THOMPSON Juslin Sloboda-Ch-27-Libre

Download as pdf or txt
Download as pdf or txt
You are on page 1of 34

c u .

v 1 v v : ,
CROSS-CULTURAL
SIMILARITIES AND
DIFFERENCES
william forde thompson and
laura-lee balkwill
Over the past century, research on the relation between music and emotion has
overwhelmingly focused on the perceptions and experiences of Western listeners in
response to Western tonal music. This focus has largely been a pragmatic consequence
of the challenges of carrying out research in non-Western contexts, but it has left many
important questions unanswered. To what extent are emotional aspects of music simi-
lar across cultures? Are there general principles that might account for the connection
between music and emotion in all (or most) cultures, or is that connection unique to
each culture, and perhaps non-existent in some cultures? How might we conduct a
cross-cultural study of emotion and music without the conclusions being corrupted
by cultural biases?
In this chapter, we review empirical studies of music and emotion that involve a
cross-cultural comparison, and we outline prevailing views on the implications of such
studies. Cross-cultural investigations can provide important insights into the contri-
bution of universal and cultural associations between music and emotion. They can
also be used to validate psychological theories of music and emotionwhich are over-
whelmingly supported by research that privileges Western tonal music.
Reecting the current state of research in this area, our focus is on the perception of
emotion in music (decoding) rather than the induction of affective states by music.
The question of whether individuals from different cultures have comparable affec-
tive experiences in response to music is largely unknown, but is a rich area for future
research (see Chapter o, this volume). We also discuss cross-cultural studies of
27-Juslin & Sloboda-Ch-27.indd 755 7/24/2009 7:50:51 PM
,,o cross-cultural similarities and differences
emotional prosody. Speech prosody refers to the vocal qualities of speech and includes
intonation (pitch variation), stress, and timing. It signals points of emphasis, indicates
a statement or question, and conveys emotional connotations (Darwin, I8,:; Frick,
I,8,; Juslin & Laukka, :oo,). Prosodic communication of emotion can occur indepen-
dently of verbal comprehension (Kitayama & Ishii, :oo:).
We discuss research on speech prosody for two reasons. First, music and speech
prosody share important acoustic attributes and may use a common code for emo-
tional communication (Juslin & Laukka, :oo,). Second, current theory and evidence
on cross-cultural emotional decoding is more advanced for prosodic materials than for
musical materials and, as such, may provide a model for future cross-cultural research
on music and emotion. We do not provide a detailed examination of ethnographic
studies of specic musical traditions, as such discussions may be found in Chapters o
and , (this volume).
We begin by discussing some theoretical implications of research on cross-cultural
commonalities in the association between music and emotion. We note that cognitive,
ethnomusicological, and sociological approaches complement each other by provid-
ing different levels of explanation and different perspectives on the concept of music.
Section :,.: reviews the central questions arising from cross-cultural research on emo-
tion. In Section :,.,, we outline the cue-redundancy model, developed to account for
cross-cultural similarities and differences in the expression and recognition of emotion
in music. The model accounts for the balance of culture-transcendent and culture-
specic emotional cues across musical genres. It can be used to frame research ques-
tions, and to communicate, compare, and integrate empirical ndings.
Section :,. outlines a broader framework for summarizing existing data on emo-
tional communication, referred to as fractionating emotional systems (FES). FES
extends the cue-redundancy model by accounting for similarities and differences in
emotional communication not only across cultures but also across the auditory chan-
nels of music and speech prosody. FES also accounts for the process of enculturation
that permits the gradual division of musical and prosodic emotional coding systems
as well as the emergence of distinctive systems across cultures. Section :,., provides a
review of cross-cultural studies of music and emotion, while section :,.o reviews cross-
cultural studies of emotion in speech. Section :,., identies some future prospects for
the cross-cultural study of music and emotion.
27.1 Music as a cross-cultural construct
In order to investigate music as a cross-cultural construct, one must confront the tasks
of dening music on the one hand and culture on the other. Merriam (I,o) char-
acterized music as having three aspects: sound, behaviour, and concept. As sound,
music can be dened as a class of auditory signals that are produced by performers
27-Juslin & Sloboda-Ch-27.indd 756 7/24/2009 7:50:51 PM
27. 1 music as a cross-cultural construct ,,,
and perceived by listeners. As behaviour, music is associated with tangible activities
(e.g. performance, dance, ritual) that are often essential to music experience, and that
can be subjected to rigorous psychological, social, and historical analyses. As concept,
music is construed as having specic functions within any social group (Clayton, :ooI;
Cross, :ooo; Dissanayake, :ooI). To a large extent, cross-cultural music cognition has
focused on music as sound, but there is increasing awareness that music is a multimodal
phenomenon. The behavioural aspect of music cannot be treated as a distinct level of
analysis, but is inseparable from perceptions and experiences of music. Indeed, the
visual input from viewing the facial expressions and gestures of a music performer can
profoundly inuence a listeners emotional responses to music (Thompson, Graham,
& Russo, :oo,; Thompson, Russo, & Quinto, :oo,).
The construct of culture has also been scrutinized and debated. It refers to the set of
behaviours, beliefs, social structures, and technologies of a population that are passed
down from generation to generation. It includes social conventions related to art, dress,
manner, dance, music, religion, ritual, and morality. Like music, cultures can be exam-
ined on multiple levels, some of which are nested within others. For example, within
any cultural environment (e.g. Northern India) there are many subcultures (Hebdige,
I,,,). Each subculture, in turn, is associated with its own distinctive habitus: the habits,
beliefs, skills, schemas, and preferences that are tacitly acquired and that shape percep-
tions, behaviours, and experiences (see Chapter o, this volume). The term culture is
not equivalent to country or continent, and is not always associated with any one
geographical region. Moreover, most individuals do not belong to a single culture.
More typically, numerous cultural and sub-cultural inuences can be detected for
any one individual, colliding, merging, and making temporary appearances when the
appropriate context arises. Thus, although cross-cultural studies have often involved
an examination of music materials and practices within specic geographical regions
(e.g. Europe, Japan, Northern India), it should be recognized that the construct of
culture does not restrict studies to such strategies.
Most ethnomusicologists are sensitive to such challenges, and are dubious of attempts
to characterize nations as singular cultures and compare them with one another. In
order to appreciate this disciplinary perspective, it may be useful to note that eth-
nomusicology was known as comparative musicology at the turn of the twentieth
century. This name was abandoned for three reasons. First, the comparative method
as practised at the beginning of the twentieth century was steeped in an implied cultural
hierarchy. Many scholars at the time regarded Western art music as a cultural pinnacle,
and they interpreted music from other cultures as earlier or more primitive stages of
development in the evolution of music. Like the common distinctions between hunt-
er-gatherer, agrarian society, and nation state, comparative musicologists at the
time were preoccupied with characterizing the stage of musical development for each
culture. Second, any comparison involves the establishment of criteria for evaluating
musical practices, and the question of which criteria to adopt raises thorny methodo-
logical issues. How does one compare cultures objectively without inadvertently using
criteria that privilege the culture of the researcher? Most contemporary ethnomusi-
cologists have concluded that there are no neutral criteria. Third, ethnomusicologists
27-Juslin & Sloboda-Ch-27.indd 757 7/24/2009 7:50:51 PM
,,8 cross-cultural similarities and differences
resolved that each culture should be understood on its own terms. The very act of com-
paring musical cultures is therefore suspect. The change in name from comparative
musicology to ethnomusicology reected these concerns.
Ethnomusicologists often point to the work of Alexander Ellis as a seminal event in
the birth of their discipline (On the musical scales of various nations; Ellis, I88,). At
the time Ellis was working, many scholars believed that different musical cultures were
progressing along a developmental path that approached Western European art music.
Evidence of this could purportedly be found in the instruments of other cultures whose
tuning was considered crude approximations of the more developed Western scale.
The crude tuning of these instruments was regarded as tacit evidence of the rudi-
mentary listening skills of people in other cultures. Working at the British Museum,
Ellis measured the tuning of instruments from different cultures, and observed that
the varied tunings in other cultures were not merely poor approximations of Western
tuning. Rather, he found that different instruments pointed to unique and stable tun-
ing systems that were categorically different from the Western scale. The instruments
were not merely technologically nave efforts along some path toward the Western
scale, but sophisticated technological efforts in their own right. Elliss results thereby
conicted with notions of a cultural hierarchy.
Although the abandonment of notions of cultural hierarchies was a welcome devel-
opment, the abandonment of comparative approaches altogether by ethnomusicol-
ogy has been unfortunate. A full understanding of the cognitive basis of music is not
possible unless similarities and differences across cultures in the perception, experi-
ence, and production of music are taken into account. Comparisons across cultures
are difcult methodologically and susceptible to researcher bias, but renouncing all
comparative research is hardly a productive response to such challenges. Elliss own
achievements depended on comparative measurements within and between cultures,
and in recent years researchers have employed reciprocal or counterbalanced methods
in which test materials and subjects are recruited from two or more cultures. These
methods may not eliminate cultural bias entirely, but the empirical data provide useful
grist for theorizing about the cognitive basis of music.
There are several approaches to the cross-cultural study of music, including ethno-
musicological, anthropological, sociological, and cognitive. Among the most inu-
ential sociological approaches is that of Alan Lomax (I,,,, I,,o, I,,8). His method,
called cantometrics, involved characterizing song styles from different cultures using
rating scales for features such as intensity, tempo, rhythmic complexity, interval width,
embellishment, register, and tension. The procedure allowed Lomax to analyse and
compare song styles cross-culturally. This research led him to conclude that the emo-
tional properties of music are central to understanding cross-cultural similarities and
differences in musical behaviours: music somehow expresses emotion; therefore,
when a distinctive and consistent musical style lives in a culture or runs through sev-
eral cultures, one can posit the existence of a distinctive set of emotional needs or drives
that are somehow satised or evoked by this music (Lomax, I,o:, p. :,). Put simply,
song style mirrors and reinforces cultural style.
Cross-cultural music cognition may be dened as the exploration of similarities and dif-
ferences in cognitive processes and emotional experiences for music across cultures, and
27-Juslin & Sloboda-Ch-27.indd 758 7/24/2009 7:50:51 PM
27. 1 music as a cross-cultural construct ,,,
can be used to differentiate universal and culture-specic determinants of mature forms
of music understanding. In general, cross-cultural music cognition has a more restricted
focus than ethnomusicology, which applies research strategies adapted from cultural
anthropology, sociology, and other disciplines in order to understand specic musical
systems and traditions within a social or cultural context (see Chapter o, this volume).
All cross-cultural research has a specic set of challenges above and beyond those
that are part of any research endeavour. The difculties of identifying geographical
correlates of culture, of isolating one culture from other cultural inuences, and the
potential for researcher bias have already been noted. How effectively a study addresses
such pitfalls and challenges is an important factor in interpreting its ndings. Of critical
importance in cross-cultural research is an awareness of ones own cultural perspec-
tives and how they can bias every facet of the research, from the question being asked,
to the methodology employed, to the analysis strategy employed, to the interpretation
of observations. Consulting with members of the cultures under investigation is one
important way of increasing an awareness of ones biases.
Perhaps the greatest challenge to studies of the expression and recognition of emotion
across cultures is establishing cross-cultural equivalence of conceptual and empirical
variables. Much has been written about the use of emotion labels to conceptualize the
communication of emotions (see Wierzbeca, I,,:, for a review). Typically, research-
ers employ the strategy of back-translation to create an equivalent set of instructional
materials in the language of each culture of interest. However, even when linguistic
equivalence has been achieved, conceptual equivalence may still prove elusive. For
example, the word for anger in English and the word for anger in Japanese may not be
interpreted by members of these cultures with the same degree of intensity or as signi-
fying the same situations or behaviours. Therefore, establishing equivalence must be
handled with care and sensitivity to cultural mores.
Ethnomusicologists and anthropologists have drawn attention to many nuances of
difference in musical understandings across cultures, but evidence from cross-cultural
music cognition suggests that music experiences are also constrained in important ways
by the nature of our physical environment, the structure of the auditory system, and
evolved strategies of perceptual and cognitive processing (Huron, :ooo; Patel, :oo8;
Thompson & Schellenberg, :ooo). Such constraints provide a foundation upon which
processes of enculturation lead to additional layers of culture-specic understandings
of music. They are an important source of commonalities in music cognition across
cultures, and may exert a powerful inuence on emotional responses to music. They do
not necessarily lead to musical universals if environment-specic or culture-specic
inuences overwhelm their impact. However, the effects of constraints can sometimes
lead to universal, or near-universal, aspects of music cognition or experience, giving
rise to striking commonalities across cultures. The following observations have been
proposed as likely candidates for musical universals:
a processing advantage for music built on a small number of discrete pitch levels

that are spaced unevenly (e.g. the major scale);


greater sensitivity to pitch contour than to exact intervals;

perceived similarity of pitches separated by an octave;

27-Juslin & Sloboda-Ch-27.indd 759 7/24/2009 7:50:51 PM


,oo cross-cultural similarities and differences
greater sensitivity to pitch relations than to absolute pitch values;

sensitivity to sensory consonance and dissonance;

a tendency to perceive sequences of pitches that are proximal in pitch as part of the

same group;
a processing advantage for music that contains a regular temporal pattern of stress;

and
a tendency to perceive pitches as having different levels of stability, with one pitch

class often acting as a point of reference for other pitch classes.


