouchard (1998, experiment 2) in classical music using these means can be found in de
manipulated tempo and mode in synthesized versions of la Motte-Haber (1996, pp. 72–77).
32 excerpts of classical music and found that both factors The combination of positive valence and high arousal
affected both normal listeners’ and a brain-injured lis- is associated with fast tempo, high loudness, high pitch
tener’s perception of happiness and sadness in similar level, large pitch variation, staccato articulation, bright or
ways. sharp timbre, and sharp envelopes, while the two dimen-
Thompson and Robitaille (1992) asked five composers sions may differ with regard to mode (positive valence
to compose short melodies to express joy, sorrow, excite- rather in major mode, high arousal may be in either
ment, dullness, anger, and peace. Listeners successfully mode), in harmony (positive valence apt to consonance
recognized most of these expressions (see “Accuracy,” ear- and simple harmony, high arousal may be dissonant and
lier). Joyful melodies were strongly tonal and rhythmically contain more complex harmony), rhythm (positive va-
varied. Sad melodies were slow with implied minor or lence fluent rhythm, high arousal more complex rhythm),
chromatic harmony. Melodies for excitement were fast and formal complexity (positive valence less complex
and contained intervallic leaps and high pitches. Dull mel- than high arousal). The combination of negative valence
odies were tonal in stepwise motion. Angry melodies were and low arousal is associated with slow tempo, soft loud-
rhythmically complex and with implied chromatic har- ness, low pitch level, small pitch variation, legato articu-
mony or atonality. Melodies for peacefulness were tonal, lation, dark timbre, and round envelopes, while the two
applicable copyright law.

slow, and often involved stepwise motion leading to me- dimensions may differ with regard to mode (negative va-
lodic leaps. lence usually in minor mode, low arousal may be in either
mode), harmony (negative valence more dissonant than
Post Hoc Analyses. Variables in the musical structure re- low arousal), and formal complexity (negative valence
lated to perceived expressions have been studied post hoc more complex than low arousal).
using the score of the music (e.g., Gundlach, 1935; Krum- With regard to mode, the common association between

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.2 Summary of Determinants of Perceived Emotional Expression in Music
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Note. Authors’ names are abbreviated to the initial two letters of the (first) author, publication year to the two last digits (e.g.,
He36 ⫽ Hevner, 1936). References within parentheses refer to performance studies (Round envelope ⫽ slow tone attack/slow
decay, sharp envelope ⫽ fast tone attack/fast decay).

Emotion Related variables in musical structure


Happiness, Joy, Fast tempo Gu35, He37, Ri40b Wa42, Ri64, Kl68, Mo68, We72c, Ga73, Sc77, Ni82, Th92, Ju97b, Kr97,
Gaiety Pe98, (Ba94), (Ga/Li95), (Ga96), (Ju97a), (Ad98), (Ju99), (Ba99), (Ju00), (Ju/La00), (La00)
Moderate tempo variation (Ga/Li95), (Ju99)
Major mode He36, Ri39, Kl68, We72c, Sc77, Cr85, Kr97, Pe98
Consonance He36, Ri39, Wa42, We72c
High loudness Ri39, Ni82, (Ko76), (Ga/Li95), (Su95), (Ga96), (Oh96b), (Ju97a), (Ad98), (Ju99), (Ju00, (Ju/
La00)
Small loudness variation Wa42, Sc77, (Ju99)
High pitch He37, Ri40a, Kl68, We72c, (Ad98)
Large pitch variation/pitch range Gu35, Sc77, (Ba99)
Ascending melody Ge95, rising intonation (Ra 96)
Rhythm: regular Gu35, Wa42, varied Th92, fluent He36, We72c
Staccato articulation Ri39, Ni82, Ju97b, (Ga/Li95), (Ga96), (Ra96), (Ju99), (Ju00)
Large articulation variation (Ju99), (Ju00), (Ju/La00)
Many harmonics/bright timbre (Ga96), (Ju/La00)
Sharp envelope Sc77, (Ko76), (Ga96)
Moderate timing variation (Ga96), (Ju/La00)
Sharpened contrasts between long and short notes (Ga96), (La00)
Low formal complexity and average dynamism Im79
Sadness/Gloom Slow tempo He37, Ri40b, Wa42, Ri64, Kl68, Be72, We72c, Sc77, Ni82, Th92, Ju97b, Kr97, Pe98,
(Ko76), (Ba94), (Ga96), (Oh96b), (Ju97a), (Ad98), (Ju99), (Ba99), (Ju00), (Ju/La00), (La00)
Minor mode He36, Ri39, Ri64, Kl68, We72c, Cr85, Kr97, Pe98
Dissonance He36, Wa42, Ri64, We72c
Soft loudness Gu35, Ju97b, (Ko76), (Ba94), (Ga96), (Oh96b), (Ju97a), (Ad98), (Ju99), (Ju00), (Ju/La00),
(La00)
Loudness variation: small Wa42, moderate (Ju99)
Low pitch Gu35, He37, Ri40a, Wa42, Ri64, We72c, Sc77, (Ad98)
Narrow pitch range Gu35, (Ba99)
Descending melody/pitch contour Sc77, Ge95
Firm rhythm He36, We72c
Legato articulation Ri39, Ju97b, (Ga96), (Ju97a), (Ju99), (Ju00), (Ju/La00)
Little articulation variation (Ju99), (Ju00), (Ju/La00)
Few harmonics/soft timbre Sc77, Ju97b, (Ju00), (Ju/La00)
Round envelope Sc77, Ju97b, (Ko76), (Ga96)
Large timing variation (Ga96), (Ju/La00)
Softened contrasts between long and short notes (Ga96), (La00)
Slow vibrato (Ga96), (Ko00)
High formal complexity combined with low dynamism Im79
Activity/Energy/ Fast tempo Gu35, He37, Ri40b, Wa42, Ri64, Sc77, Th92
Excitement Dissonance He36, Ri39, Wa42, Ri64
High loudness Wa42, Ri64, We72c
Small loudness variation Sc77
Pitch: high Wa42, Sc77, low He37, Ri40a
Large pitch variation Sc77
Intervallic leaps Th92
Staccato articulation Ri39, We72c
Many harmonics Sc77
Sharp envelope Sc77
Large vibrato extent (Wa32), (Ra96)
Potency Fast tempo Sc77
applicable copyright law.

High loudness Kl68


High pitch Sc77
Rising pitch contour Sc77
Many harmonics Sc77
Round envelope Sc77

(continued)

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
Table 26.2 (continued)
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Emotion Related variables in musical structure


Tension Dissonance Ni82a, Ni83
High sound level, dynamics Ni83, Kr96
Ascending melody Ni83, Kr96
Increased note density Ni83, Kr96
Harmonic complexity Ni83
Rhythmic complexity Ni82, Ni83
Lack of melody Ni82
Various formal properties Ni83, Kr96
Relaxation/Calm/ Slow tempo Gu35, He37, Ri64, (Ba99)
Softness/Peace Consonance He36, Ri64, Li97
Soft loudness Wa42, K168, We72c
Pitch: high He37, low Gu35
Narrow pitch range Gu35, (Ba99)
Rhythm: regular Wa42, flowing He36
Legato articulation We72c
Small vibrato extent (Wa32), (Ra96)
Low formal complexity Ni83
Solemnity/Dignity Slow tempo Gu35, He37, Ri40b, Ri64, We72c, (Ga/Li95), (Ga96)
Mode: major Ri39, minor He36
Consonance He36, Ri39, Wa42, Ri64
Loudness: high We72c, moderate or high (Ga/Li95), (Ga96)
Few loudness changes Wa42
Low pitch He37, Ri40a, We72c
Narrow melodic range Gu35
Rhythm: regular Wa42, Ri64, firm He36
Legato articulation Ri39, (Ga/Li95)
Sharp envelope (Ga96)
Small timing variation (Ga96)
Anger Fast tempo Sc77, Ju97b, (Ko76), (Ba94), (Ga/Li95), (Su95), (Ga96), (Oh96b), (Ju97a), (Ju99), (Ju00), (Ju/
La00), (La00)
Minor mode Sc77
Dissonance Li97
High loudness Ju97b, (Ko76), (Ba94), (Ga/Li95), (Su95), (Ga96), (Oh96b), (Ra96), (Ju97a), (Ju99), (Ju00),
(Ju/La00), (La00)
Little loudness variation (Ju99)
Pitch: high Sc77
Small pitch variation Sc77
Ascending pitch contour Sc77
Complex rhythm Th92
Articulation: staccato Ju97b, (Ga/Li95), (Ga96), (Ju99), legato (Ju97a), (Ju00)
Moderate articulation variation (Ju99), (Ju00)
Many harmonics/sharp timbre Sc77, Ju97b, (Ga96), (Ju00), (Ju/La00)
Sharp envelope Ju97b, (Ko76), (Ga96)
Small timing variation (Ju/La00)
Sharpened contrasts between long and short notes (Ga96)
Large vibrato extent (Me32), (Ko00)
High complexity combined with high dynamism Im79
Fear Tempo: fast Sc77, (Ko76), (Oh96b), (Ju99), (La00), slow (Ju97a), (Ju00)
Large tempo variation (Ga96), (Ju99), (La00)
Dissonance Kr97
Loudness: soft Ju97b, (Ko76), (Ga96), (Ju97a), (Ju00), (Ju/La00), (La00), moderate (Oh96b)
Large loudness variation Sc77, (Ga96), (Ju99)
Rapid loudness changes Kr97
applicable copyright law.

