You are on page 1of 21


Current Advances in the Cognitive

Neuroscience of Music
Daniel J. Levitin and Anna K. Tirovolas
McGill University, Montreal, QC Canada

The study of music perception and cognition is one of the oldest topics in experimental
psychology. The last 20 years have seen an increased interest in understanding the func-
tional neuroanatomy of music processing in humans, using a variety of technologies
including fMRI, PET, ERP, MEG, and lesion studies. We review current findings in the
context of a rich intellectual history of research, organized by the cognitive systems
underlying different aspects of human musical behavior. We pay special attention to the
perception of components of musical processing, musical structure, laterality effects,
cultural issues, links between music and movement, emotional processing, expertise,
and the amusias. Current trends are noted, such as the increased interest in evolu-
tionary origins of music and comparisons of music and language. The review serves
to demonstrate the important role that music can play in informing broad theories of
higher order cognitive processes such as music in humans.

Key words: music; language; emotion; structure; evolutionary psychology; expertise

Introduction dress questions about part−whole relationships

in music and melody (Ehrenfels 1890/1988).
The field of music cognition traces its ori- The past decade has seen an exponential
gins to the 4th century BCE, long before increase in studies of music cognition. Musi-
the establishment of experimental psychol- cal behaviors that are typically studied include
ogy itself, through the ideas of Aristoxenus, listening, remembering, performing, learning,
an Aristotelian philosopher. Contrary to the composing, and, to a lesser extent, movement
Pythagoreans of that time, Aristoxenus ar- and dancing. The largest paradigm shift has
gued that musical intervals should be classi- been the increased use of neuroimaging and
fied by their effects on listeners as opposed neural case studies to inform theories about
to merely examining their mathematical ra- the brain basis for musical behaviors. A second
tios (Griffiths 2004; Levitin 1999). This notion theme over the past decade has been an in-
brought the scientific study of music into the creased interest in the origins of music and its
mind, followed by the first psychophysics exper- connection with language, both evolutionarily
iments at the dawn of experimental psychology, and functionally.
which mapped changes in the physical world In cognitive neuroscientific studies of lan-
onto changes in the psychological world (e.g., guage, mathematical ability, or visual percep-
Fechner 1860; Helmholtz 1863/1954). Indeed, tion, one rarely encounters a definition of the
many of the earliest studies in experimental psy- capacity being studied, yet the question of
chology concerned music, and the Gestalt psy- just what is music (and by implication, what
chology movement was formed in part to ad- it is not) is one that emerges more often in
this field of inquiry than in the others. Those
who study music cognition often rely on the
theorist Leonard Meyer, who defined it as
Address for correspondence: Daniel J. Levitin, Ph.D., Department of
Psychology, McGill University, 1205 Avenue Dr. Penfield, Montreal, QC
a form of emotional communication, or on
H3A 1B1 Canada. the definition of the composer Edgar Varése,
The Year in Cognitive Neuroscience 2009: Ann. N.Y. Acad. Sci. 1156: 211–231 (2009).
doi: 10.1111/j.1749-6632.2009.04417.x  C 2009 New York Academy of Sciences.

212 Annals of the New York Academy of Sciences

Figure 1. Core brain regions associated with musical activity. Based on Tramo 2001 and updated in
2006 (from Levitin 2006).

who famously defined it as “organized sound.” The Origins of Music

Music can be seen as a form of artistic expres-
sion, communication, self-expression and self- At the annual meeting of the Society of Mu-
discovery, or as an auditory art form. Music sic Perception and Cognition (SMPC) held at
most typically involves variations in pitch and M.I.T. during the summer of 1997, the cog-
rhythm that are composed or improvised with nitive scientist Steven Pinker made a now fa-
the purpose of inducing emotional responses in mous declaration that music cognition is “not
the listener. However, these are neither neces- worth studying” because, he said, it is “auditory
sary nor sufficient conditions, and one is usu- cheesecake,” an evolutionary byproduct of the
ally left with a Wittgensteinian conclusion that adaptation for human language. The phrase
a new exemplar can be considered music if it auditory cheesecake derives from a rhetorical chal-
bears a “family resemblance” to other exam- lenge to evolutionary theory: If evolution se-
ples that are generally agreed to be “music.” lects those behaviors that are maximally adap-
As studied in the laboratory, researchers typ- tive, how do you explain that many of us like
ically examine variations in one musical at- fats and sweets (as in cheesecake, for example),
tribute while holding the others constant so which can actually lead to obesity, diabetes, and
as to maintain experimental control. A re- other clearly maladaptive outcomes?
view of the literature can be parsed in at least “We enjoy strawberry cheesecake, but not
three ways: by the discipline of those who because we evolved a taste for it,” Pinker ar-
study it (e.g., psychology, neuroscience, musi- gues. “We evolved circuits that gave us trickles
cology, music theory, sociology, anthropology, of enjoyment from the sweet taste of ripe fruit,
biology); by the attribute of the musical sig- the creamy mouth feel of fats and oils from
nal studied (rhythm, pitch, melody, timbre); or nuts and meat, and the coolness of fresh water.
by those mental processes involved. Here, we Cheesecake packs a sensual wallop unlike any-
choose to organize this review using the lat- thing in the natural world because it is a brew
ter, cognitive−systematic approach, with a brief of megadoses of agreeable stimuli which we
opening discussion of the origins of music. We concocted for the express purpose of press-
limit our discussion to music in humans. ing our pleasure buttons” (Pinker 1997,
Levitin & Tirovolas: Cognitive Neuroscience of Music 213