Such commonalities can be explained with reference to one or more cognitive, envi-
ronmental, genetic, and biophysical constraints, which may be collectively referred to
as system constraints. The similarity across cultures of sung phrase lengths, for example,
may arise in part because human short-term memory has limited capacity and would
be burdened by excessively long phrases (an evolved feature of cognitive architecture).
In addition, the ability to produce a melodic phrase vocally is constrained by the bio-
physics of singing, in that air resistance, breathing apparatuses, and oxygen require-
ments of humans are similar across cultures.
Experience and learning invariably build upon, expand, and in some cases can
counter the constraints imposed by cognitive, environmental, genetic, and biophysical
factors. For example, there is considerable diversity in musical scales across cultures,
including the amount of tonal material and the number of scale notes (e.g. I: tones
in the Western chromatic scale, seven tones in the Western major, Byzantine, Gypsy
and Hungarian scales, ve tones in Celtic, West African and Indonesian scales), the
intervals formed by scale notes, and the tuning of scale notes (Patel, :oo8). Indeed, the
inuence of experience and learning is vast and leads to patterns of thought, behaviour,
and aesthetic sensibility that are often unique to specic subcultures, as documented
by the thick descriptions provided by ethnomusicologists and anthropologists (Geertz,
I,,,; Nettl, I,8,). When differences in musical systems, behaviours, and experiences
are observed from one culture or subculture to the next, they can usually be interpreted
as effects of enculturation.
Cross-cultural music cognition is a valuable strategy for understanding the com-
plex interplay between early predispositions and enculturation, but other scientic
approaches are also essential to a full understanding of the issues involved. In particu-
lar, developmental approaches, neuroscientic data, and evolutionary psychology are
particularly important for interpreting the nature of culture-transcendent and culture-
specic sources of musical understanding (see Cross, :oo,; Livingstone & Thompson,
in press; and Chapters ,, I:, and :,, this volume).
Considerable caution must be taken when attempting to draw conclusions from
cross-cultural similarities in musical practices and structures. In particular, obser-
vations of apparent universals in music have often been mistakenly interpreted as
implicating the involvement of genetic (innate) or hard-wired biological determi-
nants of musical systems and behaviours. There are important qualications that
must accompany this interpretation. First, genetic constraints on musical behav-
iours that result from a process of phylogenesis represent just one of several potential
27-Juslin & Sloboda-Ch-27.indd 760 7/24/2009 7:50:51 PM
27. 1 music as a cross-cultural construct ,oI
system constraints. Commonalities in musical behaviours across cultures also arise
from biophysical constraints, such as oxygen consumption requirements, or similari-
ties across cultures in acoustic environments, such as the tendency for sequences of
pitched sounds emanating from a single source to exhibit pitch proximity. Second,
many apparent constraints on musical systems may not be specic to music. Some
may arise as a by-product of general auditory processing or the limits of short-term
memory (Justus & Hutsler, :oo,; Patel, :oo8). Third, commonalities in music behav-
iours across cultures frequently arise from transcultural diffusion and interaction. In
the twenty-rst century, it is difcult to nd individuals who have not been exposed
to a large number of musical styles and tuning systems from a wide range of cultures
and subcultures. Therefore, arguments for innately determined constraints on musical
systems are compelling only when they are grounded in evolutionary biology, when
other potential constraints have been ruled out, or when accompanied by direct genetic
evidence. Indeed, many phenomena that appear to be innate may prove to be learned,
because learning takes place in an interaction between organisms and physical envi-
ronments that is ubiquitous around the world. As such, many apparent universals of
music probably have no genetic basis.
Conversely, differences in musical practices across cultures do not imply that genetic
or innate factors play no role in such practices. Rather, cultural differences result
from ontogenetic processes, which are themselves products of natural selection that
permit cognitive mechanisms to adapt to environmental conditions. Such processes
are guided by knowledge and expectancies, and their operation changes with experi-
ence (Heyes, :oo,). Ontogenetic processes account for differences across cultures in
the communication of emotion in music, but they do not guarantee such differences
because some environmental and biophysical inuences on musical systems are ubiq-
uitous and can lead to musical universals in the absence of any innate inuences on
music cognition. This conception of phylogenetic and ontogenetic inuences has the
potential to resolve some of the abiding concerns surrounding the study of similarities
and differences in the musical practices of different cultures.
The connection between music and emotion may be examined at several levels
of analysis relevant to cross-cultural comparisons, with some levels relying more on
phylogenetic determinants and other levels relying more on ontogenetic processes.
Environmental signals that remain stable across geographical regions and over long
periods of evolutionary time are most economically handled by phylogenetic determi-
nants. For example, the startle response that follows an unexpected and intensely loud
sound is instinctive; it is a genetically determined reex that requires no process of
learning. Other phylogenetically-determined mechanisms include early feature detec-
tion and orienting reexes. More generally, psychological processes, including those
associated with emotional responses to music, vary in the extent to which they permit
adaptive modication in response to changing environmental conditions.
Environmental signals that are unstable across time or geographical location are
most effectively handled by ontogenetic processes. Indeed, ethnomusicologists have
revealed that there is a high degree of semiotic variability in our auditory environ-
ment, such that the role of learning and enculturation is indispensable to full (adult)
27-Juslin & Sloboda-Ch-27.indd 761 7/24/2009 7:50:51 PM
,o: cross-cultural similarities and differences
understanding of the many connections between music and emotion. For example,
the increased tension that is experienced by a North American listener following the
occurrence of a nondiatonic tone results from an ontogenetic process: it requires a pro-
cess of enculturation to a diatonic tonal system and is manifested only after a certain
stage of development. Indeed, children are not sensitive to differences in mode until
after the age of ve (Dalla Bella, Peretz, Rousseau, & Gosselin, :ooI).
Phylogenetic determinants are responsible for predispositions that shape initial
perceptions of musical attributes, such as early preferences for consonance over dis-
sonance, and they exert culture-transcendent forces on the connection between music
and emotion (Dowling & Harwood, I,8o). Ontogenetic processes build upon such
predispositions and permit cultural fractionation to occur in the emotional connota-
tions of music across musical traditions, cultures, styles, and time periods. Examples
include the banning of the tritone interval from Western tonal music in sacred music
during the middle ages, and the characterization of certain kinds of music as socially or
emotionally dangerous at various points in history (e.g. music composed in the mix-
olydian scale in ,,o BC, ragtime in the I,oos, heavy metal in the I,8os). Mature forms
of musical understanding reect a network of cognitive processes that combine phy-
logenetic and ontogenetic sources of adaptive features. Adaptive features that have a
phylogenetic source were favoured by natural selection and are specialized to respond
to highly stable features in the environment. Adaptive features that have an ontoge-
netic source were not favoured by natural selection but are generated in the course of
development, and vary across cultures.
All musical systems implicate a dense combination of emotional and structural
signals, some of which are culture-transcendent and handled by phylogenetically
determined cognitive mechanisms, others that are culture-transcendent but nonethe-
less handled by ontogenetically determined cognitive mechanisms, and still others that
are historically and culturally specied and again handled by ontogenetically deter-
mined cognitive mechanisms. Phenomena that are culturally transcendent are often
identied by experimental strategies developed in cognitive science. Behaviours that
are historically and culturally specied are often identied through ethnographic strat-
egies, commonly applied in ethnomusicology, sociology, and anthropology.
27.2 Emotion as a cross-cultural
construct
Emotions are, in an important sense, biological mechanisms of preparedness (Darwin,
I,,:; Ekman, I,,:; Huron, :ooo; Lazarus, I,,I; Tooby & Cosmides, I,,o), and this
functional property of emotions has important implications for cross-cultural com-
parisons. Emotional states often have tangible biological functions, facilitating survival
responses such as ghting, eeing, or freezing. Emotional experiences and behaviours
27-Juslin & Sloboda-Ch-27.indd 762 7/24/2009 7:50:51 PM
27. 2 emotion as a cross-cultural construct ,o,
are observed in all cultures, and may be interpreted as universal responses to signicant
events. Some events are signicant regardless of ones cultural environment, and may
generate consistent emotional responses across cultures. Being in imminent danger
of attack by a wild animal is likely to generate heightened arousal and fear in any indi-
vidual, whether from Bali, China, Finland, or Canada. Many events acquire mean-
ing through social construction, however, and an understanding of their signicance
requires knowledge of cultural traditions. The convention of bowing as a signal of
respect and politeness in some cultures can be very specic: bowing too briey or for
too long can have signicant connotations that elicit a strong emotional response.
According to Matsumoto (I,8,), emotions are biologically programmed, but the
process of learning to perceive and express emotions is dependent on enculturation.
Categories of emotion are also inuenced by cultural factors. Broad dimensions such
as valence and arousal may be applicable across cultures, while specic emotional cat-
egories such as the Indian rasa adbhutam (meaning wondrous) may not translate well
across cultures (Deva, I,,,; Russell, I,,). Moreover, the notion that typical interac-
tions involve the communication of a singular denable emotion may be unrealistic.
In many if not most contexts, people experience varying combinations of emotions,
and musical expressions of human experience often reect such mixed emotions. If
decoding individual emotions across cultures is difcult, decoding expressions of
mixed emotions is an even greater challenge. To our knowledge, investigations of
sensitivity to mixed emotions across cultures have yet to be conducted.
Landmark studies by Ekman and colleagues suggested that facial emotion recogni-
tion is universal for a number of basic emotions (Biehl et al, I,,,; Ekman, I,,o, I,,:;
Ekman, Sorenson, & Friesen, I,o,). Ekman also emphasized cultural differences in
display rules for emotional communication, but it has nevertheless been suggested
that many of the early studies of emotional communication underestimated the role
of enculturation by relying on forced-choice response formats and posed expressions
(Russell, I,,), and because the researchers chose to focus on agreement across cultures
rather than disagreement. The challenge of interpretation inherent in such studies is
exemplied by a study that revealed that members of Bahinemo tribes perceived all
faces of Americans as angry (Sorensen, I,,,). Such ndings underscore the difculty
of decoding emotions across cultural boundaries. Indeed, some scholars have even
argued that certain emotions are unique to particular cultures, such that cross-cultural
decoding is essentially impossible (Briggs, I,,o; Lutz, I,88).
A number of researchers have emphasized the signicance of such cross-cultural
differences (Matsumoto, I,8,; Mesquita & Frijda, I,,:; Mesquita et al, I,,,). This
research has led to the speculation that many cultures have distinctive norms for dis-
playing and decoding emotions, or even culture-specic emotions. Such norms may
function to maintain positive social interactions, and are especially apparent for the
expression and recognition of negative emotions, which have the greatest potential
to disrupt social interactions (Matsumoto, I,8,). The role of cultural norms can also
explain the phenomenon of ethnic or in-group bias. For example, when emotional
speech is recorded from an individual, decoding accuracy is often highest by mem-
bers of the same cultural group from which the utterance was recorded (Kilbride
27-Juslin & Sloboda-Ch-27.indd 763 7/24/2009 7:50:51 PM
,o cross-cultural similarities and differences