Pitch: high level, ascending contour, small variation Sc77


Large pitch contrasts Kr97
Staccato articulation Ju97b, (Ga96), (Ju97a), (Ju/La00)
Large articulation variation (Ju99), (Ju/La00)
Harmonics: many Sc77, few (Ju00), (Ju/La00)
Large timing variation (Ga96), (Ju/La00), (La00)
Envelope: round Sc77, Ju97b, sharp (Ko76)
Vibrato: irregular (Ra96), (Ga96), fast (Ko00)

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 523

Table 26.2 (continued)


Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Emotion Related variables in musical structure


Tenderness/Love Tempo: slow Ju97b, (Ga/Li95), (Su95), (Ga96), moderate (La00)
Soft loudness Ju97b, (Ga/Li95), (Su95), (Ga96), (La00)
Rising intonation (Ra96)
Little loudness variation (Ga/Li95)
Legato articulation Ju97b, (Ga/Li95), (Ga96)
Round envelope Ju97b (Ga96)
Few harmonics/soft timbre Ju97b (Ga96)
Large timing variation (Ga96)
Softened contrasts between long and short notes (Ga/Li95), (Ga96)
Boredom Slow tempo, low pitch level, few harmonics, descending pitch contour, round envelope, small pitch
variation Sc77
Disgust Many harmonics, small pitch variation, round envelope, slow tempo Sc77
Surprise Fast tempo, high pitch level, ascending pitch contour, sharp envelope, many harmonics, large pitch
variation Sc77

major mode and happiness and between minor mode and sociated with a particular emotional expression. Gabriel
sadness was questioned by Heinlein (1928). Listening to (1978) generated the 16 basic terms using sinusoidals, con-
isolated major and minor chords, his subjects gave many stant tempo, and uniform rhythm. Students listened to
“erroneous” responses—that is, sad-type responses to ma- these “reduced” basic terms and were given a response
jor chords and happy-type responses to minor chords (cf. sheet with Cooke’s description of each term. Their task
discussion in Crowder, 1984). He also found that loud was to rate the agreement between each basic term and
chords and chords at high pitch level evoked more happy- the description on a five-point scale (5 ⫽ perfect match, 1
type responses than soft chords and chords at low pitch ⫽ no match at all). In one condition the description of
level, irrespective of mode. However, Crowder (1985) each term was according to Cooke, but in another condi-
found the conventional major-happy and minor-sad asso- tion Cooke’s descriptions were distributed randomly
ciations in young adults who listened to sine-wave triads, across the basic terms. It turned out that the mean agree-
and these associations appeared in many of the investi- ment ratings were about the same under both conditions,
gations using real music (Table 26.2). See also children’s thus failing to support Cooke’s theory. However, this ex-
responses in the section on “Developmental Studies.” periment has, in turn, been discussed and criticized re-
This summary should preferably be supplemented by garding the choice of sinusoidal stimuli, the lack of mu-
the review in Gabrielsson and Lindström (2001), which sical context, and the use of musically untrained listeners
summarizes the same results but in terms of the effects (Cazden, 1979; Gabriel, 1979; Hampson, 2000; Kaminska
each structural factor has on respective emotion judgment. & Woolf, 2000; Nettheim, 1979; Sloboda, 1985, pp. 60–64).
It also points out several gaps and uncertainties in re- Hampson (2000) and Kaminska and Woolf (2000) again
search so far, especially regarding the effects of timbre, tried to test the validity of Cooke’s theory. Hampson used
intervals, melody, rhythm, harmony, and properties of mu- selected pieces of 19th-century piano music (e.g., by Cho-
sical form. The most obvious gap concerns potential in- pin, Schumann), thus real music, featuring nine of Cooke’s
teractions between different factors. It has been repeatedly basic terms. Musically experienced listeners rated on a
emphasized (e.g., Hevner, 1935b, 1937; Rigg, 1964) that no five-point scale how well Cooke’s descriptions matched
factor works in isolation; its effects are highly dependent their own perception of the emotion in the respective
on what other factors and levels are present in the actual piece. Kaminska and Woolf first reduced Cooke’s 16 basic
context. Despite these problems, the findings clearly sup- terms to four bipolar dimensions (joy–sorrow, assertive-
port the general assumption made in most theories of ex- ness–submissiveness, finality–continuation, outburst–con-
pression in music—namely, that specific patterns of mu- stancy) and then had nonmusicians listen to the basic
sical factors correspond to specific emotions. terms using sampled piano sound and keeping tempo, vol-
ume, and articulation constant (similar to Gabriel’s ap-
applicable copyright law.

proach). The listeners’ task was to rate each basic term on


Tests of Theories
each of the four bipolar scales. Although these two studies
Cooke’s Theory. Some attempts have been made to test the used different stimuli and procedures, both revealed that
validity of Cooke’s (1959) theory. As explained in the sec- the different melodic lines in Cooke’s basic terms carry
tion “Modern Theories of Emotional Expression in Mu- different emotional meanings, but not in the precise way
sic,” Cooke postulated 16 so-called basic terms, each as- claimed by Cooke. Only one of Cooke’s statements was

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
524 PART IV. EXPRESSION OF EMOTION

confirmed, namely, the association of major and minor of music and push the disk on a sentograph at the metrical
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

mode to positive and negative emotion, respectively, al- accents of the music, “the way you do when tapping along
ready known from other research. with a piece of music” (p. 48). After compensating for in-
However, the effect of “melodic line” on emotional ex- dividual differences in exerted force and calculating val-
pression cannot be separated from other factors because ues for five parameters of the sentograms, de Vries found
melody always implies some kind of rhythm (even if tones that sentograms were different for different pieces, and
have equal duration), further a certain tempo, loudness, that there were similarities between the sentograms for
timbre, articulation, and so on. Variables in the musical pieces with a certain emotional expression (anger, grief,
structure may interact in many different ways (Gabrielsson joy, love) and Clynes’s sentograms for the corresponding
& Lindström, 2001; Hevner, 1935b, 1937; Rigg, 1964), and emotions. Some (less formalized) evidence in the same di-
their relative importance may therefore vary with different rection was found in musicians’ performances on the sen-
contexts. tograph (without sound) of the same piece of music with
different emotional expressions; the resulting pressure
Clynes’s Theory. Clynes’s suggestion to use the sentograph patterns showed some similarities with Clynes’s sento-
to study emotional expression in music was followed up grams for the corresponding emotions (Gabrielsson &
in his own studies. This gave some indications of pressure Lindström, 1995). In sum, it seems that further research is
patterns (e.g., for joy, sex), similar to those he had obtained needed to fully evaluate Clynes’s theory.
without music, in listeners’ sentograph responses to the
pulse in pieces of ethnic and rock music (Clynes & Walker,
Developmental Studies
1982, pp. 202–207; see also examples in Gabrielsson,
1993, p. 107). Children’s perception of emotional expression has been
Independent attempts to test Clynes’s theory about spe- investigated in a number of recent studies (e.g., Cunning-
cific spatiotemporal patterns (sentograms) for different ham & Sterling, 1988; Dolgin & Adelson, 1990; Giomo,
emotions have yielded mixed results. Asked to match a 1993; Kratus, 1993; Terwogt & van Grinsven, 1991). Chil-
randomized list of the emotion labels (i.e., anger, love, sex, dren of varying age—from 4 to 12 years across these
hate, joy, reverence) with the corresponding sentograms investigations—were asked to indicate which emotion
presented in graphic illustrations, subjects showed poor they perceived in the music. The youngest children usu-
matching accuracy—barely above chance (Gorman & ally answered by choosing among faces with different
Crain, 1974). Trussoni, O’Malley, and Barton (1988) and emotional expressions. Older children could choose
Nettelbeck, Henderson, and Willson (1989) asked partici- among verbal alternatives, and responses by adolescents
pants to produce sentograms for eight emotions, and ex- or adults were used as reference. The music was mostly
amined their similarity to the spatiotemporal curves sug- classical music or specially composed pieces (Dolgin &
gested by Clynes (1973). In both studies, the results were Adelson, 1990). Emotions to choose among were usually
mainly negative. Trussoni et al. (1988), for instance, noted limited to happiness, sadness, anger, and fear (also “ex-
“the definite absence of a significant correlation between cited” and “calm” in Kratus, 1993). Even the youngest
our sentograms and those of Clynes” (p. 423). However, children were able to recognize these emotions with fair
Hama and Tsuda (1990) provided some support for the accuracy. However, most studies indicated better recog-
theory: Sentograms produced in response to a sad face nition with increasing age and better recognition of hap-
lasted longer than those to an angry face, and sentograms piness and sadness than of anger and fear; the latter two
produced in response to an angry face exhibited higher may be confused (also seen in Robazza et al., 1994). Girls
pressure than sentograms to faces representing joy, sorrow, achieved better recognition than boys, generally or for
or no emotion. some emotions, in most studies, but Dolgin and Adelson
A more important test of Clynes’s theory for our present (1990) and Kratus (1993) found no consistent gender dif-
purposes would involve the use of sounds shaped by the ference. There were certain differences in the results be-
production of pressure patterns. Nettelbeck et al. (1989) tween the different studies, which may be due to differ-
used such “sentic” sound material from the soundsheet ences in the type of music, instruments, and procedures
included in Clynes (1982) to replicate a study by Clynes used.
and Nettheim (1982). Only three of the six sound stimuli Do children associate major mode with happiness and
(i.e., joy, anger, grief) were decoded with a fair degree of minor mode with sadness? Infants (6 months old) showed
applicable copyright law.

accuracy. Nettelbeck and associates therefore concluded no difference in their preference for a major or minor
that “Clynes and Nettheim’s (1982) results are properly chord, but preference for a consonant chord over a dis-
regarded as artefactual” (p. 36). sonant chord (Crowder, Reznick, & Rosenkrantz, 1991).
Some support for Clynes’s theory was found by de Children from preschool and elementary school associated
Vries (1991), who asked 30 subjects to listen to 10 pieces major melodies to positive emotions (happy, contented)