p. 525). Moreover, in the quantities that fats motor coordination (because in evolutionary
and sweets would have been available to time frames music was usually accompanied by
our hunter−gatherer ancestors, they posed no dance).
Pinker argues that music exploits circuits that
evolved for spoken language: that language was Perception and Musical Structure
the evolutionary adaptation, music the byprod-
uct or spandrel. He feels similarly about lit- Music is characterized by eight perceptual
erature and the other arts, that the pleasures attributes, or dimensions, each of which can be
afforded by them are incidental (cf. Carroll varied independently: pitch, rhythm, timbre,
1998). Michael Gazzaniga (2008) and others tempo, meter, contour, loudness, and spatial lo-
(e.g., Tooby & Cosmides 2001) believe that cation (Levitin 1999; Pierce 1983). Perceptual
artistic thinking in general would have been grouping in music occurs as a function of princi-
essential to early human development. An abil- ples similar in some ways to those for grouping
ity to engage in and enjoy fictional thinking in vision. Grouping by similarity of timbre and
would have conferred an evolutionary advan- loudness has been demonstrated, as has group-
tage to our ancestors. They could consider hy- ing by proximity of pitch or onset time, and
pothetical scenarios and plan their responses by good continuation of pitch (Bregman 1990).
to them ahead of time, without having to try Temporal grouping of tones into subsequences
various alternatives during a moment of im- gives rise to the perception of meter (common
minent danger, such as a confrontation with meters in Western music are based on group-
a predator. Music, and indeed all art, derives ings of 2, 3, or 4 primary beats). The points
from three abilities that are hallmarks of hu- over time at which one would naturally tap a
man cognition: theory of mind, recursion, and foot or snap a finger to accompany music are
abstract representation (Levitin 2008, see also called the tactus, the underlying beat or pulse.
Cosmides & Tooby 1995). Each human culture develops its own tradi-
Music composition and improvisation can tions for the ways in which the eight percep-
be seen as a preparatory activity for training tual attributes are employed to create music.
cognitive flexibility—arranging and rearrang- The system of rules or conventions by which
ing the elements of pitch and rhythm over time sounds are strung together in a given culture
is now believed to exercise attentional networks can be thought of as the grammar for that music
(or more generally, executive function; H. J. (Lerdahl & Jackendoff 1983; Lerdahl 2001) and
Neville, personal communication, November, as reflecting a musical style. Musical and lin-
2006) and from an evolutionary standpoint guistic grammar allow for the generation of a
can be seen as an “honest signal” for men- theoretically infinite number of songs or sen-
tal and emotional flexibility and fitness (Cross tences through combinations and rearrange-
& Woodruff, in press, Miller 2000; Sluming & ments of elements. Tonality occurs as a function
Manning 2000). Moreover, our primitive ances- of either the simultaneous or sequential sound-
tors who could sing and dance for hours on end, ing of tones. In Western tonal music, major and
creating variations on themes, were indicating minor tonalities are the principal distinctions.
to potential mates their cognitive and physical Other musical traditions use different concep-
flexibility and fitness—skills that could come in tions; we restrict our discussion here to Western
handy if the food supply ran out or one needed tonal music.
to hastily build a new shelter or escape from a Early reports (e.g., Bever & Chiarello
predator. Music-dance, among other cognitive 1974) stated that music is predominantly a
displays, would have indicated the presence of right-hemisphere activity and language, left
the creative mind as well as physical fitness and (in neurologically intact right-handed listeners).
214 Annals of the New York Academy of Sciences