& Yarczower, I,8,; Mesquita & Frijda, I,,:). Although the study of ethnic bias has
received considerable attention, a full understanding of the phenomenon is unlikely to
emerge soon, because results are inconsistent across studies, emotional channels, and
cultures under consideration (Elfenbein & Ambady, :oo:).
Differences in emotion recognition across cultures may also arise from language
differences. For example, some languages have highly expressive terms for certain
emotional concepts and lack terms for expressing other emotional concepts. Such dif-
ferences in language apply to the expression of emotion across communication chan-
nels. Note, however, that the lack of an emotional term does not imply the absence of
that emotion in a culture, but reects a lack of emphasis on a particular subordinate
category (e.g. frustration) within a broader emotion category (e.g. anger) in the social
discourse of that culture (Boucher, I,,,; Johnson-Laird & Oatley, I,8,; Russell, I,,I).
In addition, each communication channel is associated with channel-specic prop-
erties that can sometimes interfere with cross-cultural emotional communication.
With respect to speech prosody, differences in emotional decoding accuracy across
cultures can also arise because culture-transcendent prosodic cues are masked dif-
ferentially by attributes of voice quality that are unfamiliar to non-speakers of that
language. As an example, pitch changes involved in tonal languages like Mandarin
may confuse or mask the interpretation of emotional intent to individuals who are
unfamiliar with that language. Similarly, difculties in emotional decoding of music
can arise because unfamiliar conventions of composition and performance overwhelm
impressions of the music, masking culture-transcendent emotional cues.
27.3 The cue-redundancy model
A basic property of music that is shared by all cultures is that it involves multiple
features that must be analysed by the auditory system. Even silence is registered by the
brain and contrasted with the possibility of sound. One or more acoustic features such
as intensity, tempo, pitch height, timbre, sensory dissonance, rhythmic complexity,
and harmony may be present in the music, and each feature varies with the emotional
connotations of the music. Just as instrument membranes have resonant frequencies
that allow them to vibrate in sympathy with different sounds, the various acoustic
features of music can be said to resonate to varying degrees with the emotional char-
acter of the music. In turn, the auditory receptors for these music features themselves
resonate with the emotional character of the music.
Each emotion is correlated to some extent with different types of changes in acoustic
features, but many of these associations are moderate or even weak, and so no indi-
vidual feature can be relied upon to determine an intended emotion. Rather, the full
set of features available must be evaluated in a probabilistic manner, with different
features weighted according to the strength of their association with each emotion
27-Juslin & Sloboda-Ch-27.indd 764 7/24/2009 7:50:51 PM
27. 3 the cue-redundancy model ,o,
(see Chapter I, this volume). A yet-to-be-discovered mechanism that summarizes
the available evidence from multiple emotional cues determines the emotion that is
ultimately perceived and experienced. Because cues overlap in their emotional implica-
tions, the process of emotional decoding has the advantage of cue redundancy, allow-
ing emotional communication to occur through convergent evidence and appraisals
based on family resemblances rather than on dening features (Rosch & Mervis, I,,,;
Wittgenstein, I,,,).
Thus, for example, one might imagine two musicians from disparate cultures
(A and B) each composing a piece of music on the theme of joy. They would probably
compose in a scale they know to be associated with joy in their culture. Musician A
might use intricate harmonies, a simple rhythm, and compose for a high-pitched wind
instrument that is instantly recognized by her people as the joy ute. Musician B
might compose an unaccompanied melodic line, a complex rhythm, and compose for
a stringed instrument that is widely associated with joy by his people. They might both
perform their pieces at a fast tempo and moderate amplitude.
Although there would probably be many surface differences between the two com-
positions, each piece would contain psychophysical cues (e.g. rate, intensity, pitch
height) that have emotional connotations across cultures, and culture-specic cues
(e.g. cadences, specic instruments, harmonic progressions) that enhance the recogni-
tion process for members of the same culture. Balkwill and Thompson (I,,,) dened
psychophysical cues as any property of sound that can be perceived independent of
musical experience, knowledge, or enculturation (p. ). However, psychophysical
cues are not merely the result of innate processes but also reect other system con-
straints, as described earlier. For example, music that communicates high-arousal
emotions may be characterized by attributes that reect the increased oxygen require-
ments associated with high-arousal states. Because the connection between oxygen
requirements and arousal states is ubiquitous, certain attributes of high-arousal music
are likely to occur across cultures. The presence of such attributes is unlikely to be dic-
tated by direct genetic encoding, but may emerge indirectly as a result of biophysical
constraints.
Examples of psychophysical cues include sound intensity, rate (tempo), melodic
complexity, melodic contour, pitch range, rhythmic complexity, dynamics, and
timbre. Listeners have the advantage of cue redundancy because most music commu-
nicates emotion through multiple redundant cues (e.g. Juslin, :ooo; :ooI; Thompson
& Robitaille, I,,:). Listeners who share cultural experiences with the singer/songwriter
have additional benets of cue redundancy, whereas listeners from another culture
must rely on psychophysical cues in order to recognize the emotional connotation of
the music.
Figure :,.I illustrates the cue-redundancy model (CRM) proposed by Balkwill and
Thompson (I,,,). A central tenet of the CRM is that listeners can appreciate affec-
tive qualities of unfamiliar music by attending to psychophysical cues. Consciously or
intuitively, composers and performers draw upon these culture-transcendent cues as
well as culture-specic conventions in order to express emotion in music. Listeners,
in turn, attend to either or both sources of emotional meaning. When psychophysical
27-Juslin & Sloboda-Ch-27.indd 765 7/24/2009 7:50:51 PM
,oo cross-cultural similarities and differences
cues in the music are not overwhelmed by culture-specic cues, listeners from outside
that culture are more likely to be able to decode the emotional intent.
Emotional decoding is affected by familiarity with the conventions of the tonal sys-
tem and sensitivity to psychophysical cues. Listeners who are familiar with a musical
style should nd it relatively easy to decode emotional meaning in that music, because
they can draw from both culture-specic and psychophysical cues. As such, in-group
advantages are a basic prediction of the CRM. However, because studies of in-group
advantages have yielded inconsistent results for other channels of emotional commu-
nication, their validity for emotional messages in music should be subjected to rigor-
ous cross-cultural evaluation.
When cultural-specic cues are unfamiliar or absent, listeners may still attend to
psychophysical cues such as tempo and intensity. These cues provide listeners with
a general understanding of the intended emotion even for unfamiliar musical styles.
Psychophysical cues are especially powerful signals because their interpretation
requires no knowledge of musical conventions. Indeed, cues such as tempo and inten-
sity have emotional signicance in other channels such as speech prosody, animal calls,
and alert signals.
The CRM can also be extended to the domain of speech prosody. Vocal expres-
sions of anger are often marked by fast tempo and greater intensity; vocal expressions
of joy are often fast in tempo and have a large pitch range, and vocal expressions of
sadness are typically slow in tempo and low in pitch register (Juslin & Laukka, :oo,).
When speech prosody and music convey the same emotion, they also tend to share
many of the same psychophysical cues (Ilie & Thompson, :ooo). These common-
alities suggest that emotions are communicated through psychophysical cues that
Fig. 27.1 The cue-redundancy model (CRM), origi-
nally proposed by Balkwill & Thompson (1999).
27-Juslin & Sloboda-Ch-27.indd 766 7/24/2009 7:50:51 PM
27. 4 fractionating emotional systems ,o,
not only transcend cultural boundaries but are manifested in different channels of
communication.
27.4 Fractionating emotional systems
The concept of fractionating emotional systems (FES) extends the implications of the
CRM, providing a broader framework for summarizing and interpreting cross-cultural
data on emotional communication. Like the CRM, the FES framework acknowledges
culture-transcendent and culture-specic inuences on the emotional dimension of
music and prosody. Ontogenetic processes allow for enculturation to occur, giving
rise to culture-specic and domain-specic emotional cues and thereby differentiating
or fractionating emotional systems in different cultures and domains. Phylogenetic
processes provide a point of convergence between emotional channels and cultural
groups; ontogenetic processes provide a mechanism for distinguishing and focusing
emotional channels and cultural groups.
Psychophysical cues are represented as a foundation for emotional communication
that is common across cultures and auditory domains (music and speech prosody).
This foundation is presumed to result from phylogenetic adaptations and other system
constraints (ubiquitous environmental or biophysical inuences), and is an important
source of commonalities in emotional communication across cultures and auditory
domains. For example, evidence suggests that infants prefer infant-directed song to
infant-directed speech, just as they prefer infant-directed speech to affectively neu-
tral speech (Nakata & Trehub, :oo; Singh et al, :oo:). Such preferences may reect
early sensitivity to psychophysical cues along with an instinctive preference for positive
emotional messages.
Emotional communication is maximally adaptive when it involves a combination of
phylogenetic processes that handle stable acoustic features in the environment (e.g. the
correlation between sound intensity and object proximity) and exible ontogenetic
processes for handling emotional signals that vary across environments. As displayed
in Figure :,.:, enculturation builds culture-specic conventions and experiences of
music from the foundations provided by phylogenetic contraints. The fanning out of
triangular shapes represents the differentiating effects of ontogenetic processes, which
promote the development of culture-specic experiences of the links between music
and emotion, and between speech prosody and emotion.
Early in development, before enculturation has had the opportunity to exert a sig-
nicant inuence, there are few differences between individuals across cultures and
few differences in the cues that must be used to decode emotional meaning in music
and speech. With development, enculturation becomes increasingly evident, and leads
to sensitivity to culture-specic and domain-specic emotional cues. This process
27-Juslin & Sloboda-Ch-27.indd 767 7/24/2009 7:50:52 PM
,o8 cross-cultural similarities and differences
permits the partial fractionation of different systems of emotional communication
across cultures and emotional communication systems.
Throughout development, psychophysical cues bootstrap learning in complex ways.
For language acquisition, they interact with prosodic cues to segmentation and syntac-
tic relationships, and may be recruited to draw attention to statistical regularities in the
transitional probabilities among acoustic events (Palmer & Hutchins, :ooo; Saffran,
Aslin, & Newport, I,,o). Moreover, different languages and musical styles permit dif-
ferent degrees of variation in psychophysical cues. For example, the use of pitch varia-
tion for emotional communication is constrained in tone languages and in musical
styles that reply on percussive instruments. Thus, the role of each psychophysical cue
for emotional communication varies as a function of the learning environment.
The cue redundancy model may be understood as a cross section of FES at any one
stage of development, from early development to full maturity. Early in development,
an individual will be predominantly sensitive to psychophysical cues that transcend
cultures and emotional channels. This stage would be represented in the CRM as a
relatively large proportion of overlap in the cues used for emotional communication.
The process of enculturation leads to an increasing contribution of culture-specic
and domain-specic cues to emotional meaning, and is represented in the CRM by an
increased proportion of such cues.
Fig. 27.2 An illustration of the concept of fractionat-
ing emotional systems (FES). The cue-redundancy model
addresses the distribution of psychophysical and culture-
specic cues for an enculturated individual, and may be
understood as a cross-section of FES (dotted line).
27-Juslin & Sloboda-Ch-27.indd 768 7/24/2009 7:50:52 PM