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 525

and minor melodies to negative emotions (sad, angry), not nese music does not primarily aim at expressing emotion
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

perfect but significantly different from what would hap- but rather at describing the beauty and harmony of nature.
pen on chance basis (Kastner & Crowder, 1990). Gregory,
Worrall, and Sarge (1996), too, found the happy–major
Emotional Expression in Performance
and sad–minor association for older children (7–8 years)
but not for 3- to 4-year-olds. Thus, the conventional As should be evident from the review so far, studies of
happy–major and sad–minor associations are found in emotional expression have almost exclusively been con-
older children, whereas the results for younger children cerned with the impact of particular pieces of music,
are still inconclusive. whereas they have ignored the impact of specific perfor-
mances. However, the performance of a piece of music is
also important in shaping the emotional expression. By
performing the music in a certain way, a performer may
Cross-Cultural Studies
intend to make fully manifest a quality of emotion that he
There are only a few cross-cultural studies. Gundlach or she considers to be important for the piece in question.
(1932) collected and analyzed transcriptions of 334 songs Throughout history, there has been plenty of anecdotal
from different North American Indian tribes, including evidence about the extraordinary expressive skills of par-
healing songs, love songs, and war songs. He then at- ticular performers. Empirical research on emotional ex-
tempted to find similarities between European art music, pression in performance, however, is a fairly recent phe-
Indian songs, and European folk songs (Gundlach, 1935, nomenon.
p. 637). Comparisons were hard to do, but some similari- Most studies have used the following paradigm. Pro-
ties were suggested. For example, high pitch used to char- fessional performers are asked to play or sing brief melo-
acterize animated and glad feelings in European music oc- dies to express different emotions chosen by the re-
curred in Indian songs for happy love, and slow tempo searcher. The performances are recorded, and then judged
used to express melancholy or mourning in European mu- by listeners to see whether they can infer the intended
sic occurred for sad love in Indian songs. expressions. Each performance is also analyzed to study
Morey (1940) played recordings of European art music, what means the performer used to accomplish each emo-
selected to express fear, reverence, rage, love, and no par- tional expression. The assumption is that because the mel-
ticular emotion, to 20 native West Africans, asking them ody remains the same in the different versions, whatever
about the activities and emotional expression suggested to effects appear in listeners’ judgments or in acoustic mea-
them by these pieces. Nine of them left without answering, sures should primarily be the result of the performer’s ex-
and the answers given by the remaining listeners showed pressive intention.
little, if any, correspondence to the hypothesized expres- There are now about 30 studies of emotional expression
sions. in performance. They cover a wide range of musical styles,
Kleinen (1994) had German and Chinese students judge including opera (e.g., “Ardi gli incensi” from Lucia di
the expression in two European (Mozart, Brahms) and two Lammermoor; Siegwart & Scherer, 1995), classical music
Chinese folklore pieces of music, using 11 scales referring (Violin Concerto by Mendelsohn; Senju & Ohgushi, 1987),
to emotion (e.g., cheerful), motion, and structure (e.g., or- folk music (“Greensleeves”; Juslin, 2000), jazz (e.g.,
der). The results showed commonalities as well as marked “When the Saints”; Juslin, 1997a), and popular tunes
differences between the two subject groups; differences (“Happy Birthday”; Gabrielsson & Lindström, 1995). The
were most obvious in aesthetic evaluation. emotions investigated have usually included happiness,
Gregory and Varney (1996) asked English and Indian sadness, anger, fear, solemnity, and love/tenderness. A va-
students (living in England) to listen to Western classical, riety of musical instruments have also been included,
classical Indian, and New Age pieces of music and check such as the flute, violin, clarinet, electric guitar, piano,
one or more adjective groups with regard to emotional ex- trumpet, drums, synthesizer, and singing voice.
pression. There was an overall good agreement between
the two subject groups, but also significant differences in
Communication Accuracy
perceived expression of many pieces.
Hoshino (1996) studied perceived expression of Japa- The primary question facing researchers was whether per-
nese musical modes and Western major and minor modes formers could communicate emotions to listeners at all. In
applicable copyright law.

using color associations and verbal associations as re- what is probably the first study, Kotlyar and Morozov
sponses from different groups of Japanese subjects. While (1976) asked 10 opera singers to sing phrases from various
major and minor modes were perceived as contrasting in pieces of music in such a way that they would convey joy,
emotional character, the Japanese modes had weaker ef- anger, sorrow, and fear to listeners. Then, 10 musically
fects, which was related to the fact that traditional Japa- trained listeners were asked to judge the emotional ex-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
526 PART IV. EXPRESSION OF EMOTION

pression of each performance. The results showed that lis- (e.g., Gabrielsson & Juslin, 1996; Gabrielsson & Lindström,
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

teners were highly successful at decoding the intended ex- 1995; Juslin, 1997a, 2000; Juslin & Laukka, 2000; Juslin &
pression. Since then, many studies have confirmed that Madison, 1999; Kotlyar & Morozov, 1976; Ohgushi & Hat-
professional music performers are able to communicate tori, 1996b; Rapoport, 1996).
specific emotions to listeners (e.g., Baroni & Finarelli, One important finding from this research is that the
1994; Behrens & Green, 1993; Gabrielsson, 1994, 1995; Ga- performer’s expressive intention affects almost every as-
brielsson & Juslin, 1996; Gabrielsson & Lindström, 1995; pect of the performance. That is, emotional expression in
Juslin, 1997a, 1997b, 1997c, 2000; Juslin & Madison, 1999; performance seems to involve a whole set of cues (pieces
Ohgushi & Hattori, 1996a; Sundberg, Iwarsson, & Hage- of information) that are used by performers and listeners.
gård, 1995). Listeners are even able to decode certain emo- Most of these cues are included in Table 26.2: tempo,
tions with better than chance accuracy from single voice sound level, timing, articulation, timbre, vibrato, intona-
tones (Konishi, Imaizumi, & Niimi, 2000; Sherman, 1928). tion, attacks, and decays. Thus, for example, expressions
However, many studies have also revealed marked indi- of sadness are associated with slow tempo, low sound
vidual differences in encoding and decoding accuracy, level, legato articulation, slow tone attacks, and “soft” tim-
consistent with the findings from studies of other nonver- bre (i.e., little high-frequency energy in the spectrum),
bal communication channels (Buck, 1984; Wallbott & whereas expressions of anger are associated with fast
Scherer, 1986). tempo, high sound level, staccato articulation, fast tone
Few studies have yet reported decoding accuracy in a attacks, and “sharp” timbre (i.e., much high-frequency en-
manner that makes more precise estimates possible. How- ergy in the spectrum). Of course, the number of cues avail-
ever, Juslin (1997b) used a forced-choice format in order able depends on the specific instrument used (for a more
to make the results comparable to the results from studies extensive review, see Juslin, 2001a).
of vocal expression. He found that decoding accuracy was Performance analyses have shown that performers can
75% correct, or about four times higher than what would use a number of cues to express specific emotions. How-
be expected by chance alone. This suggests that, at least ever, these analyses do not show that listeners actually use
under certain circumstances, the accuracy of communi- the same cues in their judgments. To test the validity of
cation of emotion via musical performance may approach hypotheses derived from performance studies, it is nec-
the accuracy of facial or vocal expression (see chapter 22, essary to carry out listening experiments with synthesized
this volume; chapter 28, this volume). and systematically varied performances (Gabrielsson,
One objection to this finding might be that the accuracy 1985). Such listening experiments have confirmed that lis-
of the communicative process is boosted because of the teners make the same associations between cues and emo-
response format used. In most studies, listeners made their tions as performers do (see Bresin & Friberg, 2000; Juslin,
judgments by means of forced choice or adjective ratings— 1997b; Juslin, Friberg, & Bresin, in press).
methods that offer only a limited number of response op-
tions. To address this problem, Juslin (1997c) investigated
Tests of Predictions: The Functionalist Perspective
the generalizability of decoding accuracy across response
formats by means of a parallel enrichment procedure (Ro- In this section, we summarize evidence with regard to the
senthal, 1982)—that is, quantitative data from forced- theoretical predictions of the functionalist perspective (see
choice judgments of performances were augmented by “Emotional Expression in Performance,” earlier in chap-
qualitative data from free labeling of the same stimuli. The ter). One major prediction of this approach is that there
two methods converged on the conclusions that (1) com- should be similarities regarding the acoustic cues used to
munication was reliable regardless of the response format communicate emotions in vocal expression and music
and (2) what could be communicated reliably were the performance, because it is assumed that performers use
basic-level emotion categories, but not specific nuances the same code as is used in vocal expression. Juslin (1999)
within these categories (cf. Brown, 1981; Campbell, 1942; conducted a review of a large number of studies from both
Downey, 1897). domains. This review revealed numerous similarities in
code usage. For instance, anger is expressed by means of
increased speed, intensity, and high-frequency spectral en-
Code Usage
ergy in both vocal expression and music performance.
The early studies were mainly concerned with exploring One assumption of the functionalist perspective is that
applicable copyright law.

whether performers are able to communicate emotions to decoding of vocal expression of emotion (and hence also
their listeners at all. However, if we want to study com- expression in performance) is done largely in terms of a
munication as a process, we need also to consider its set of basic emotion categories. Support for this view
mechanisms, particularly the code used by performers and comes from evidence of categorical perception of vocal
listeners. Accordingly, researchers have tried to describe expression of emotion (de Gelder & Vroomen, 1996). Thus,
the means by which performers express specific emotions another prediction of the functionalist perspective is that

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 527

basic emotions should be easier to communicate than ute in an additive fashion to listeners’ judgments (Juslin,
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

other emotions in music performances. Indeed, a survey 1997b). All this supports a conceptualization of the com-
of the literature suggests that commonly suggested basic municative process in terms of a Brunswikian lens model.
emotions, such as sadness, anger, and happiness, are eas- Performance studies have shown that the cue utiliza-
ier to communicate than other states, such as solemnity, tion is inconsistent both across and within performers
tenderness, and dreaminess (Gabrielsson & Juslin, 1996; (Juslin, 2000). Similarly, listening tests have shown that
Gabrielsson & Lindström, 1995; Senju & Ohgushi, 1987). the cue utilization is inconsistent both across and within
The functionalist perspective holds that the vocal code listeners (Juslin & Madison, 1999). This really presents
on which performers base their communication of emo- something of a puzzle: How can the communicative pro-
tions is innate. Accordingly, it is predicted that there cess be accurate if the cue utilization is inconsistent? This
should be cross-cultural similarities in how performers apparent paradox can be explained by the lens model: Be-
employ cues in the performance to express emotions. cause cues are intercorrelated (and hence partly redun-
While no study thus far has (explicitly) aimed to address dant), many different cue utilization strategies can lead to
this issue, there is evidence of cross-cultural similarities a similar level of accuracy (cf. Dawes & Corrigan, 1974).
in performers’ utilization of cues from investigations in Indeed, a study by Juslin (2000) has shown that two per-
India (Balkwill & Thompson, 1999), Russia (Kotlyar & Mo- formers can reach a similar level of communication ac-
rozov, 1976), Italy (Baroni & Finarelli, 1994), Japan (Ohgu- curacy despite differences in their cue utilization (for fur-
shi & Hattori, 1996), and Sweden (Juslin, 2000). These ther discussion, see Juslin & Persson, in press).
studies indicate that cues, such as tempo, sound level, and
timbre, are used in a similar manner in different cultures.
However, when it comes to cross-cultural studies that in- Discussion
volve other musical variables, such as harmony, there are
usually larger differences.
Methodological Issues
One further prediction of the functionalist perspective
is that the ability to decode basic emotions from perfor-
Selection of Stimuli
mance should be present early in development. Although
there are no developmental studies that focus specifically Ideally, the music stimuli should be a broad and represen-
on children’s perception of emotional expression in per- tative sample from the universe of music, which is, of
formance, there are studies that suggest that young chil- course, infinite. The number of musical pieces included
dren are able to decode basic emotions in pieces of music in the investigations varies from 2 to 40, and the total
(see “Developmental Studies”). Considering that these number in all investigations together is still less than
studies involved both performance variables (e.g., tempo, 1,000. Moreover, most of these are either very short pieces
articulation, timbre) and composed structure (e.g., har- or excerpts from longer pieces, sometimes only a single
monic progression, tonality), it seems plausible that young phrase. The large majority belong to Western tonal music,
children are able to decode emotional expression from and mostly art music (for discussions of popular music,
performance variables as such. Studies of children’s abil- see, e.g., Frith, 1996; Tagg, 1979). Representativity is thus
ity to encode emotions in music are complicated by the so far very limited, in terms of both number of pieces and
fact that very young children may not be able to play an musical genres. The rationale for the selection of stimuli
instrument. However, Adachi and Trehub (1998) found is sometimes explained, sometimes not. A certain arbi-
that children 4–12 years old were able to use cues to ex- trariness is almost certainly present depending on the re-
press emotions in singing. For instance, they used fast searcher’s musical knowledge and preferences, and on
tempo and high intensity in happy expressions, whereas practical and economical matters. An advantage of using
they used slow tempo and low intensity in sadness ex- music unknown to the listeners is that it does not preju-
pressions. dice their judgments due to common or private associa-
There are also a number of predictions that derive from tions and memories.
the conceptualization of the communicative process in The short duration of most music stimuli is usually mo-
terms of a Brunswikian lens model. For instance, the pre- tivated by the need to have stimuli with homogeneous ex-
diction that cues are only probabilistically related to per- pression, so as to optimize the conditions for consistency
formers’ expressive intentions and listeners’ judgments is in the listener’s judgments. Brief stimuli also make it eas-
applicable copyright law.

supported by evidence in Juslin (1997a, 1997b, 2000) and ier to investigate which musical factors influence the per-
Juslin and Madison (1999). It has further been shown that ceived expression. On the other hand, brief stimuli mean
there are moderate intercorrelations between the cues (Jus- a limitation in ecological validity. With recent proposals
lin, 2000; see also Kratus, 1993). Finally, there is evidence for continuous recording of perceived expression there are
from an experiment that used systematic manipulation of now possibilities for studying how expression varies in
synthesized performances that performance cues contrib- extended pieces of music.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
528 PART IV. EXPRESSION OF EMOTION

tion to what sensitive listeners perceive. Even world-


Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

Response Formats
renowned authors complain about the elusiveness of
Listeners’ descriptions of perceived expression in music music to verbal expression. In a way, music begins where
are usually obtained by verbal responses. Nonverbal re- language ends.
sponses, such as drawings (Francès, 1988; Smith & Wil-
liams, 1999), finger pressure on a sentograph (Clynes &
Emotions Expressed in Music
Walker, 1982), pressing a pair of tongs (Nielsen, 1983),
moving a slider (Krumhansl, 1996) or a computer mouse A great variety of emotion descriptors have been used in
(Schubert, 1999a), and the like, have rarely been used so the investigations reviewed (Table 26.1). Reliable recog-
far and may be difficult to interpret. Moreover, language nition of intended emotions or high interindividual agree-
is still used to define which aspect the listener should at- ment on expressed emotion has been reported for many of
tend to (e.g., tension). them. Among so-called basic emotions, results are most
There are several problems with verbal responses. The convincing concerning happiness/gaiety/joy and sadness/
meanings of words differ among people, investigators as gloom. Anger and fear have been little investigated except
well as subjects. Verbal labels attached to clusters or di- in performance studies, in which they are usually recog-
mensions of emotional expression vary among research- nized well, but disgust has rarely appeared. The concept
ers, a problem further aggravated when terms have to be of basic emotions and other influences from emotion psy-
translated into another language. The selection of descrip- chology in general have come into studies of musical ex-
tive terms is crucial. Some investigators rely on descrip- pression quite recently and then mainly in studies of per-
tors used in earlier studies; others collect a large number formance. Rather, in the bulk of investigations, the
of descriptors from various sources and have them rated emotions considered have been chosen relying on state-
for suitability to describe expression in music (e.g., Schu- ments by philosophers, music theorists, suggestions from
bert, 1999a; Terwogt & van Grinsven, 1991), keeping those earlier studies on expression, and, not the least, on folk
that are highest rated. There is no generally agreed clas- psychology and personal experiences. All together the
sification of emotions, or emotion words, to rely upon emotion descriptors used in these investigations are
(Plutchik, 1994). counted in hundreds.
Verbal methods appear in many variants: forced choice The most far-reaching attempts to achieve reduction
between two descriptive terms; choice of any number of and order among these myriad terms are studies involving
a given set of descriptive terms; numerical ratings using multivariate analysis techniques—factor analysis, cluster
bipolar or unipolar scales; or free phenomenological de- analysis, correspondence analysis, multidimensional scal-
scription. Using choice or ratings of given descriptors al- ing. Together these investigations suggest a valence di-
lows the investigator to study just those descriptors that mension (gaiety–gloom, positive vs. negative feelings) and
he or she considers relevant, and data can usually be suit- an arousal dimension (tension/energy–relaxation), in
ably analyzed by common statistical methods. However, agreement with the results from studies in the emotion
the given descriptors may not be the ones that the listener domain (Russell, 1980). The status of the potency dimen-
finds most adequate or that he or she would focus on in sion must be left open, because the reviewed studies do
free phenomenological description. The latter are harder not provide unequivocal results to confirm its existence.
to analyze and to present in a condensed way; it may be There is perhaps another dimension called solemnity ver-
necessary to present them in extenso to fully apprehend sus triviality (e.g., Wedin, 1972c). In Hevner’s adjective
the contents. Which response format should be used must clock this might correspond to an axis from cluster 1 (in-
be considered in relation to the questions under scrutiny, cluding solemn) to cluster 5 (fanciful, playful, whimsical;
the music used, the type of listeners, and so on. Using see Figure 26.4). Wedin (1972c) pointed out that solem-
forced choice among a few alternative descriptors may in- nity–triviality may reflect a stylistic dimension, solemnity
dicate high agreement among listeners; however, with free used to characterize certain kinds of art music but rarely
descriptions by the same listeners agreement may be lower applicable to popular music.
(e.g., Rigg, 1937a; see also Watt & Ash, 1998, p. 39). Ex- Some basic emotions, such as anger and fear, do not
pression in music should preferably be studied in a mul- appear at all in Hevner’s adjective circle and hardly in the
timethod approach (cf. Juslin, 1997c). dimensional studies, which may reflect both that these
Finally, it may be questioned whether it it is possible emotions are relatively little expressed in classical music
applicable copyright law.

to describe all subtleties in music by words. Many would and that studies on emotional expression in music have
probably agree with Gilman’s (1892, p. 72) statement that had little contact with emotion psychology in general (and
“what music expresses is literally unutterable.” So far, the vice versa). Moreover, most descriptive terms in Kleinen’s
terminology developed in research for describing expres- and Wedin’s representations (Figures 26.1 and 26.2) be-
sion in music seems meager and void of nuances in rela- long to the upper-right and lower-left quadrants, whereas