27. 5 cross-cultural studies of emotion and music ,o,
Nonetheless, a basic premise of FES is that psychophysical cues continue to exert an
inuence on emotional communication throughout the lifespan, providing a means
for cross-cultural emotional communication and for commonalities between music
and speech prosody in their use of emotional cues.
27.5 Cross-cultural studies of
emotion and music
Cross-cultural research on emotion and music includes studies that directly assess the
capacity of individuals to interpret the emotional meaning of music across cultures,
as well as studies that provide indirect evidence for cross-cultural commonalities in
the association between music and emotion. The latter studies include developmen-
tal approaches and investigations of tension and complexity. These three sources of
evidence will be reviewed in turn.
27.5.1 Direct assessments
Gundlach (I,,:) conducted one of the early psychological studies of emotional com-
munication that considered music outside the Western classical tradition. His aim was
to determine whether or not there are objective characteristics of music in aboriginal
cultures that form the basis of communicating specic moods. He analysed archives
of songs collected from several North American aboriginal tribes. Songs were chosen
for their correspondence with important aspects of aboriginal life (healing songs, love
songs, and war songs) that were likely to express a specic emotion. Analyses of pitch,
range, tempo, and rhythm revealed that war songs were typically low in pitch, fast in
tempo, and wide in pitch range, whereas love songs were usually high in pitch, slow in
tempo, and had a moderate pitch range. Healing songs often had the fewest number of
rhythmic changes, whereas war songs had the greatest number of rhythmic changes.
Gundlach concluded that there were measurable objective differences between these
songs as a function of their emotional character.
In a later study, Gundlach (I,,,) presented a phrase from o different musical pieces
to groups of listeners with varying levels of musical training. Their task was to select
from a list of terms the verbal descriptor that best characterized the mood expressed
by the composer, or to ll in their own description if they could not nd an appropri-
ate term. Agreement among participants was observed for each phrase (mean agree-
ment o.,,). Analyses of tempo, rhythm, intervals, range, intensity, mean pitch, and
melody range revealed several objective features of phrases associated with emotional
communication. However, the extent to which such features can be decoded across
cultures was not determined.
27-Juslin & Sloboda-Ch-27.indd 769 7/24/2009 7:50:52 PM
,,o cross-cultural similarities and differences
Morey (I,o) extended this work by presenting samples of Western classical music
(Schubert, Davies, Handel, and Wagner) to II male members of the Loma, a tribe in
Liberia, West Africa. The pieces were selected on the basis that they conveyed emotions
such as fear, reverence, rage, and love. Participants were teachers and students at the
Holy Cross Mission School at Bolahun. They were asked to indicate which emotions
were expressed by each music selection. Morey observed that the music often did not
appear to elicit any emotionsor at least they did not evoke emotional judgements
that are typical of Western listeners. In another experiment, Morey presented his music
samples to :o members of the Zealua Loma tribe outside the school. He reported that
they did not nd the music interesting and many of them left while it was still playing.
With ubiquitous media and increasing globalization in the twenty-rst century,
Western tonal music is so pervasive that it is rapidly losing its value as music stimuli
in cross-cultural investigations. Research that involves Western tonal music therefore
requires participants from extremely isolated cultures who are nave to this music.
Because radio and television are almost universally available, it is reasonable to wonder
whether such cultures exist. Tom Fritz, together with Stefan Koelsch and colleagues,
performed a cross-cultural study with participants from a native African population
known as the Mafa (Fritz, Sammler, & Koelsch, :ooo; Fritz et al, in press). The Mafa
are one of approximately :,o ethnic groups that make up the population of Cameroun.
They are culturally isolated and live in the Extreme North Province in the Mandara
mountain range. The remote Mafa settlements do not even have electrical supply,
and are inhabited by individuals who pursue a traditional lifestyle and have not been
exposed to Western music. Twenty German and :I Mafa listeners judged the emotions
conveyed in a sample of Western music. The music excerpts, which varied in duration
(,I, s), mode, tempo, pitch range, tone density, and rhythmic regularity, were com-
posed to express the emotions happy, sad, and scary.
Music stimuli were classied using depictions of facial expressions from the Ekman
archive (happy, sad, scary) (Ekman, I,,o). Three Mafa participants could not recog-
nize the facial expressions on the two-dimensional paper and were thus excluded from
the test. Both German and Mafa listeners recognized all three emotions in the music
samples at levels that were above chance performance (,, per cent). Among German
listeners, recognition of happy music (,, per cent) was better than recognition of sad
music (,, per cent) or scary music (8I per cent). Similarly, among the Mafa listeners,
recognition of happy music (o, per cent) was better recognized than recognition of sad
music (, per cent) or scary music (8 per cent). This advantage for happy music did
not arise because the two negative emotions were confused with one another: errors
for each target emotion were distributed equally among the other two categories.
Interestingly, the capacity to decode emotional expressions was signicantly correlated
with the degree of appreciation of the compositions. The authors proposed that mean-
ingfulness in music (clear emotional expression) leads to increased appreciation.
Hindustani music offers an ideal stimulus for cross-cultural comparisons, because
of its strong association with emotional communication, and because it is easy to locate
individuals who are unfamiliar with this music. Central to Hindustani music is the
raga-rasa system in which each raga (set piece) is associated with one or more rasas
27-Juslin & Sloboda-Ch-27.indd 770 7/24/2009 7:50:52 PM
27. 5 cross-cultural studies of emotion and music ,,I
(moods or essences). Each raga is also associated with a distinctive colour, season, and
time of day. Deva and Virmani (I,,,) compared the intended mood of Hindustani
ragas with mood ratings made by Indian listeners. In one experiment, four excerpts
from ragas (roughly two minutes each) were played to ,, listeners, who were asked to
choose from a list of mood adjectives the emotion that they believed was expressed. The
authors reported that the perceptions of listeners often matched the intended mood of
the ragas. In a second experiment, ::8 listeners were asked to rate excerpts of a raga on
expressed mood, associated colour, season, and time of day. Listeners judged the raga
(Bhairav) as dominant in empathy and compassion, and as reecting vitality, positive
affect, courage, and tranquillity. The majority of listeners associated the raga with the
colour white or yellow, the season of autumn, and the time of day as early morning or
early evening. According to Deva and Virmani, ratings generally approximated the
intended mood of the raga.
In a later study, Gregory and Varney (I,,o) examined whether Western listeners
who are unfamiliar with Hindustani music can interpret the emotional connotations
of that music. Gregory and Varney asked British residents of Western and Indian heri-
tage to judge the emotional messages implied in commercially recorded excerpts of
Hindustani ragas, Western classical music, and Western new age music. After hear-
ing each excerpt, listeners selected adjectives from a list developed by Hevner (I,,o;
see Chapter I, this volume). Western and Indian listeners were sensitive to intended
emotions in Western music, but not Hindustani ragas. The nding suggests that emo-
tional cues in unfamiliar music are not always interpretable, and suggests that culture-
specic properties of music can mask culture-transcendent emotional cues.
Balkwill and Thompson (I,,,) performed a similar study, but observed far better
emotional decoding of Hindustani music by Western listeners. They asked Canadian
listeners to judge the emotional content of eld recordings of Hindustani ragas, and
to rate the presence of several psychophysical attributes in the music. Recordings were
obtained of solo performances of Hindustani ragas that were explicitly intended to
evoke specic emotions. Excerpts intended to convey joy/hasya were assigned high
ratings of joy; excerpts intended to convey sadness/karuna were assigned high ratings
of sadness; and excerpts intended to convey anger/raudra were assigned high ratings of
anger. High ratings of joy were associated with high ratings of tempo and low ratings of
melodic complexity. High ratings of sadness were associated with low ratings of tempo
and high ratings of melodic complexity. Anger ratings were not signicantly associated
with ratings of tempo or complexity, but ragas performed on stringed instruments
were rated much higher in anger than those performed on the ute.
A subsequent study conducted in Japan extended this work and provided addi-
tional evidence for the cue-redundancy model. Whereas Balkwill and Thompson
(I,,,) examined judgements by Canadian listeners of Hindustani music, Balkwill,
Thompson, and Matsunaga (:oo) examined judgements by Japanese listeners of
Japanese, Western, and Hindustani music. Music samples consisted of emotive music
by two Canadian musicians in Toronto, Ontario, Canada, and six Japanese musicians
in Sapporo, Hokkaido, Japan. The music provided by the Canadian musicians was a
compilation of improvisations on each target emotion. Each musician was asked to
27-Juslin & Sloboda-Ch-27.indd 771 7/24/2009 7:50:52 PM
,,: cross-cultural similarities and differences
draw upon their musical experience and repertoire and to focus on the task of con-
veying each of the three specied emotions. This procedure was adopted to minimize
extra-musical associations that many listeners may have formed between past pairings
of familiar music with their own experiences. Because Japanese listeners are generally
well acquainted with classical and popular Western music, the use of improvisations
also served to decrease the level of familiarity with the Western music materials for this
listener group.
The music provided by the Japanese musicians was a compilation of several genres
of traditional composed music (gagaku, shinto and minyo). Each musician was asked
to choose pieces from their repertoire that they felt would best convey each target emo-
tion. These two stimulus sets, along with the Hindustani music previously collected
(from Balkwill & Thompson, I,,,) were then edited to create short excerpts (mean
duration = ,o s) intended to evoke anger, joy, or sadness.
One hundred and forty-seven Japanese listeners (,o women, ,I men, mean age =
:,.,) rated the degree of anger, joy, sadness, complexity, loudness, and tempo in each
sample of music from Western, Japanese, and Hindustani stimulus sets. As expected,
Japanese listeners were sensitive to the intended emotion of music from Japanese,
Western, and Hindustani music, and their judgements were associated with the pres-
ence of psychophysical cues. High ratings of joy were associated with a fast tempo
and a simple melody; high ratings of sadness were associated with a slow tempo and
a complex melody; and high ratings of anger were associated with high intensity and
a complex melody. The ndings conrmed that Japanese listeners were sensitive to
the emotions communicated in familiar and unfamiliar music, and their sensitivity
was associated with the perception of psychophysical cues that transcended cultural
boundaries.
More recently, we compared the judgements of Canadian and Japanese listeners for
the same music stimuli (Balkwill, :ooo). One hundred and thirty-nine Western listen-
ers (8: women, ,, men, mean age = :,.,:) participated in the study. As with Japanese
listeners, the Canadian group was remarkably sensitive to intended emotions expressed
in Japanese, Western, and Hindustani music, and their judgements were associated
with the presence of psychophysical cues. Both groups were able to recognize music
intended to evoke anger, joy, and sadness in all three tonal systems, and judgements
for every emotion were signicantly associated with at least one of the psychophysical
attributes of complexity, intensity, tempo, and timbre.
An interesting difference observed between the Canadian and Japanese listener
groups was the number of psychophysical cues associated with some emotion judge-
ments. For example, in the case of anger ratings the degree of perceived complexity,
tempo, and intensity were all signicant predictors for Japanese listeners, while for
Canadian listeners, only degree of perceived intensity was signicant. Previous research
involving descriptions of visual stimuli have indicated that Japanese participants tend
to process scenes more holistically, reporting more background stimuli than North