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 529

there are few terms in the remaining quadrants. However, of music to a small number of main dimensions have
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

in Russell’s (1980) model, emotions appear about equally yielded dimensions corresponding to those discovered in
in all quadrants. This may again reflect that certain emo- other domains of emotion research, as well as some di-
tions are more likely to be expressed in music than others mensions that may be more typical of music. It is quite
(which does not mean that the latter cannot be expressed obvious, however, that these dimensions cannot capture
by musical means). all the nuances that we can perceive in music.
Many of music’s most expressive qualities relate to There are several areas of musical expression that are
changes over time, which makes possible expression of, not treated in this chapter: expression in film music (Co-
say, shifting emotions, blending emotions, or conflicting hen, 2001; Vol. 13 of Psychomusicology, 1994), music in
emotions. Such changes involve elusive qualities that are advertising (e.g., North & Hargreaves, 1997), music ther-
best described in dynamic terms, such as growth, decay, apy (e.g., Pratt & Jones, 1987 for historical account; Bonny
tension and release, crescendo–diminuendo, and acceler- & Savary, 1990, and Wigram, Saperston, & West, 1995, for
ando–ritardando. They recall what Stern (1985) termed contemporary examples). Works in musical aesthetics
“vitality affects,” which seem intuitively appealing in the (e.g., Budd, 1985; Davies, 1994; Kivy, 1990; Scruton, 1997)
study of musical expression (e.g., Sloboda & Juslin, 2001). or musical semiotics (e.g., Tarasti, 1994) have been only
Phenomena like these are little investigated in research so cursorily referred to. There is little, if any, reference to
far; the closest examples are studies on tension and release empirical studies in these works. Similarly, it is rare to
by means of continuous recording. Continuous recording find references to works in musical aesthetics or musical
of listeners’ judgments seems a necessary prerequisite to semiotics in empirical music psychology. We believe that
study these phenomena. much can be gained from the cross-fertilization of all these
Many studies indicate little or no difference in results different domains.
between musically experienced and musically inexperi- The musical means that can be used to express various
enced subjects (e.g., Campbell, 1942; Gabrielsson, 1973; emotions constitute practically all variables in musical
Hevner, 1936; Juslin, 1997a; Rigg, 1937b; Robazza et al., structure and performance, including several interactions,
1994; Wedin, 1969). However, Brown (1981) concluded as evident in Table 26.2 (see also Gabrielsson & Lindström,
that with pieces that are close in expression, agreement 2001; Juslin, 2001a). Previous studies have focused mainly
could only be achieved by highly trained listeners; if the on rather simple physical variables (e.g., tempo, loudness,
expressions are not too similar, agreement may be reached pitch), whereas much remains to be done concerning the
also by musically untrained persons. Developmental stud- effects of more specific rhythmic, melodic, and harmonic
ies indicate that even very young children are able to rec- features, as well as various aspects of musical form. The
ognize basic emotions expressed in music, happiness and possibilities afforded by sound synthesis have rendered
sadness earlier than anger and fear. Finally, gender differ- systematic manipulation of such musical factors quite fea-
ences have been reported in a number of studies (e.g., sible.
Cunningham & Sterling, 1988; Dolgin & Adelson, 1990; One problem in previous research is that representativ-
Gabrielsson & Juslin, 1996; Giomo, 1993; Juslin, 1997c). ity has been limited, in terms of both the number of pieces
and the number of musical genres studied. Future research
should encompass a wider range of musical genres. Fur-
Conclusions and Directions for Future Research
thermore, cross-cultural research is badly needed because
We believe that the present review allows us to draw some only through such research is it possible to determine
tentative conclusions. The majority of empirical studies whether there are universal relationships between musical
have investigated whether listeners agree in their descrip- factors and emotional expression.
tions of expression in music, and what musical character- Finally, much of the empirical work is characterized by
istics correlate with listeners’ judgments of expression. a lack of theories. This has probably impeded the progress
The results suggest that agreement about a perceived ex- in this field. That musical training is not required to per-
pression varies greatly depending on a number of factors ceive emotional expression in music suggests that more
(e.g., the piece of music, the musical genre, the response general mechanisms of emotion perception are involved.
format, the type of listener). There is generally greater Brain imaging techniques, such as PET and fMRI, might
agreement or accuracy with regard to the expression of help to discriminate among different theoretical accounts
emotions than with regard to the expression of other as- of the mechanisms underlying emotional expression in
applicable copyright law.

pects (e.g., in program music). Also, many studies suggest music. Only recently have music researchers turned to
that there is good agreement among listeners regarding the emotion psychology for theoretical guidance. However,
broad emotional category expressed in music, but less we believe that a closer alliance between music research-
agreement regarding the nuances or variants within this ers and emotion researchers affords the royal road to pro-
category. Attempts to reduce perceived emotions in pieces gress in the field of musical expression.3

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
530 PART IV. EXPRESSION OF EMOTION

Brown, R. (1981). Music and language. In Documentary


NOTES
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

report of the Ann Arbor symposium. National sympo-


1. Unless otherwise stated, the text in this section is sium on the applications of psychology to the teaching
mainly based on Benestad (1978), Lippman (1986, 1988, and learning of music (pp. 233–265). Reston, VA: Music
1990), and on articles in the New Grove Dictionary of Mu- Educators National Conference.
sic and Musicians (Sadie, 1980): Aesthetics of music, Af- Brunswik, E. (1956). Perception and the representative de-
fections, Empfindsamkeit, Expression, Figures, Greece, sign of psychological experiments. Berkeley: University
Hermeneutics, Programme music, Rhetoric and music, Ro- of California Press.
mantic, Tempo and expression marks. Further theories on Buck, R. (1984). The communication of emotion. New
musical expression developed within musicology are re- York: Guilford Press.
viewed by de la Motte-Haber (1996, pp. 25–79). A histor- Budd, M. (1985). Music and the emotions. The philosoph-
ical account of performance practices used to enhance ical theories. London: Routledge.
emotional expression is given by Hudson (1994). Buelow, G. J. (1983). Johann Mattheson and the invention
2. For an explanation of various modes in ancient of the Affektenlehre. In G. J. Buelow & H. J. Marx (Eds.),
Greek music, see article “Greece” in Sadie (1980). New Mattheson studies (pp. 393–407). Cambridge, En-
3. The writing of this chapter was supported by The gland: Cambridge University Press.
Bank of Sweden Tercentenary Foundation. Campbell, I. G. (1942). Basal emotional patterns expressi-
ble in music. American Journal of Psychology, 55, 1–
17.
REFERENCES
Capurso, A. (1952). The Capurso study. In A. Capurso, V.
Adachi, M., & Trehub, S. E. (1998). Children’s expression R. Fisichelli, L. Gilman, E. A. Gutheil, J. T. Wright, & F.
of emotion in song. Psychology of Music, 26, 133–153. Paperte (Eds.), Music and your emotions (pp. 56–86).
Allesch, C. G. (1987). Geschichte der psychologischen Äs- New York: Liveright.
thetik [History of psychological aesthetics]. Göttingen: Cazden, N. (1979). Can verbal meanings inhere in frag-
Hogrefe. ments of melody? Psychology of Music, 7, 34–38.
Asada, M., & Ohgushi, K. (1991). Perceptual analyses of Clynes, M. (1973). Sentics: Biocybernetics of emotion
Ravel’s Bolero. Music Perception, 8, 241–250. communication. Annals of the New York Academy of
Asmus, E. P. (1985). The development of a multidimen- Sciences, 220, 55–131.
sional instrument for the measurement of affective re- Clynes, M. (1977). Sentics: The touch of emotions. New
sponses to music. Psychology of Music, 13, 19–30. York: Anchor Press/Doubleday.
Bach, C. P. E. (1753, 1762). Versuch über die wahre Art, Clynes, M. (1980). The communication of emotion: Theory
das Clavier zu spielen. (Facsimile edition 1957, Breit- of sentics. In R. Plutchik & H. Kellerman (Eds.), Emo-
kopf & Härtel, Leipzig.) tion: Theory, research, and experience. Vol. 1: Theories
Balkwill, L-L., & Thompson, W. F. (1999). A cross-cultural of emotion (pp. 271–301). New York: Academic Press.
investigation of the perception of emotion in music: Clynes, M. (Ed.). (1982). Music, mind, and brain: The neu-
Psychophysical and cultural cues. Music Perception, ropsychology of music. New York: Plenum.
17, 43–64. Clynes, M., & Nettheim, N. (1982). The living quality of
Baroni, M., & Finarelli, L. (1994). Emotions in spoken lan- music: Neurobiologic basis of communicated feeling. In
guage and vocal music. In I. Deliège (Ed.), Proceedings M. Clynes (Ed.), Music, mind, and brain: The neuro-
of the Third International Conference for Music Percep- psychology of music (pp. 47–82). New York: Plenum.
tion and Cognition (pp. 343–345). University of Liége, Clynes, M., & Walker, J. (1982). Neurobiologic functions of
Liége, Belgium. rhythm, time, and pulse in music. In M. Clynes (Ed.),
Batel, G. (1976). Komponenten musikalischen Erlebens. Music, mind, and brain: The neuropsychology of music
Göttingen, Germany: Göttinger Musikwissenschaft- (pp. 171–216). New York: Plenum.
licher Arbeiten. Cohen, A. (2001). Music as a source of emotion in film. In
Batson, C. D., Shaw, L. L., & Oleson, K. C. (1992). Differ- P. N. Juslin & J. A. Sloboda (Eds.), Music and emotion:
entiating affect, mood, and emotion: Toward function- Theory and research (pp. 249–272). New York: Oxford
ally based conceptual distinctions. In M. S. Clark (Ed.), University Press.
Review of personality and social psychology: Emotion Cooke, D. (1959). The language of music. London: Oxford
(pp. 294–326). Newbury Park, CA: Sage. University Press.
Behne, K. E. (1972). Der Einfluss des Tempos auf die Beur- Crowder, R. G. (1984). Perception of the major/minor dis-
teilung von Musik. Köln, Germany: Arno Volk Verlag. tinction: I. Historical and theoretical foundations. Psy-
Behrens, G. A., & Green, S. B. (1993). The ability to iden- chomusicology, 4, 3–12.
tify emotional content of solo improvisations per- Crowder, R. G. (1985). Perception of the major/minor dis-
formed vocally and on three different instruments. Psy- tinction: III. Hedonic, musical, and affective discrimi-
chology of Music, 21, 20–33. nations. Bulletin of the Psychonomic Society, 23, 314–
Benestad, F. (1978). Musik och tanke. Huvudlinjer i mu- 316.
sikestetikens historia från antiken till vår egen tid [Mu- Crowder, R. G., Reznick, J. S., & Rosenkrantz, S. L. (1991).
sic and thought. Main lines in the history of musical Perception of the major/minor distinction: V. Prefer-
applicable copyright law.