American participants, who tend to focus on primary focal elements (Masuda &
Nisbett, :ooI). Whether similar differences between Japanese and North American
groups in the quantity and quality of psychophysical cues associated with emotion
27-Juslin & Sloboda-Ch-27.indd 772 7/24/2009 7:50:52 PM
27. 5 cross-cultural studies of emotion and music ,,,
recognition in this study can be attributed to cultural differences in attention focus
or cognitive style is an intriguing question (see Nisbett, Peng, Choi, & Norenzayan,
:ooI, for a review). Further experiments with music and other auditory stimuli more
clearly focused on this question may yield some interesting contributions to this area
of research.
27.5.2 Developmental evidence
There is little doubt that long-term exposure to the music of ones culture has a pow-
erful effect on our perceptions and experiences of music (Hannon & Trainor, :oo,).
Memory for music is better as a function of increasing age (Krumhansl & Keil, I,8:;
Trainor & Trehub, I,,) and of familiarity or conventionality of the music itself
(Cuddy, Cohen, & Mewhort, I,8I). Music processing is also enhanced by formal music
training (Lynch & Eilers, I,,I).
However, there is also considerable evidence that basic auditory skills are well devel-
oped at birth, and such skills are similar in all humans across cultures. Infants possess
a number of perceptual skills for music that are remarkably similar to those of listen-
ers who have had years of exposure to music. These apparent predispositions exert a
signicant inuence on music experiences throughout the lifespan, including emo-
tional responses to music and speech (Trehub, :ooo, :oo,; Trehub & Nakata, :oo:).
As examples, infants show evidence of processing advantages for contour patterns
over precise intervals, for consonant over dissonant sounds, and for temporal pattern
relations over absolute durations (Trehub & Hannon, :ooo). Because these skills are
observed very early in development, it is possible that their emergence reects innate
biases rather than environmental stimulation. In so far as such early skills do not rely
heavily on processes of enculturation, they may form the basis for many important
cross-cultural similarities in the perception of music.
It should be emphasized that it is not possible to rule out explanations of early audi-
tory skills based on rapid adaptation of the auditory system to music exposure. To
illustrate, the AI tonotopic frequency map is fully developed (i.e. equivalent to adult
organization) in rats two days after they are born (de Villers-Sidani, Chang, Bao, &
Merzenich, :oo,). If newborn rats are exposed to broadband noise for up to ,o days, no
tonotopic map is evident. But within two days of turning off the noise, the AI tonotopic
map will be fully formed. Two conclusions may be drawn from these observations:
(I) exposure to pitched tones is necessary for tonotopic perceptions to emerge, and
(:) the rat learns pitch with only two days of auditory exposure. Clearly, the effects of
learning can be unexpectedly rapid, such that one might be forgiven for assuming that
the musical dispositions of newborns are innately determined.
An important inuence on the emotional connotation of music is the relative balance
of consonance and dissonance. In Western tonal music, consonance usually signals
stability and positive emotional valence, whereas dissonance signies tension, insta-
bility, and negative emotional valence. Musicologists have suggested that dissonance
is understood and heard somewhat differently across different musical traditions,
27-Juslin & Sloboda-Ch-27.indd 773 7/24/2009 7:50:52 PM
,, cross-cultural similarities and differences
cultures, styles, and time periods, but evidence suggests that infants begin life with
predispositions for processing consonance and dissonance, and these predispositions
may remain inuential into adulthood. Specically, infants are highly sensitive to the
distinction between consonance and dissonances, and even newborn infants prefer
music that is consonant to music that is dissonant.
Zentner and Kagan (I,,o) exposed -month-old infants to consonant and disso-
nant versions of two melodies. Infants looked longer at the source of sound and moved
less when consonant versions of melodies were played than when dissonant versions
of melodies were played. Moreover, when the dissonant version was played, infants
were more likely to fret and turn away from the music source. The researchers argued
that infants are biologically predisposed to prefer consonance to dissonance (Crowder,
Reznick, & Rosenkrantz, I,,I; Trainor & Heinmiller, I,,8).
Infant-directed singing (lullaby) is prevalent across cultures (Trehub, Unyk, &
Trainor, I,,,; Unyk, Trehub, Trainor, & Schellenberg, I,,:), and plays many impor-
tant functions in infant development, such as the acquisition of attentional skills, the
maintenance of emotional stability, and the enhancement of infantcaregiver bonds
(see also Chapter :,, this volume). These functions have implications for the survival
of the infant, and are thought to have a biological basis. A number of researchers have
proposed that all music may have ultimately originated from this form of musical
activity, which conferred a survival advantage for ancestral populations. Unlike other
primates, infants cannot cling to their mothers fur while the mother forages with both
hands. Singing would have allowed mothers to soothe their infants (keeping them quiet
to avoid predators) while foraging. If such conjectures are correct, then the diverse
musical systems across all cultures should have a common foundation that is reected
in infant-directed songs. Moreover, cross-cultural similarities should be particularly
evident for the emotional qualities of music.
In support of this view, Trehub and colleagues demonstrated that infants are natu-
rally attracted to lullabies. Nakata and Trehub (:oo) found that o-month-old infants
look longer at audio-visual recordings of their mother if she is singing a lullaby than
if she is speaking, suggesting that lullabies generate positive affective states in infants.
Moreover, cortisol levels (associated with stress) dropped signicantly following
infant-directed song, indicating that lullabies are naturally soothing.
Infant-directed song contains many of the emotional cues that are observed in
infant-directed speech, including relatively high pitch, slow tempo, and simple and
repeating contours (Fernald, I,,I; Fernald & Simon, I,8). Is there a general con-
nection between the emotional cues used in music and speech prosody? Evidence by
Thompson, Schellenberg, and Husain (:oo) suggests that there is. Musically trained
and untrained adults were given a test of sensitivity to emotional prosody for phrases
spoken in their own language (English) or in a foreign language (Tagalog, a language of
the Philippines). The test required listeners to classify the emotional connotation of each
utterance as happy, sad, fearful, or angry. Musically trained adults were signicantly
better than untrained adults at classifying the emotions, including those expressed in
the unfamiliar language of Tagalog. That is, cross-cultural sensitivity to vocal emotions
was enhanced for musically trained adults. The results suggest that psychophysical
27-Juslin & Sloboda-Ch-27.indd 774 7/24/2009 7:50:52 PM
27. 5 cross-cultural studies of emotion and music ,,,
cues to emotion are instantiated in both music and speech prosody. Thus, long-term
training in music gives rise to enhanced sensitivity to such cues, whether they occur in
familiar music, unfamiliar music, familiar speech, or unfamiliar speech.
To address the possibility that the trained adults were naturally more sensitive to
vocal emotions irrespective of their training, the researchers assessed the sensitivity
of o-year-olds to emotional prosody after one year of music lessons, drama lessons,
or no lessons. The results revealed that both drama and keyboard lessons resulted in
enhanced sensitivity to emotional prosody, including prosodic materials from an unfa-
miliar language (Tagalog). Thus, drama lessons, with their emphasis on learning to
express different emotions, had a direct effect on learning, whereas music lessons had
a transfer effect to the domain of speech prosody. One explanation for these ndings
is that music and speech prosody implicate the same acoustic cues to emotion. Music
training may not only enhance sensitivity to culture-specic properties of a musical
system; it may also engage and rene neural processes that respond to universal acous-
tic cues to emotion, thereby enhancing sensitivity to emotions in music and speech
across cultures.
27.5.3 Tension and stability
Musical tension provides another source of emotional meaning in music. Tension can
arise from several sources, including sensory dissonance, violations of expectations,
complexity, and tonal instability. When we listen to music from an unfamiliar genre,
our levels of uncertainty are generally higher, giving rise to perceived complexity and
tension. Huron (:ooo) used a betting paradigm to evaluate the level of uncertainty of
continuations in a melody. Balinese and American musicians listened to an unfamiliar
traditional Balinese melody and provided judgements, via a betting game, of their abil-
ity to anticipate future pitches. Not surprisingly, uncertainty was greater for American
listeners than for Balinese listeners, but the levels of uncertainty also showed a differ-
ent trajectory for the two groups. For Balinese listeners, the level of uncertainty was
maximal in the middle of the melody, and minimal near the beginning and end of the
melody. No such rising and falling pattern was observed for American listeners, who
simply showed a subtle decline in uncertainty as the melody progressed.
In a series of studies of the implication-realization (I-R) model (Narmour, I,,o),
American and Chinese listeners were asked to provide ratings of continuation fol-
lowing fragments of British folk songs, Webern lieder, and Chinese pentatonic songs
(Krumhansl, I,,,; Schellenberg, I,,o). Certain predictors based on the I-R model
accounted for a signicant proportion of the variance in ratings for the two listener
groups and three musical styles, suggesting cross-cultural similarities in the formation
of expectancies. For example, listeners tended to expect continuation tones that were
close in pitch to preceding tones (see also Carlsen, I,8I; Thompson & Stainton, I,,8). If
such expectancies are hard-wired, then they could provide a basis for emotional com-
munication across cultures, whereby large changes in pitch in a melody universally
give rise to heightened arousal (i.e. surprise) and increased tension.
27-Juslin & Sloboda-Ch-27.indd 775 7/24/2009 7:50:52 PM
,,o cross-cultural similarities and differences
Perceived tension in music can also be inuenced by the overall complexity of the
music. Cross-cultural comparisons suggest that a range of factors inuence perceived
complexity, and that these factors are weighted differently for judgements of famil-
iar and unfamiliar music. Eerola, Himberg, Toiviainen, and Louhivuori (:ooo) asked
Finnish and African listeners to judge the complex of Western and African folk songs.
A range of melodic complexity measures was developed, including average interval
size, entropy of interval distribution, entropy of note duration, rhythmic variability,
note density, tonal ambiguity, and contour entropy (where entropy is a measure of
uncertainty).
Overall patterns of judgements were similar for the two groups and the two styles
of music, supporting the idea that complexity is a culture-transcendent construct.
However, culture-specic inuences on judgements of complexity were also evident.
Whereas the style of the Western folk songs was familiar to both groups, the style of
the African folk songs was familiar to only the African group. Complexity measures
reected these differences in familiarity, with differences in the predictive power
of complexity variables differing between groups for the African folk songs. Taken
together, the results suggest that melodic complexity transcends cultural boundaries,
but musical enculturation still plays an important role.
Tension is also inversely related to tonal stability (Lerdahl & Krumhansl, :oo,).
Cross-cultural studies of tonal stability suggest that in the absence of knowledge of a
tonal system, the relative stability of tones (tonal hierarchies) formed by non-native
listeners were inuenced mainly by pitch distribution. When Western listeners pro-
vide goodness-of-t judgements for probe tones presented in the context of North
Indian music (Castellano, Bharucha, & Krumhansl, I,8) and Balinese music (Kessler,
Hansen, & Shepard, I,8), their judgements are primarily inuenced by the frequency
of occurrence of pitches. Similarly, when Western listeners provide judgements
of North Sami yoiks the inuence of Western schematic knowledge is strongly evi-
dent; when indiginous Sami listeners judge the same materials there is relatively little
evidence for an inuence of Western schematic knowledge (Krumhansl et al, :ooo).
27.6 Cross-cultural studies of
emotion and prosody
Because emotion is expressed in music and speech prosody using very similar cues,
cross-cultural studies of speech prosody are highly relevant to cross-cultural investi-
gations of music and emotion (Ilie & Thompson, :ooo; Juslin & Laukka, :oo,; Patel,
:oo8). Indeed, tempo, intensity, and timbre affect judgements of emotion in both
music and speech. For example, vocal expressions of anger are often marked by fast
tempo and greater intensity (in the case of hot anger); vocal expressions of joy are
often fast in tempo with a higher pitch range; and vocal expressions of sadness are
27-Juslin & Sloboda-Ch-27.indd 776 7/24/2009 7:50:52 PM
27. 6 cross-cultural studies of emotion and prosody ,,,
typically slow in tempo and low in pitch (Bachorowski & Owren, I,,,; Frick, I,8,;
Scherer, I,8o; Scherer & Oshinsky, I,,,; Thompson & Balkwill, :ooo).
Several investigations of emotional prosody have converged on the view that listen-
ers are highly sensitive to the prosodic expression of anger, joy, and sadness (Juslin
& Laukka, :oo,). Conversely, a relatively low success rate has been reported for the
recognition of fear and disgust in speech. Emotional sensitivity for prosodic materials
is similar to emotional sensitivity for musical materials, although the two modalities
differ in the emotions that are most commonly conveyed. Anger is commonly commu-
nicated through speech prosody, and it is often the best recognized emotion for spoken
stimuli (e.g. Banse & Scherer, I,,o). Interestingly, anger is also recognized rather well
in music (Juslin & Laukka, :oo,), even though this emotion is communicated in only
certain genres of music, and composers have difculty expressing anger with pitch or
rhythmic structure (Thompson & Robitaille, I,,:).
Although the CRM was developed to account for cross-cultural studies of music and
emotion, it readily applies to emotional prosody, suggesting that a broader framework
is needed to account for the emergence of emotional systems. The concept of fraction-
ating emotional systems assumes that psychophysical cues provide a common point
of departure for the development of emotional communication in music and speech
prosody. According to FES, individuals should be able to recognize emotions expressed
prosodically even when utterances are spoken in an unfamiliar language. Moreover, an
ethnic bias should be observed for decoding emotional messages expressed prosodi-
cally, because of the presence of culture-specic cues.
Prosodic cues for emotion are well documented for English and some Western
European languages (German, Dutch). Joy is expressed with a comparatively rapid
speaking rate, high average pitch, large pitch range, and bright timbre; sadness is
conveyed with a slow speaking rate, low average pitch, narrow pitch range, and low
intensity; anger is expressed with a fast speaking rate, high average pitch, wide pitch
range, high intensity, and rising pitch contours; and fear is conveyed with a fast speak-
ing rate, high average pitch, large pitch variability, and varied loudness (Juslin &
Laukka, :oo,; Scherer, I,8o).
Analyses of recognition rates indicate that prosodic cues alone allow listeners to
identify the emotion being conveyed, with identication rates of roughly four to ve
times the rates that would be predicted by guessing (Banse & Scherer, I,,o; Frick,
I,8,; Scherer, I,,,, I,8o; Standke, I,,:; van Bezooijen et al, I,8,). Not all emotions are
decoded equally well from speech prosody, however. For example, anger and sadness
are typically decoded more reliably than joy and fear (Banse & Scherer I,,o; Johnstone
& Scherer :ooo).
A large number of studies have revealed sensitivity to emotional prosody across
cultures and in unfamiliar languages. Kramer (I,o) demonstrated that English-
speaking American listeners could identify vocal expressions of emotion rendered in
English and Japanese (content-ltered) at a rate better than chance. Beier and Zautra
(I,,:) found that American, Polish, and Japanese listeners were able to identify emo-
tion in American English speech. Similar positive results were reported by McCluskey
and Albas (I,8I) for their study of Canadian and Mexican children and adult listeners
27-Juslin & Sloboda-Ch-27.indd 777 7/24/2009 7:50:52 PM
,,8 cross-cultural similarities and differences
judging vocally expressed emotion provided by Canadian and Mexican actors. In some
cases, listeners were more accurate in recognizing the emotion in the unfamiliar lan-
guage than in their own. In an experiment with samples of emotive speech spoken in
Dutch, judges from the Netherlands, Taiwan, and Japan were also able to recognize
each intended emotion with better-than-chance accuracy (van Bezooijen et al, I,8,),
although there were differences in level of accuracy as a function of emotion and lan-
guage. More recently, Scherer et al (:ooI) recorded German actors (two male, two
female) portraying ve emotions (joy, anger, sadness, fear, and neutral). The voice
samples were elicited using a scenario approach in which the actors were given text
descriptions of emotive scenarios and asked to act them out (e.g. death of a loved one)
in the performance of two standard sentences. These sentences were designed to cir-
cumvent the issue of semantic meaning by combining phonemes from six different
languages (German, English, French, Italian, Spanish, and Danish; e.g. Fee gott laish
jonkill gosterr).
By enlisting collaborators in nine countries (Germany, Switzerland, Great Britain,
the Netherlands, Italy, France, Spain, the United States, and Indonesia), emotion
judgements from groups of participants in each country were obtained (n ranged from
,: to ,o, total = :8; age I8,o). All materials were translated into the language of each
country. Participants were asked to focus on the emotion rather than the content of the
sentence. After each utterance, they were given six seconds to select up to two emotion
labels on the list of ve target emotions. They were urged to select ones they thought
best described the emotion the actor was trying to convey.
The authors reported an accuracy of oo per cent across all emotions and all coun-
tries. Rate of accuracy was affected by country, type of emotion, and gender of the
speakers. Confusion matrices were generated to assess patterns of errors as well as
accuracy. German listeners had the highest rates of accuracy (Cohens Kappa = .o,,
p < .ooI). Listeners from the other European countries also did well, with accuracy
ranging from .,: to .o: (Cohens Kappa, p < .ooI). Listeners from the one non-Euro-
pean country, Indonesia, had the lowest accuracy rate (Cohens Kappa = .,,, p < .ooI).
The highest recognition rate was associated with expressions of anger; the lowest rec-
ognition rate was associated with expressions of joy.
Elfenbein and Ambady (:oo:) conducted a meta-analysis of ,, experiments explor-
ing cross-cultural recognition of emotion in visual and auditory modes. In each study
analysed, emotional stimuli (speech, facial expression) from at least one culture were
presented to members and non-members of that culture. Cross-cultural emotion rec-
ognition accuracy was lower in studies of prosody than in studies of facial expression
and body language, although most studies of prosody reported better-than-chance
rates of recognition. Very few cross-cultural studies of emotional prosody have
involved non-Western languages.
Examination of non-Western languages is important, because there is a long history
of close contact between speakers of Western languages that might explain similar uses
of prosody. Similar uses of prosody in Swedish and Norwegian, for example, can be
explained by the history of interaction between speakers of these languages, and would
not implicate universal principles of emotional prosody. Similar uses of prosody in
27-Juslin & Sloboda-Ch-27.indd 778 7/24/2009 7:50:52 PM
27. 6 cross-cultural studies of emotion and prosody ,,,
Western and non-Western languages, on the other hand, would provide compelling
evidence for universal principles of emotional prosody.
Elfenbein and Ambady (:oo,) proposed a cultural proximity hypothesis to predict
how well people of different cultures recognize emotional expression. According to
their hypothesis, members of cultures who share cultural elements such as degree of
individualism or collectivism, power structure, and gender roles, should be more suc-
cessful at decoding each others emotional expressions than members of cultures that
are less similar. The cultural proximity hypothesis predicts, for example, that Japanese
people should be better at recognizing the emotional expressions of Chinese people
than the emotional expressions of North Americans, because Japanese and Chinese
cultures are more similar to each other on relevant dimensions than Japanese and
American cultures. Conversely, it predicts that English-speaking listeners should nd it
difcult to decode emotional prosody in Japanese and Chinese speech. This prediction
contrasts with that of the cue-redundancy model, which suggests that psychophysi-
cal cues associated with emotions will allow individuals to decode emotions across
cultural boundaries, including individualist and collectivist cultural boundaries.
Thompson and Balkwill (:ooo) investigated how well certain prosodic cues
(frequency, intensity, and event density) were associated with the ability of English
speakers to recognize emotion in ve languages: English, German, Mandarin Chinese,
Tagalog, and Japanese. Recordings were obtained of semantically neutral sentences (e.g.
The bottle is on the table) spoken in a way that communicated each of four intended
emotions: joy, sadness, anger, and fear. The use of prosody was not exaggerated or dra-
matic, but merely typical for speakers of each language. The ve languages represent
two members of the Germanic branch of the Indo-European language family (English
and German), one language from the Sino-Tibetan family (Mandarin-Chinese), one
language from the Altaic family (Japanese), and one from the Austronesian family
(Tagalog) (Katsiavriades & Qureshi, :oo,). German and English are Western languages
spoken in individualistic societies, whereas Japanese and Chinese are Asian languages
spoken in collectivist societies. Tagalog has been inuenced by both the colonizing
languages of Spanish and English, as well as by the languages of its Asian neighbours,
Japan, Korea, and China.
Three predictions were evaluated. First, assuming that sensitivity to emotional pros-
ody is partially dependent on psychophysical cues, listeners should be able to decode
emotional prosody in any language at rates higher than that predicted by chance.
Second, based on the idea that certain prosodic cues of emotion are culture specic,
listeners should have higher rates of recognition for emotional prosody in their own
language. Third, in view of previous research on emotional prosody, we predicted that
listeners should be able to decode anger and sadness more reliably than joy and fear in
all ve languages (Johnstone & Scherer :ooo).
English-speaking listeners recognized the four emotions at a rate signicantly
better than chance in all ve languages, achieving the highest recognition scores within
their own language. The analysis of speech stimuli revealed several psychophysical cues
associated with specic emotions, providing potential cues to the decoding of emo-
tional meaning. Regression analyses conrmed the predictive power of these cues for
27-Juslin & Sloboda-Ch-27.indd 779 7/24/2009 7:50:52 PM
,8o cross-cultural similarities and differences
emotive judgements. Across languages, the ability to decode emotions was associated
with psychophysical cues such as mean frequency, intensity, and event density.
We also assessed the ability of adult, native speakers of Mandarin Chinese and
Japanese to identify intended emotions in English, Chinese, and Japanese speech sam-
ples (Balkwill & Thompson, :oo,; Balkwill, :ooo). Emotional decoding from these
listeners was compared with the results from our English-speaking listeners. Again,
semantically neutral sentences were used so that emotional intentions could be deter-
mined only by attending to prosodic aspects of speech.
Chinese and Japanese groups recognized anger, joy, and sadness at rates that were
higher than chance performance in all ve languages. Across languages and listener
groups, the ability to decode emotions was associated with acoustic qualities such as
mean frequency, intensity, and event density. Although both groups recognized fear
above chance in English and Chinese, Japanese listeners recognition of fear fell below
chance in their own language. More data are needed to draw strong conclusions from
such ndings, but it is interesting to note that research on facial expressions indicates
that Japanese judges also have difculty decoding fear from Japanese facial expressions
(Biehl et al, I,,,; Matsumoto, I,,:, I,,o; Russell, Suzuki, & Ishida, I,,,). It has been
suggested that the emphasis in Japan on group cohesion and harmony discourages
acknowledgement of emotions that threaten group harmony (Matsumoto, I,,o).
We were able to address the cultural proximity hypothesis by comparing recognition
rates for English, Chinese, and Japanese listeners. Support for the hypothesis would
be evident if: (a) all listener groups are better at recognizing emotion in their own
language than in other languages (even when verbal cues provide no information);
(b) Chinese listeners are better than English-speaking listeners at recognizing emotion
in Japanese speech; and (c) Japanese listeners are better than English listeners at recog-
nizing emotion in Chinese speech. China and Japan have a long history of interaction
between their peoples in times of peace and conict (Rose, :oo,). Emotional display