aesthetics from antiquity to our time]. Stockholm: Ra- ences among infants. Bulletin of the Psychonomic So-
bén & Sjögren. ciety, 29, 187–188.
Bonny, H. L., & Savary, L. M. (1990). Music and your mind Cunningham, J. G., & Sterling, R. S. (1988). Developmental
(Rev. ed.). Barrytown, NY: Station Hill Press. changes in the understanding of affective meaning in
Bresin, R., & Friberg, A. (2000). Emotional coloring of music. Motivation and Emotion, 12, 399–413.
computer-controlled music performance. Computer Davies, S. (1994). Musical meaning and expression. Ithaca,
Music Journal, 24, 44–62. NY: Cornell University Press.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 531

Dawes, R. M., & Corrigan, B. (1974). Linear models in de- sion in music performance: Between the performer’s in-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

cision making. Psychological Bulletin, 81, 95–106. tention and the listener’s experience. Psychology of Mu-
De Gelder, B., & Vroomen, J. (1996). Categorical perception sic, 24, 68–91.
of emotional speech. Journal of the Acoustical Society Gabrielsson, A., & Lindström, E. (1995). Emotional ex-
of America, 100, 2818. pression in synthesizer and sentograph performance.
Dolgin, K., & Adelson, E. (1990). Age changes in the ability Psychomusicology, 14, 94–116.
to interpret affect in sung and instrumentally-presented Gabrielsson, A., & Lindström, E. (2001). The influence of
melodies. Psychology of Music, 18, 87–98. musical structure on emotional expression. In P. N. Jus-
Downey, J. E. (1897). A musical experiment. American lin & J. A. Sloboda (Eds.), Music and emotion: Theory
Journal of Psychology, 9, 63–69. and research (pp. 223–248). New York: Oxford Univer-
Ekman, P. (1992). An argument for basic emotions. Cog- sity Press.
nition and Emotion, 6, 169–200. Gilman, B. I. (1891). Report on an experimental test of mu-
Ekman, P., & Davidson, R. J. (Eds.). (1994). The nature of sical expressiveness. American Journal of Psychology,
emotion. New York: Oxford University Press. 4, 558–576.
Farnsworth, P. R. (1954). A study of the Hevner adjective Gilman, B. I. (1892). Report of an experimental test of mu-
list. Journal of Aesthetics and Art Criticism, 13, 97–103. sical expressiveness (continued). American Journal of
Fisk, J. (Ed.). (1997). Composers on music. Eight centuries Psychology, 5, 42–73.
of writings (2nd ed.). Boston: Northeastern University Giomo, C. J. (1993). An experimental study of children’s
Press. sensitivity to mood in music. Psychology of Music, 21,
Foros, P. B. (1975). Hva er musikalsk kommunikasjon? 141–162.
[What is musical communication?] In J. P. Jensen & M. Gorman, B. S., & Crain, W. C. (1974). Decoding of “sen-
Poulsen (Eds.), Musikopleven og musikformidling [Ex- tograms.” Perceptual and Motor Skills, 39, 784–786.
perience and communication of music] (pp. 178–203). Gregory, A. H., & Varney, N. (1996). Cross-cultural com-
Copenhagen: Akademisk Vorlag. parisons in the affective response to music. Psychology
Francès, R. (1988). The perception of music. Hillsdale, NJ: of Music, 24, 47–52.
Erlbaum. (French original: La perception de la mu- Gregory, A. H., Worrall, L., & Sarge, A. (1996). The devel-
sique, 1958. Paris: Libraire Philosophique J. Vrin.) opment of emotional responses to music in young chil-
Fredrickson, W. E. (1997). Elementary, middle, and high dren. Motivation and Emotion, 20, 341–348.
school student perceptions of tension in music. Journal Grout, D. J., & Palisca, C. V. (1996). A history of Western
of Research in Music Education, 45, 626–635. music (5th ed.). New York: Norton.
Frith, S. (1996). Performing rites. On the value of popular Gundlach, R. H. (1932). A quantitative analysis of Indian
music rites. Cambridge, MA: Harvard University Press. music. American Journal of Psychology, 44, 133–145.
Gabriel, C. (1978). An experimental study of Deryck Gundlach, R. H. (1935). Factors determining the charac-
Cooke’s theory of music and meaning. Psychology of terization of musical phrases. American Journal of Psy-
Music, 6, 13–20. chology, 47, 624–644.
Gabriel, C. (1979). A note on comments by Nettheim and Hama, H., & Tsuda, K. (1990). Finger pressure waveforms
Cazden. Psychology of Music, 7, 39–40. measured on Clynes’ sentograph distinguish among
Gabrielsson, A. (1973). Adjective ratings and dimension emotions. Perceptual and Motor Skills, 70, 371–376.
analysis of auditory rhythm patterns. Scandinavian Hampson, P. (2000). A naturalistic empirical investigation
Journal of Psychology, 14, 244–260. of Deryck Cooke’s theory of music and meaning. In C.
Gabrielsson, A. (1985). Interplay between analysis and Woods, G. Luck, R. Brochard, F. Seldon, & J. A. Sloboda
synthesis in studies of music performance and music (Eds.), Proceedings of the Sixth International Confer-
experience. Music Perception, 3, 59–86. ence on Music Perception and Cognition (CD-ROM).
Gabrielsson, A. (1988). Timing in music performance and University of Keele, UK.
its relations to music experience. In J. A. Sloboda (Ed.), Hampton, P. J. (1945). The emotional element in music.
Generative processes in music: The psychology of per- Journal of General Psychology, 33, 237–250.
formance, improvisation, and composition (pp. 27–51). Hanslick, E. (1854/1989). Vom musikalisch Schönen (21st
Oxford, England: Clarendon Press. ed.). English translation, The beautiful in music (New
Gabrielsson, A. (1993). The complexities of rhythm. In T. York: Liberal Arts Press, 1957) or On the musically
J. Tighe & W. J. Dowling (Eds.), Psychology and music. beautiful (Indianapolis: Hackett, 1986).
The understanding of melody and rhythm (pp. 93–120). Hare, F. G. (1977). Dimensions of music perception. Sci-
Hillsdale, NJ: Erlbaum. entific Aesthetics, 1, 271–288.
Gabrielsson, A. (1994). Intention and emotional expres- Heinlein, C. P. (1928). The affective character of the major
sion in music performance. In A. Friberg, J. Iwarsson, and minor modes in music. Journal of Comparative
E. Jansson, & J. Sundberg (Eds.), Proceedings of the Psychology, 8, 101–142.
Stockholm Music Acoustics Conference 1993 (pp. 108– von Helmholtz, H. L. F. (1863/1954). On the sensations of
111). Stockholm: Publications issued by the Royal tone as a physiological basis for the theory of music.
Swedish Academy of Music, No. 79. New York: Dover.
Gabrielsson, A. (1995). Expressive intention and perfor- Hermerén, G. (1986). Musik som språk [Music as lan-
applicable copyright law.

mance. In R. Steinberg (Ed.), Music and the mind ma- guage]. Swedish Journal of Musicology, 68, 7–16.
chine (pp. 35–47). Heidelberg, Germany: Springer. Hevner, K. (1935a). The affective character of the major
Gabrielsson, A. (2001). Emotions in strong experiences and minor modes in music. American Journal of Psy-
with music. In P. N. Juslin & J. A. Sloboda (Eds.), Music chology, 47, 103–118.
and emotion: Theory and research (pp. 431–449). New Hevner, K. (1935b). Expression in music: A discussion of
York: Oxford University Press. experimental studies and theories. Psychological Re-
Gabrielsson, A., & Juslin, P. N. (1996). Emotional expres- view, 47, 186–204.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
532 PART IV. EXPRESSION OF EMOTION

Hevner, K. (1936). Experimental studies of the elements of emotion: Theory and research. New York: Oxford Uni-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

expression in music. American Journal of Psychology, versity Press.


48, 246–268. Kaminska, Z., & Woolf, J. (2000). Melodic line and emo-
Hevner, K. (1937). The affective value of pitch and tempo tion: Cooke’s theory revisited. Psychology of Music, 28,
in music. American Journal of Psychology, 49, 621– 133–153.
630. Kastner, M. P., & Crowder, R. G. (1990). Perception of the
Hoshino, E. (1996). The feeling of musical mode and its major/minor distinction: IV. Emotional connotations in
emotional character in a melody. Psychology of Music, young children. Music Perception, 8, 189–202.
24, 29–46. Kivy, P. (1990). Music alone: Philosophical reflections on
Huber, K. (1923). Der Ausdruck musikalischer Elementar- the purely musical experience. Ithaca, NY: Cornell Uni-
motive. Leipzig: Johann Ambrosius Barth. versity Press.
Hudson, R. (1994). Stolen time. The history of tempo ru- Kleinen, G. (1968). Experimentelle Studien zum musikal-
bato. Oxford, England: Clarendon Press. ischen Ausdruck. Hamburg, Germany: Universität
Imberty, M. (1979). Entendre la musique. Paris: Dunod. Hamburg.
Izard, C. E. (1977). The emotions. New York: Plenum. Kleinen, G. (1994). Musikalischer Ausdruck und ästhetis-
Jürgens, U., & von Cramon, D. (1982). On the role of the che Wertung als interkulturelle Qualität und Differenz.
anterior cingulate cortex in phonation: A case report. Musikpsychologie. Jahrbuch der Deutschen Gesell-
Brain and Language, 15, 234–248. schaft für Musikpsychologie, 11, 76–101.
Juslin, P. N. (1995). Emotional communication in music Konishi, T., Imaizumi, S., & Niimi, S. (2000). Vibrato and
viewed through a Brunswikian lens. In G. Kleinen emotion in singing voice (abstract). In C. Woods, G.
(Ed.), Musical expression. Proceedings of the Confer- Luck, R. Brochard, F. Seddon, & J. A. Sloboda (Eds.),
ence of ESCOM and DGM 1995 (pp. 21–25). University Proceedings of the Sixth International Conference on
of Bremen, Bremen, Germany. Music Perception and Cognition, August 2000. (CD-
Juslin, P. N. (1997a). Emotional communication in music rom). University of Keele, UK.
performance: A functionalist perspective and some Kotlyar, G. M., & Morozov, V. P. (1976). Acoustic correlates
data. Music Perception, 14, 383–418. of the emotional content of vocalized speech. Soviet
Juslin, P. N. (1997b). Perceived emotional expression in Physics. Acoustics, 22, 370–376.
synthesized performances of a short melody: Capturing Kratus, J. (1993). A developmental study of children’s in-
the listener’s judgment policy. Musicae Scientiae, 1, terpretation of emotion in music. Psychology of Music,
225–256. 21, 3–19.
Juslin, P. N. (1997c). Can results from studies of perceived Krumhansl, C. L. (1996). A perceptual analysis of Mozart’s
expression in musical performances be generalized Piano Sonata K. 282: Segmentation, tension, and mu-
across response formats? Psychomusicology, 16, 77– sical ideas. Music Perception, 13, 401–432.
101. Krumhansl, C. L. (1997). An exploratory study of musical
Juslin, P. N. (1999). Communication of emotion in vocal emotions and psychophysiology. Canadian Journal of
expression and music performance: Different channels, Experimental Psychology, 51, 336–352.
same code? Manuscript submitted for publication. Krumhansl, C. L., & Lynn Schenk, D. (1997). Can dance
Juslin, P. N. (2000). Cue utilization in communication of reflect the structural and expressive qualities of music?
emotion in music performance: Relating performance A perceptual experiment on Balanchine’s choreography
to perception. Journal of Experimental Psychology: Hu- of Mozart’s Divertimento No. 15. Musicae Scientiae, 1,
man Perception and Performance, 26, 1797–1813. 63–85.
Juslin, P. N. (2001a). Communicating emotion in music Langer, S. K. (1953). Feeling and form. London: Routledge.
performance: A review and a theoretical framework. In Langer, S. K. (1957). Philosophy in a new key (3rd ed.).
P. N. Juslin & J. A. Sloboda (Eds.), Music and emotion: Cambridge, MA: Harvard University Press.
Theory and research (pp. 309–337). New York: Oxford Laukka, P., & Gabrielsson, A. (2000). Emotional expression
University Press. in drumming performance. Psychology of Music, 28,
Juslin, P. N. (2001b). A Brunswikian approach to emo- 181–189.
tional communication in music performance. In K. R. Levinson, J. (1996). The pleasures of aesthetics. Ithaca,
Hammond & T. R. Stewart (Eds.), The essential Bruns- NY: Cornell University Press.
wik: Beginnings, explications, applications (pp. 426– Lindström, E. (1997). Impact of melodic structure on emo-
430). New York: Oxford University Press. tional expression. In A. Gabrielsson (Ed.), Proceedings
Juslin, P. N., Friberg, A., & Bresin, R. (in press). Toward a of the Third Triennial ESCOM Conference, Uppsala,
computational model of expression in music perfor- June 1997 (pp. 292–297). Uppsala University, Uppsala,
mance: The GERM model. Musicae Scientiae. Sweden.
Juslin, P. N., & Laukka, P. (2000). Improving emotional Lippman, E. A. (Ed.). (1986). Musical aesthetics: A histor-
communication in music performance through cogni- ical reader. Volume 1: From Antiquity to the eighteenth
tive feedback. Musicae Scientiae, 4, 151–183. century. New York: Pendragon Press.
Juslin, P. N., & Madison, G. (1999). The role of timing pat- Lippman, E. A. (Ed.). (1988). Musical aesthetics: A histor-
terns in recognition of emotional expression from mu- ical reader. Volume 2: The nineteenth century. Stuy-
applicable copyright law.

sical performance. Music Perception, 17, 197–221. vesant, NY: Pendragon Press.
Juslin, P. N., & Persson, R. S. (in press). Emotional com- Lippman, E. A. (Ed.). (1990). Musical aesthetics: A histor-
munication. In R. Parncutt & G. E. McPherson (Eds.), ical reader. Volume 3: The twentieth century. Stuyves-
The science and psychology of music performance: ant, NY: Pendragon Press.
Creative strategies for teaching and learning. New York: Madsen, C. K. (1997). Emotional responses to music. Psy-
Oxford University Press. chomusicology, 16, 59–67.
Juslin, P. N., & Sloboda, J. A. (Eds.). (2001). Music and Madsen, C. K., & Fredrickson, W. E. (1993). The experi-

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
CHAPTER 26. EMOTIONAL EXPRESSION IN MUSIC 533

ence of musical tension: A replication of Nielsen’s re- tions of human and animal emotions. New York: Ox-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

search using the continuous response digital interface. ford University Press.
Journal of Music Therapy, 30, 46–63. Papoušek, M. (1996). Intuitive parenting: A hidden source
Mattheson, J. (1739). Der vollhommene Capellmeister. of musical stimulation in infancy. In I. Deliège & J. A.
(Facsimile edition 1954, Bärenreiter Verlag, Kassel, Sloboda (Eds.), Musical beginnings: Origins and devel-
Germany, and Basel.) opment of musical competence (pp. 89–112). Oxford,
Metfessel, M. (1932). The vibrato in artistic voices. In C. England: Oxford University Press.
E. Seashore (Ed.), University of Iowa studies in the psy- Peretz, I., Gagnon, L., & Bouchard, B. (1998). Music and
chology of music. Vol. I: The vibrato (pp. 14–117). Iowa emotion: Perceptual determinants, immediacy, and iso-
City: University of Iowa. lation after brain damage. Cognition, 68, 111–141.
Meyer, L. B. (1956). Emotion and meaning in music. Chi- Ploog, D. (1986). Biological foundations of the vocal ex-
cago: University of Chicago Press. pressions of emotions. In R. Plutchik & H. Kellerman
Morey, R. (1940). Upset in emotions. Journal of Social Psy- (Eds.), Emotion: Theory, research, and experience. Vol-
chology, 12, 333–356. ume 3: Biological foundations of emotion (pp. 173–
de la Motte-Haber, H. (1968). Ein Beitrag zur Klassifikation 197). New York: Academic Press.
musikalischer Rhythmen. Köln, Germany: Arno Volk Plutchik, R. (1994). The psychology and biology of emo-
Verlag. tion. New York: HarperCollins.
de la Motte-Haber, H. (1996). Handbuch der Musikpsy- Pratt, C. (1931). The meaning of music: A study in psy-
chologie (2nd ed.). Laaber: Laaber Verlag. chological aesthetics. New York: McGraw-Hill.
Namba, S., Kuwano, S., Hatoh, T., & Kato, M. (1991). As- Pratt, R. R., & Jones, R. W. (1987). Music and medicine: A
sessment of musical performance by using the method partnership in history. In R. Spintge & R. Droh (Eds.),
of continuous judgment by selected description. Music Music in medicine (pp. 377–388). Berlin: Springer-
Perception, 8, 251–276. Verlag.
Nettelbeck, T., Henderson, C., & Willson, R. (1989). Com- Quantz, J. J. (1752). Versuch einer Anweisung die Flöte
municating emotion through sound: An evaluation of traversiere zu spielen. (Revised new printing with
Clynes’ theory of sentics. Australian Journal of Psy- comments by Arnold Schering, 1926, C. P. Kahnt,
chology, 41, 25–36. Leipzig.)
Nettheim, N. (1979). Comment on a paper by Gabriel on Rapoport, E. (1996). Emotional expression code in opera
Cooke’s theory. Psychology of Music, 7, 32–33. and lied singing. Journal of New Music Research, 25,
Nielsen, F. V. (1983). Oplevelse av musikalsk spœnding 109–149.
[Experience of musical tension]. Copenhagen: Akadem- Rigg, M. G. (1937a). Musical expression: An investigation
isk Forlag. (Includes summary in English.) of the theories of Erich Sorantin. Journal of Experimen-
Nielsen, F. V. (1987). Musical “tension” and related con- tal Psychology, 21, 442–455.
cepts. In T. A. Sebeok & J. Umiker-Sebeok (Eds.), The Rigg, M. G. (1937b). An experiment to determine how ac-
semiotic web ’86. An international yearbook (pp. 491– curately college students can interpret the intended
513). Berlin: Mouton de Gruyter. meanings of musical compositions. Journal of Experi-
Nielzén, S., & Cesarec, Z. (1981). On the perception of mental Psychology, 21, 223–229.
emotional meaning in music. Psychology of Music, 9, Rigg, M. G. (1939). What features of a musical phrase have
17–31. emotional suggestiveness? Publications of the Social
Nielzén, S., & Cesarec, Z. (1982). Emotional experience of Science Research Council of the Oklahoma Agricul-
music as a function of musical structure. Psychology of tural and Mechanical College, No. 1.
Music, 10, 7–17. Rigg, M. G. (1940a). The effect of register and tonality
North, A., & Hargreaves, D. J. (1997). Music and consumer upon musical mood. Journal of Musicology, 2, 49–61.
behaviour. In D. J. Hargreaves & A. C. North (Eds.), The Rigg, M. G. (1940b). Speed as a determiner of musical
social psychology of music (pp. 268–289). Oxford: Ox- mood. Journal of Experimental Psychology, 27, 566–
ford University Press. 571.
Oatley, K. (1992). Best laid schemes. The psychology of Rigg, M. G. (1964). The mood effects of music: A compar-
emotions. Cambridge, MA: Harvard University Press. ison of data from four investigators. Journal of Psy-
Ohgushi, K., & Hattori, M. (1996a). Emotional communi- chology, 58, 427–438.
cation in performance of vocal music. In B. Pennycook Robazza, C., Macaluso, C., & D’Urso, V. (1994). Emotional
& E. Costa-Giomi (Eds.), Proceedings of the Fourth In- reactions to music by gender, age, and expertise. Per-
ternational Conference on Music Perception and Cog- ceptual and Motor Skills, 79, 939–944.
nition (pp. 269–274). McGill University, Montreal, Can- Rosenthal, R. (1982). Judgment studies. In K. R. Scherer,
ada. & P. Ekman (Eds.), Handbook of methods in nonverbal
Ohgushi, K., & Hattori, M. (1996b). Acoustic correlates of behavior research (pp. 287–361). Cambridge: Cam-
the emotional expression in vocal performance. Paper bridge University Press.
presented at the Third Joint Meeting of the Acoustical Ross, B. H., & Spalding, T. L. (1994). Concepts and cate-
Society of America and the Acoustical Society of Japan, gories. In R. J. Sternberg (Ed.), Thinking and problem
Honolulu, Hawaii, 2–6 December. solving (2nd ed., pp. 119–150). New York: Academic
applicable copyright law.