rules in China and Japan are also similar in that negative emotions are considered inap-
propriate to display in public. North Americans, in contrast, have a higher tolerance of
negative emotional displays (Eid & Deiner, :ooI; Matsumoto, I,,o).
Only our English-speaking listeners showed an in-group advantage across emotion
categories. Chinese and Japanese listeners had an in-group advantage for the recogni-
tion of joy in their own language, but on the whole did not decode emotion better in
their own language than in the English language. The high recognition of emotion in
English speech by Chinese and Japanese listeners was not surprising, as all participants
had exposure to English speech during their secondary school English-language train-
ing, as well as ongoing exposure through television, movies, and other media (Albas et
al, I,,o; Rosenthal et al, I,,,; Scherer et al, :ooI). That is, passive exposure to English
speech among Japanese and Chinese listeners may have overwhelmed any decoding
advantage arising from cultural proximity.
Across languages, judgements of joy were associated with greater range of
fundamental frequency; judgements of sadness were associated with lower mean
intensity, lower range of intensity, and lower event density; and judgements of anger
were associated with greater mean and range of intensity, and greater event density.
27-Juslin & Sloboda-Ch-27.indd 780 7/24/2009 7:50:52 PM
AU: Bill and Laura-Lee
pls check at proof stage
whether this should be
three or five.
27. 7 summary and future prospects ,8I
Consistent with the cue-redundancy model, the type of acoustic parameter and the
direction of association with emotion judgements were often similar in Canadian,
Japanese, and Chinese listeners.
27.7 Summary and future prospects
Cross-cultural studies of music and speech prosody suggest that emotional communi-
cation occurs through a complex interplay between system-specic and psychophysi-
cal cues. Psychophysical cues have similar connotations across cultures (Balkwill &
Thompson, I,,,; Balkwill, Thompson, & Matsunago, :oo) and similar functions in
music and speech prosody (Ilie & Thompson, :ooo; Juslin & Laukka, :oo,; Patel, :oo8).
Reliance on psychophysical inuences wanes throughout the lifespan, with encultura-
tion gradually leading to increased sensitivity to culture-specic and channel-specic
cues. Enculturation can lead to an understanding of abstract cues that have no psy-
chophysical basis, but it can also inuence the interpretation of psychophysical cues
themselves, overlaying their signicance with culture-specic nuance or altering their
interpretation entirely.
To summarize, the FES model contains three components, as follows:
I. Phylogenetic base. Infants start life with an understanding of basic psychophysical
cues to emotion. Psychophysical cues are determined by system constraints that
may include genetic factors, but may also be determined by ubiquitous environ-
mental and biophysical constraints. Musical and prosodic channels of communi-
cation share these psychophysical cues.
:. Ontogenetic process. Throughout development is a process of renement and
fractionation of different emotional communication systems. This fractionation
applies not only to the modalities of music and prosody (which fractionate from
each other), but also to musical and prosodic systems across cultures. That is, the
process of fractionation permits the emergence of culture-specic systems of emo-
tional coding, some prosodic and others musical. System-specic cues are learned
through exposure to the conventions of ones culture, giving rise to culture-
specic understandings of emotional communication.
,. Cross-cultural communication. Psychophysical cues remain available for decod-
ing at all stages of development, with culture-specic and domain-specic cues
overlaid increasingly throughout development. When an individual attempts to
decode an emotional message (from prosody or music) across cultural boundar-
ies, it is usually optimal to focus attention on psychophysical cues while ignor-
ing culture-specic cues. Culture-specic cues can mask the ability to attend to
psychophysical cues, and emotional communication is weaker across cultures
than within cultures because psychophysical cues provide only one part of the
27-Juslin & Sloboda-Ch-27.indd 781 7/24/2009 7:50:52 PM
,8: cross-cultural similarities and differences
full emotional code. Nonetheless, it generally enables above-chance decoding of
emotional messages.
Research is needed to evaluate the capacity for enculturation to alter the emotional
connotation of psychophysical cues entirely. Existing data suggest that this capacity is
either limited or seldom realized: individuals at all stages of development are capable
of decoding basic emotions such as anger, joy, and sadness in unfamiliar music or
speech, because both channels of emotional communication contain a number of psy-
chophysical cues that transcend cultural boundaries, including pitch level and vari-
ability, sound intensity, rate (tempo), timbre, and complexity. Difculties in decoding
do occur, but they usually arise because culture-specic conventions and cues mask
the detection of psychophysical cues. Training listeners to selectively attend to psy-
chophysical cues may lead to improved cross-cultural emotional communication in
music and speech prosody.
Further cross-cultural studies are also needed to classify the many known emotional
cues as either psychophysical and having a phylogenetic basis, or culture-specic and
having an ontogenetic basis. Juslin and Laukka (:oo,) have identied a large number
of musical and prosodic cues such as jitter, vibrato, formant structure, the presence
of pauses, onset characteristics, and articulation. More recently, Juslin and Vstfjll
(:oo8) proposed a set of mechanisms for understanding emotional induction by music
(see Chapter ::, this volume). One such mechanismthe brain stem responseis
likely to respond in similar ways across all cultures to psychophysical features such
as high intensity. Nonetheless, the potential role of this mechanism for cross-cultural
emotional communication has yet to be fully understood.
How might one account for the existence of culture-transcendent connections
between psychophysical cues and emotions? Huron (:ooo) provided a number of evo-
lutionary explanations, and Scherers (I,8o) component process model describes physi-
ological changes that occur in the vocal apparatus during the expression of emotions.
For example, speaking of something that is deeply unpleasant often manifests as faucal
and pharyngeal constriction and a tensing of the vocal tract. The acoustic outcome is
higher-frequency energy. Based on this model, Banse and Scherer (I,,o) predicted
several associations between the acoustic attributes of speech and emotional intent.
Differences between expression and recognition of emotion in music and speech
are also of interest. Are some emotions easy to convey through speech prosody, but
difcult to convey in music? One might speculate that because speech carries the
weight of semantic as well as emotional meaning, it may be more susceptible to within-
culture norms as well as in-group and out-group biases. The FES framework accounts
for the dynamic interplay between psychophysical cues and culture-specic conven-
tions throughout development, and complements the CRM and other models of emo-
tional communication. For example, the Brunswikian lens model (Brunswik, I,,o)
provides another valuable framework for understanding the ability to decode emo-
tional meaning from either speech prosody (Scherer I,8:) or music (Juslin, :ooo).
Originally designed to describe visual perception, the model has been adapted to many
types of human judgements (see also Chapter I,, this volume).
27-Juslin & Sloboda-Ch-27.indd 782 7/24/2009 7:50:52 PM
references ,8,
In a Brunswikian framework the intent of the encoder to express an emotion is facili-
tated by the use of a large set of cues that are probabilistic and partially redundant.
Each cue by itself is not a reliable indicator of the expressed emotion, but combined
with other cues, each one contributes in a cumulative fashion to the communication
and recognition of emotion. The model incorporates the exibility of the decoder to
shift expectations from unavailable to available cues (Juslin, :ooo). The CRM and FES
are consistent with this model, but focus on the signicance of psychophysical cues for
the expression of emotions, their interaction with culture-specic convention that can
enhance or hinder the ability of listeners to decode emotions across cultures, and the
gradual fractionation of emotional communication systems across cognitive domains
and cultural environments.
Recommended further reading
I. Elfenbein, H. A., & Ambady, N. (:oo:). On the universality and cultural specicity of
emotion recognition: A meta-analysis. Psychological Bulletin, I:8, :o,,,.
:. Juslin, P. N., & Vstfjll, D. (:oo8a). Emotional responses to music: The need to consider
underlying mechanisms. Behavioral and Brain Sciences, ,I, ,,,,,.
,. Reyes, A. (:oo,). What do ethnomusicologists do? An old question for a new century.
Ethnomusicology, ,,, II,.
References
Albas, D. C., McCluskey, K. W., & Albas, C. A. (I,,o). Perception of the emotional content
of speech: A comparison of two Canadian groups. Journal of Cross-Cultural Psychology ,,
8I,o.
Bachorowski, J., & Owren, M. (I,,,). Vocal expression of emotion: Acoustic properties of
speech are associated with emotional intensity and context. Psychological Science, o,
:I,:.
Balkwill, L-L. (:ooo). Perceptions of emotion in music across cultures. Paper presented at
Emotional Geographies: The Second International & Interdisciplinary Conference, May
:ooo, Queens University, Kingston, Canada.
Balkwill, L-L., & Thompson, W. F. (I,,,). A cross-cultural investigation of the perception of
emotion in music: Psychophysical and cultural cues. Music Perception, I,, ,o.
Balkwill, L-L., & Thompson, W. F. (:oo,). Recognition of emotion in ve languages. Paper pre-
sented at Auditory Perception Cognition and Action Meeting, November :oo,, Toronto,
Canada.
Balkwill, L-L., Thompson, W. F., & Matsunaga, R. (:oo). Recognition of emotion in Japanese,
North Indian and Western music by Japanese listeners. Japanese Journal of Psychological
Research, o, ,,,,.
Banse, R., & Scherer, K. R. (I,,o). Acoustic proles in vocal emotion expression. Journal of
Personality and Social Psychology, ,o, oI,o.
27-Juslin & Sloboda-Ch-27.indd 783 7/24/2009 7:50:52 PM
,8 cross-cultural similarities and differences
Biehl, M., Matsumoto, D., Ekman, P., & Hearn, V. (I,,,). Matsumoto and Ekmans Japanese
and Caucasian Facial Expressions of Emotion (JACFEE): Reliability data and cross-national
differences. Journal of Nonverbal Behavior, :I, ,:I.
Bezooijen, R. van., Oto, S. A., & Heenan, T. A. (I,8,). Recognition of vocal expressions of emo-
tion: A three-nation study to identify universal characteristics. Journal of Cross-Cultural
Psychology, I, ,8,oo.
Boucher, J. D. (I,,,). Culture and emotion. In A. J. Marsella, R. G. Tharp, & T. V. Ciborowski
(eds), Perspectives on cross-cultural psychology (pp. I,,,8). San Diego, CA: Academic
Press.
Briggs, J. L. (I,,o). Never in anger: Portrait of an Eskimo family. Cambridge, MA: Harvard
University Press.
Brunswik, E. (I,,o). Perception and the representative design of psychological experiments.
Berkeley, CA: University of California Press.
Carlsen, J. C. (I,8I). Some factors which inuence melodic expectancy. Psychomusicology,
I, I::,.
Castellano, M. A., Bharucha, J. J., & Krumhansl, C. L. (I,8). Tonal hierarchies in the music of
North India. Journal of Experimental Psychology: General, II,, ,,I:.
Clayton, M. (:ooI). Introduction: towards a theory of musical meaning (in India and else-
where). British Journal of Ethnomusicology, Io, II8.
Cross, I. (:oo,). Music, cognition, culture and evolution. In I. Peretz & R. J. Zatorre (eds), The
cognitive neuroscience of music (pp. :,o). Oxford: Oxford University Press.
Cross, I. (:ooo). The origins of music: Some stipulations on theory. Music Perception, :,
,,8:.
Crowder, R. G., Resnick, J. S., & Rosenkrantz, S. L. (I,,I). Perception of the major/minor
distinction: V. Preferences among infants. Bulletin of the Psychonomic Society, :,, I8,8.
Cuddy, L. L., Cohen, A., & Mewhort, D. J. (I,8I). Perception of structure in short melodic
sequences. Journal of Experimental Psychology: Human Perception and Performance, ,,
8o,8,.
Dalla Bella, S., Peretz, I., Rousseau, L., & Gosselin, N. (:ooI). A developmental study of the
affective value of tempo and mode in music. Cognition, 8o, BIBIo.
Darwin, C. (I8,:). Expression of the emotions in man and animals. London: John Murray.
Deva, B. C., & Virmani, K. G. (I,,,). A study in the psychological response to ragas. (Research
Report II of Sangeet Natak Akademi). New Delhi, India: Indian Musicological Society.
de Villiers-Sidani, E., Chang, E. F., Bao, S., & Merzenich, M. M. (:oo,). Critical period for spec-
tral tuning dened in the primary auditory cortex (AI) in the rat. Journal of Neuroscience,
:,, I8o,.
Dissanayake, E. (:ooI). Antecedents of the temporal arts in early mother-infant interaction. In
N. L. Wallin, B. Merker, & S. Brown (eds), The origins of music (pp. ,8,Io). Cambridge,
MA: MIT Press.
Dowling, W. J., & Harwood, D. L. (I,8o). Music cognition. New York: Academic Press.
Eerola, T., Himberg, T., Toiviainen, P., & Louhivuori, J. (:ooo). Perceived complexity of west-
ern and African folk melodies by western and African listeners. Psychology of Music, ,,
,,,,I.
Eid, M., & Diener, E. (:ooI). Norms for experiencing emotions in different cultures: Inter- and
intranational differences. Journal of Personality and Social Psychology, 8I, 8o,8,.
Ekman, P. (I,,o). Universal facial expressions of emotion. California Mental Health Research
Digest, 8, I,I8.
Ekman, P. (I,,:). Universals and cultural differences in facial expressions of emotion. In
J. Cole (ed.), Nebraska symposium on motivation (pp. :o,8,). Lincoln, NE: University of
Nebraska Press.
27-Juslin & Sloboda-Ch-27.indd 784 7/24/2009 7:50:52 PM
references ,8,
Ekman, P. (I,,o). Pictures of facial affect. Palo Alto, CA: Consulting Psychologists Press.
Ekman, P. (I,,:). An argument for basic emotions. Cognition & Emotion, o, Io,:oo.
Ekman, P., Sorenson, E. R., & Friesen, W. V. (I,o,). Pan-cultural elements in facial displays of
emotions. Science, Io, 8o8.
Elfenbein, H. A., & Ambady, N. (:oo:). On the universality and cultural specicity of emotion
recognition: A meta-analysis. Psychological Bulletin, I:8, :o,,,.
Elfenbein, H. A., & Ambady, N. (:oo,). Cultural similaritys consequences: A distance
perspective on cross-cultural differences in emotion recognition. Journal of Cross Cultural
Psychology, ,, ,:Io,.
Frick, R. W. (I,8,). Communicating emotion: The role of prosodic features. Psychological
Bulletin, ,,, I::,.
Fritz, T., Jentschke, S., Gosselin, N., Sammler, D., Peretz, I., Friederici, A. D., & Koelsch, S. (in
press). Universal recognition of three basic emotions in music. Current Biology.
Fritz, T., Sammler, D., & Koelsch, S. (:ooo). How far is music universal? An intercultural
comparison. In M. Baroni, A. R. Addessi, R. Caterina, & M. Costa (eds), ,th International
Conference on Music Perception and Cognition (p. 88). Bologna, Italy: Bononia University
Press.
Geertz, C. (I,,,). Thick description: Toward an interpretive theory of culture. In The interpre-
tation of cultures: Selected essays (pp. ,,o). New York: Basic Books.
Gundlach, R. H. (I,,:). A quantitative analysis of Indian music. American Journal of Psychology,
, I,,,.
Gundlach, R. H. (I,,,). Factors determining the characterization of musical phrases. American
Journal of Psychology, ,, o:,.
Hannon, E. E., & Trainor, L. J. (:oo,). Music acquisition: Effects of enculturation and formal
training on development. Trends in Cognitive Sciences, II, oo,:.
Hebdige, D. (I,,,). Subculture: The meaning of style. London: Routledge.
Heyes, C. (:oo,). Four routes of cognitive evolution. Psychological Review, IIo, ,I,:,.
Huron, D. (:ooo). Sweet anticipation: Music and the psychology of expectation. Cambridge, MA:
MIT Press.
Ilie, G., & Thompson, W. F. (:ooo). A comparison of acoustic cues in music and speech for
three dimensions of affect. Music Perception, :,, ,I,:,.
Johnson-Laird, P. N., & Oatley, K. (I,8,). The language of emotions: An analysis of a semantic
eld. Cognition & Emotion, ,, 8II:,.
Johnstone, T., & Scherer, K. R. (:ooo). Vocal communication of emotion. In M. Lewis &
J. M. Haviland-Jones (eds), Handbook of emotions (:nd edn, pp. ::o,,). New York:
Guilford Press.
Juslin, P. N. (:ooo). Cue utilization in communication of emotion in music performance:
Relating performance to perception. Journal of Experimental Psychology: Human Perception
and Performance, :o, I,,,8I,.
Juslin, P. N. (:ooI). Communicating emotion in music performance. A review and a theoreti-
cal framework. In P. N. Juslin & J. A. Sloboda (eds), Music and emotion: Theory and research
(pp. ,o,,,). Oxford: Oxford University Press.
Juslin, P. N., & Laukka, P. (:oo,). Communication of emotions in vocal emotion and music
performance: Different channels, same code? Psychological Bulletin, I:,, ,,o8I.
Juslin, P. N., & Vstfjll, D. (:oo8). Emotional responses to music: The need to consider
underlying mechanisms. Behavioral and Brain Sciences, ,I, ,,,,,.
Justus, T., & Hutsler, J. J. (:oo,). Fundamental issues in the evolutionary psychology of music:
Assessing innateness and domain-specicity. Music Perception, :,, I:,.
Katsiavriades, K., & Qureshi, T. (:oo,). Ten language families in detail. Retrieved from: http://
www.Krysstal.com
27-Juslin & Sloboda-Ch-27.indd 785 7/24/2009 7:50:52 PM
,8o cross-cultural similarities and differences
Kessler, E. J., Hansen, C., & Shepard, R. N. (I,8). Tonal schemata in the perception of music
in Bali and the west. Music Perception, :, I,Io,.
Kilbride, J. E., & Yarczower, M. (I,8,). Ethnic bias in the recognition of facial expressions.
Journal of Nonverbal Behavior, 8, :,I.
Kitayama, S., & Ishii, K. (:oo:). Word and voice: Spontaneous attention to emotional utter-
ances in two languages. Cognition & Emotion, Io, :,,,.
Krumhansl, C. L. (I,,,). Music psychology and music theory: Problems and prospects. Music
Theory Spectrum, I,, ,,8o.
Krumhansl, C. L., Toivanen, P., Eerola, T., Toiviainen, P., Jrvinen, T., & Louhivuori,
J. (:ooo). Cross-cultural music cognition: Cognitive methodology applied to North Sami
yoiks. Cognition, ,o, I,,8.
Krumhansl, C. L., & Keil, F. C. (I,8:). Acquisition of the hierarchy of tonal functions in music.
Memory & Cognition, Io, :,,I.
Lazarus, R. S. (I,,I). Emotion and adaptation. Oxford: Oxford University Press.
Lerdahl, F., & Krumhansl, C. L. (:ooo). Modeling tonal tension. Music Perception, :, ,:,oo.
Livingstone, S., & Thompson, W. F. (in press). The emergence of music from the Theory of
Mind. Musica Scientiae.
Lomax, A. (I,o:). Song structure and social structure. Ethnology, I, :,,I.
Lomax, A. (I,,,). The good and the beautiful in folksong. Journal of American Folklore, 8o,
:I,:,,.
Lomax, A. (I,,,). Cantometrics: An approach to the anthropology of music. Audiocassettes and a
handbook. Berkeley, CA: University of California Media Extension Center.
Lomax, A. (I,,8). Folk song style and culture. Edison, NJ: Transaction Publishers.
Lutz, C. (I,88). Unnatural emotions: everyday sentiments on a Micronesian atoll and their
challenge to western theory. Chicago, IL: University of Chicago Press.
Lynch, M. P., & Eilers, R. E. (I,,I). Childrens perception of native and nonnative musical
scales. Music Perception, ,, I:I,I.
Masuda, T., & Nisbett, R. E. (:ooI). Attending holistically versus analytically: Comparing the
context sensitivity of Japanese and Americans. Journal of Personality and Social Psychology,
8I, ,::,.
Matsumoto, D. (I,8,). Cultural inuences on the perception of emotion. Journal of Cross-
Cultural Psychology, :o, ,:Io,.
Matsumoto, D. (I,,o). Cultural similiarities and differences in display rules. Motivation and
Emotion, I, I,,:I.
Matsumoto, D. (I,,:). American-Japanese cultural differences in the recognition of universal
facial expressions. Journal of Cross-Cultural Psychology, :,, ,:8.
Matsumoto, D., Yoo, S. H., Hirayama, S., & Petrova, G. (:oo,). Development and validation
of a measure of display rule knowledge: the Display Rule Assessment Inventory. Emotion,
,, :,o.
Mesquita, B., & Frijda, N. H. (I,,:). Cultural variations in emotions: A review. Psychological
Bulletin, II:, I,,:o.
Mesquita, B., Frijda, N. H., & Scherer, K. R. (I,,,). Culture and emotion. In J. W. Berry,
P. R. Dasen, & T. S. Saraswathi (eds), Handbook of cross-cultural psychology, Vol. :: Basic
processes and human development (:nd edn, pp. :,,,,). Needham Heights, MA: Allyn &
Bacon.
Morey, R. (I,o). Upset in emotions. Journal of Social Psychology, I:, ,,,,o.
Nakata, T., & Trehub, S. E. (:oo). Infants responsiveness to maternal speech and singing.
Infant Behavior and Development, :,, ,,o.
Narmour, E. (I,,o). The analysis and cognition of basic melodic structures: the Implication-
Realization Model. Chicago, IL: Chicago University Press.
27-Juslin & Sloboda-Ch-27.indd 786 7/24/2009 7:50:52 PM
references ,8,
Nettl, B. (I,8,). The study of ethnomusicology. Chicago, IL: University of Illinois Press.
Nisbett, R. E., Peng, K., Choi, I., & Norenzayan, A. (:ooI). Culture and systems of thought:
Holistic versus analytic cognition. Psychological Review, Io8, :,I,Io.
Palmer, C., & Hutchins, S. (:ooo). What is musical prosody? The Psychology of Learning and
Motivation, o, :,,8.
Rosch E., & Mervis, C. (I,,,). Family resemblances: studies in the internal structure of catego-
ries. Cognitive Psychology, ,, ,,,oo,.
Rose, C. (:oo,). Sino-Japanese relations: facing the past, looking to the future. London:
Routledge.
Rosenthal, R., Hall, J. A., DiMatteo, M. R., Rogers, P. I., & Archer, D. (I,,,). Sensitivity to non-
verbal communication: the PONS test. Baltimore, MD: Johns Hopkins University Press.
Russell, J. A. (I,,I). Culture and the categorization of emotions. Psychological Bulletin, IIo,
:o,o.
Russell, J. A. (I,,). Is there universal recognition of emotion from facial expression? A review
of cross-cultural studies. Psychological Bulletin, II,, Io:I.
Russell, J. A., Suzuki, N., & Ishida, N. (I,,,). Canadian, Greek, and Japanese freely produced
emotion labels for facial expressions. Motivation and Emotion, I,, ,,,,I.
Saffran, J. R., E. K., Aslin, R. N., & Newport, E. L. (I,,o). Statistical learning by 8-month-old
infants. Science, :,, I,:o8.
Schellenberg, E. G. (I,,o). Expectancy in melody: Tests of the implication-realization model,
Cognition, ,8, ,,I:,.
Scherer, K. R. (I,,,). Non-linguistic vocal indicators of emotion and psychopathology. In
C. E. Izard (ed.), Emotions in personality and psychopathology (pp. ,,,:,). New York:
Plenum Press.
Scherer, K. R. (I,8:). Methods of research on vocal communication: Paradigms and parame-
ters. In K. R. Scherer & P. Ekman (eds), Handbook of methods in nonverbal behavior research
(pp. I,o,8). Cambridge, UK: Cambridge University Press.
Scherer, K. R. (I,8o). Vocal affect expression: A review and a model for future research.
Psychological Bulletin, ,,, I,o,.
Scherer, K. R. (I,8,). Vocal correlates of emotional arousal and affective disturbance. In
H. Wagner & A. Manstead (eds), Handbook of social psychophysiology (pp. Io,,,).
New York: Wiley.
Scherer, K. R., & Oshinsky, J. S. (I,,,). Cue utilization in emotion attribution from auditory
stimuli. Motivation and Emotion, I, ,,Io.
Singh, L., Morgan, J. L., & Best, C. T. (:oo:). Infants listening preferences: Baby talk or happy
talk? Infancy, ,, ,o,,.
Sorenson, E. R. (I,,,). Culture and the expression of emotion. In T. R. Williams (ed.),
Psychological anthropology (pp. ,oI,:). Chicago, IL: Aldine.
Standke, R. (I,,:). Methods of digital speech analysis in research on vocal communication.
Frankfurt, Germany: Peter Lang.
Thompson, W. F., & Balkwill, L-L. (:ooo). Decoding speech prosody in ve languages.
Semiotica, I,8, o,:.
Thompson, W. F., Graham, P., & Russo, F. A. (:oo,). Seeing music performance: Visual inu-
ences on perception and experience. Semiotica, I,o, :o,:,.
Thompson, W. F., & Robitaille, B. (I,,:). Can composers express emotions through music?
Empirical Studies of the Arts, Io, ,,8,.
Thompson, W. F., Russo, F. A., & Quinto, L. (in press). Audio-visual integration of emotional
cues in song. Cognition & Emotion.
Thompson, W. F., & Schellenberg, E. G. (:ooo). Listening to music. In R. Colwell (ed.), MENC
handbook of music cognition & development (pp. ,:I:,). Oxford: Oxford University Press.
27-Juslin & Sloboda-Ch-27.indd 787 7/24/2009 7:50:53 PM
,88 cross-cultural similarities and differences
Thompson, W. F., Schellenberg, E. G., & Husain, G. (:oo). Decoding speech prosody: Do
music lessons help? Emotion, , oo.
Thompson, W. F., & Stainton, M. (I,,8). Expectancy in Bohemian folksong melodies:
Evaluation of implicative principles for implicative and closural intervals. Music Perception,
I,, :,I,:.
Tooby, J., & Cosmides, L. (I,,o). The past explains the present: Emotional adaptations and the
structure of ancestral environments. Ethology and Sociobiology, II, ,,,:.
Trainor, L. J., & Trehub, S. E. (I,,). Key membership and implied harmony in Western tonal
music: Developmental perspectives. Perception & Psychophysics, ,o, I:,,:.
Trainor, L. J., & Heinmiller, B. M. (I,,8). The development of evaluative responses to music:
Infants prefer to listen to consonance over dissonance. Infant Behavior & Development, :I,
,,88.
Trehub, S. E. (:ooo). Human processing predispositions and musical universals. In
N. L. Wallin, B. Merker, & S. Brown (eds), The origins of music (pp. :,8). Cambridge,
MA: MIT Press.
Trehub, S. E. (:oo,). The developmental origins of musicality. Nature Neuroscience, o,
oo,,,.
Trehub, S. E., & Hannon, E. E. (:ooo). Infant music perception: Domain-general or domain-
specic mechanisms. Cognition, Ioo, ,,,,.
Trehub, S. E., & Nakata, T. (:oo:). Emotion and music in infancy. Musicae Scientiae, Special
Issue :ooI:, ,,oI.
Trehub, S. E., Schellenberg, E. G., & Hill, D. S. (I,,,). The origins of music perception and
cognition: A developmental perspective. In I. Delige & J. A. Sloboda (eds), Perception and
cognition of music (pp. Io,:8). Hove, UK: Psychology Press.
Trehub, S. E., Unyk, A., & Trainor, L. J. (I,,,). Maternal singing in cross-cultural perspective.
Infant Behaviour and Development, Io, :8,,,.
Unyk, A. M., Trehub, L. J., & Schellenberg, E. G. (I,,:). Lullabies and simplicity: A cross-
cultural perspective. Psychology of Music, :o, I,:8.
Wittgenstein, L. (I,,,/:ooI). Philosophical investigations. Oxford: Blackwell.
Zentner, M. R., & Kagan, J. (I,,o). Perception of music by infants. Nature, ,8,, :,.
27-Juslin & Sloboda-Ch-27.indd 788 7/24/2009 7:50:53 PM

You might also like