Osborne, J. W. (1989). A phenomenological investigation Press.


of the musical representation of extra-musical ideas. Rousseau, J. J. (1761/1986). Essay on the origin of lan-
Journal of Phenomenological Psychology, 20, 151–175. guages. In J. H. Moran & A. Gode (Eds.), On the origin
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The of language: Two essays (pp. 5–74). Chicago: University
measurement of meaning. Urbana, IL: University of Il- of Chicago Press.
linois Press. Russell, J. A. (1980). A circumplex model of affect. Journal
Panksepp, J. (1998). Affective neuroscience: The founda- of Personality and Social Psychology, 39, 1161–1178.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451
534 PART IV. EXPRESSION OF EMOTION

Sadie, S. (Ed.). (1980). The new Grove dictionary of music expression of emotions in sung performance. In O. Fu-
Copyright © 2003. Oxford University Press. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or

and musicians (6th ed.). London: Macmillan. jimura & M. Hirano (Eds.), Vocal fold physiology: Voice
Scherer, K. R. (1995). Expression of emotion in voice and quality control (pp. 217–229). San Diego, CA: Singular
music. Journal of Voice, 9, 235–248. Press.
Scherer, K. R., & Oshinsky, J. S. (1977). Cue utilization in Tagg, P. (1979). Kojak—50 seconds of television music. To-
emotion attribution from auditory stimuli. Motivation wards the analysis of affect in popular music. Studies
and Emotion, 1, 331–346. from Gothenburg University, Department of Musicol-
Scherer, K. R., & Zentner, M. R. (2001). The emotional ef- ogy, No. 2. Gothenburg, Sweden: Department of Musi-
fects of music: Production rules. In P. N. Juslin & J. A. cology.
Sloboda (Eds.), Music and emotion: Theory and re- Tarasti, E. (1994). A theory of musical semiotics. Indian-
search (pp. 361–392). New York: Oxford University apolis: Indiana University Press.
Press. Terwogt, M. M., & van Grinsven, F. (1991). Musical ex-
Schubert, E. (1999a). Measurement and time series anal- pression of mood states. Psychology of Music, 19, 99–
ysis of emotion in music. Unpublished doctoral disser- 109.
tation. University of South Wales, Sydney, Australia. Thompson, W. F., & Robitaille, B. (1992). Can composers
Schubert, E. (1999b). Measuring emotion continuously: express emotions through music? Empirical Studies of
Validity and reliability of the two-dimensional the Arts, 10, 79–89.
emotion-space. Australian Journal of Psychology, 51, Tomkins, S. S. (1962). Affect, imagery, and consciousness:
154–165. The positive affects. New York: Springer.
Schubert, E. (2001). Continuous measurement of self- Trussoni, S. J., O’Malley, A., & Barton, A. (1988). Human
report emotional response to music. In P. N. Juslin & J. emotion communication by touch: A modified repli-
A. Sloboda (Eds.), Music and emotion: Theory and re- cation of an experiment by Manfred Clynes. Perceptual
search (pp. 393–414). New York: Oxford University and Motor Skills, 66, 419–424.
Press. Unger, H. H. (1979). Die Beziehungen zwischen Musik und
Scruton, R. (1997). The aesthetics of music. Oxford, En- Rhetorik im 16–18. Jahrhundert. Hildesheim, Germany:
gland: Oxford University Press. Georg Olms Verlag.
Seashore, C. E., Lewis, D., & Saetveit, J. (1939/1960). Sea- de Vries, B. (1991). Assessment of the affective response
shore measures of musical talents. New York: Psycho- to music with Clynes’s sentograph. Psychology of Mu-
logical Corporation. sic, 19, 46–64.
Senju, M., & Ohgushi, K. (1987). How are the player’s Wagner, A. H. (1932). Remedial and artistic development
ideas conveyed to the audience? Music Perception, 4, of the vibrato. In C. E. Seashore (Ed.), University of
311–324. Iowa studies in the psychology of music. Vol. I: The
Sherman, M. (1928). Emotional character of the singing vibrato (pp. 166–212). Iowa City: University of Iowa.
voice. Journal of Experimental Psychology, 11, 495– Wallbott, H. G., & Scherer, K. R. (1986). Cues and channels
497. in emotion recognition. Journal of Personality and So-
Shove, P., & Repp, B. H. (1995). Musical motion and per- cial Psychology, 51, 690–699.
formance: Theoretical and empirical perspectives. In J. Watson, K. B. (1942). The nature and measurement of mu-
Rink (Ed.), The practice of performance. Studies in mu- sical meanings. Psychological Monographs, 54, 1–43.
sical interpretation (pp. 55–83). Cambridge, England: Watt, R. J., & Ash, R. L. (1998). A psychological investi-
Cambridge University Press. gation of meaning in music. Musicae Scientiae, 2, 33–
Siegwart, H., & Scherer, K. R. (1995). Acoustic concomi- 53.
tants of emotional expression in operatic singing: The Wedin, L. (1969). Dimension analysis of emotional ex-
case of Lucia in Ardi gli incensi. Journal of Voice, 9, pression in music. Swedish Journal of Musicology, 51,
249–260. 119–140.
Sloboda, J. A. (1985). The musical mind: The cognitive Wedin, L. (1972a). Evaluation of a three-dimensional
psychology of music. Oxford: Oxford University Press. model of emotional expression in music. Reports from
Sloboda, J. A., & Juslin, P. N. (2001). Psychological per- the Psychological Laboratories, University of Stock-
spectives on music and emotion. In P. N. Juslin & J. A. holm, No. 349. Stockholm, Sweden: Department of Psy-
Sloboda (Eds.), Music and emotion: Theory and re- chology.
search (pp. 71–104). New York: Oxford University Wedin, L. (1972b). Multidimensional scaling of emotional
Press. expression in music. Swedish Journal of Musicology,
Smith, L. D., & Williams, R. N. (1999). Children’s artistic 54, 1–17.
responses to musical intervals. American Journal of Wedin, L. (1972c). Multidimensional study of perceptual-
Psychology, 112, 383–410. emotional qualities in music. Scandinavian Journal of
Sopchak, A. L. (1955). Individual differences in responses Psychology, 13, 241–257.
to different types of music in relation to sex, mood, and Weld, H. P. (1912), An experimental study of musical en-
other variables. Psychological Monographs: General joyment. American Journal of Psychology, 23, 245–
and Applied, 69, 1–20. 308.
Spintge, R., & Droh, R. (Eds.). (1987). Musik in der Medizin Wigram, T., Saperston, B., & West, R. (Eds.). (1995). The
applicable copyright law.

[Music in medicine]. Berlin: Springer-Verlag. art and science of music therapy: A handbook. Chur,
Stern, D. N. (1985). The interpersonal world of the infant. Switzerland: Harwood Academic Publishers.
New York: Basic Books. Zuckerkandl, V. (1956). Sound and symbol. Princeton, NJ:
Sundberg, J., Iwarsson, J., & Hagegård, H. (1995). A singer’s Princeton University Press.

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/15/2022 11:06 PM via UNIVERSITY COLLEGE LONDON
AN: 129691 ; Richard J. Davidson, Klaus R. Scherer, H. Hill Goldsmith.; Handbook of Affective Sciences
Account: s8454451

You might also like