This is now considered to be an oversimplifica- mation (Dowling 1978; Narmour 1990; White
tion, in part because of the distributed nature 1960), which suggests that human music per-
of specialized processing mechanisms acting on ception may be qualitatively different from that
the individual musical attributes listed above. It of most animal species. There has been a his-
is now known that music listening, perform- tory of debate in the animal learning literature
ing, and composing engage regions throughout regarding whether animals’ mental representa-
the brain, bilaterally, and in the cortex, neo- tions are relational or absolute (Hanson 1959;
cortex, paleo-, and neocerebellum (Peretz & Kohler 1918/1938; Reese 1968; Spence 1937).
Zatorre 2003; Platel et al. 1997; Sergeant 1993; Even most bird species do not recognize their
Tramo 2001). Laterality effects do exist, how- own songs in transposition (Hauser & McDer-
ever. For example, magnetic encephalography mott 2003). In human listeners the absolute
(MEG) responses to deviations in the mem- values of a tone’s pitch and duration are pro-
orized lyrics of tunes are stronger in the left cessed, and when there is more than one tone
hemisphere, while the perception of violations present it is the processing of tonal relations
of expected notes are governed by the right that gives rise to the appreciation of melody
hemisphere (Yasui et al. 2008). The act of (Dowling & Harwood 1986). To some extent
learning music causes a left hemisphere shift tonal relations are computed even when only a
(Ohnishi et al. 2001), particularly as naming single tone is presented—the listener is aware,
processes become involved (such as naming for example, that the presented tone is higher
musical intervals, chords, etc.; Zatorre et al. or lower, or longer or shorter, than a conceptual
1998). average tone encountered across the course of
Evidence supports the differential specializa- a lifetime.
tion of the left and right mesial temporal lobes Tonal relations, or musical intervals (as op-
in learning new melodies (Wilson & Saling posed to large-scale musical structure), have
2008). When presented with a learning task for been shown to be predominantly served by net-
novel melodies in tonal and atonal contexts, pa- works in the right temporal region (Liégeois-
tients with either left- or right-sided mesial tem- Chauvel et al. 1998; Zatorre 1985) and in the
poral damage were impaired in interval recog- left dorsolateral prefrontal and right inferior
nition compared to normal controls. However, frontal cortex (Zatorre et al. 1998), with par-
when memorizing melodies within a tonal con- ticular deficits noted following lesions of the
text, individuals with right mesial temporal right anterolateral part of Heschl’s gyrus (John-
damage in particular were unable to use im- srude et al. 2000; Tramo et al. 2002; Zatorre
plicit knowledge of Western musical tonality to 1988). Neuroimaging studies have shown that a
aid their memory. hierarchy of pitch processing operations seems
While our subjective experience of music to exist. Fixed pitches and noise are processed
may seem complete and seamless, this phe- in Heschl’s gyrus bilaterally (Patterson et al.
nomenological unity belies the fact that the per- 2002). Posterior regions of secondary audi-
ceptual components are processed separately. tory cortex process pitch height, and ante-
Primary auditory cortex in both cerebral hemi- rior regions process pitch chroma (pitch class)
spheres in most mammals contains a tonotopic (Warren et al. 2003). Intervals, contour, and
map—a map of pitches running from low to melody activate the superior temporal gyrus
high, which mirrors the neuronal pitch map in (STG) and planum polare (PP) (Patterson et al.
the cochlea (Bendor & Wang 2005) and allows 2002).
for the encoding of pitch height (that dimen- Neuroimaging studies have shown that pos-
sion of pitch perception that correlates with terior regions of secondary auditory cortex
frequency). Human perception of music relies process pitch height and anterior regions pro-
on pitch relations as well as absolute pitch infor- cess pitch chroma or pitch class (Warren et al.
Levitin & Tirovolas: Cognitive Neuroscience of Music 215

2003; see also (Tervaniemi & Huotilainen 2003) thus induced by the temporal context created
for converging evidence from the event-related by the stimuli rather than the perception of a
potential, or ERP). prominent tone created by deviations in pitch
An examination of blood oxygenation level- or volume; these rhythmic sequences were cre-
dependent (BOLD) responses using functional ated as metrically simple, metrically complex,
magnetic resonance imaging (fMRI) revealed and nonmetric. Metrically simple rhythms were
that major and minor tonalities recruit the bi- more accurately reproduced by the partici-
lateral inferior frontal gyri, medial thalamus, pants. Furthermore, both simple and complex
and the dorsal cingulate cortex (Mizuno & rhythms generate activity in several areas as
Sugishita 2007). The authors suggest that the measured by fMRI, including motor areas
frontal and thalamic regions are implicated in such as the supplemental motor area, basal
judging tonality, while the cingulate may be ganglia, and cerebellum. However, specific to
recruited for the resolution of mental conflict rhythms in which accents arise at regular inter-
in response when the participant differentiates vals (which give the feeling of a simple beat), the
modality. Minor, compared to major chords, basal ganglia, pre-SMA/SMA, and the ante-
shows selective activation in the amygdala, ret- rior superior temporal gyri showed greater ac-
rosplenial cortex, brain stem, and cerebellum tivation. These areas may subserve the internal,
(Pallesen et al. 2005), and in a separate study intuitive, “beat-based” timer in both musicians
of mode melodies, activation was found in left and nonmusicians. The increased performance
parahippocampal gyrus, bilateral ventral ante- ability for simple, perceptually salient rhythms
rior cingulate, and left medial prefrontal cortex suggests the existence of a metrically regular, in-
(Green et al. 2008). (We note that in this and ternal timer; such a timer in adults may prefer-
subsequent fMRI studies reported in this re- entially support the processing of small-integer
view the relation between neural activation to ratio temporal intervals typical of Western mu-
a stimulus and neural deactivation is an area sic (cf. Ivry & Hazeltine 1995; Poppel 1997;
of current inquiry and much work remains to Sternberg et al. 1982).
be done on this issue. Not all papers report The aforementioned neuroanatomical stud-
deactivation, and in those that do, the clear ies suggest a theoretical model of functional
interpretation of deactivations has not been es- architecture whereby distinct neural circuits for
tablished.) music grouped into pitch organization and tem-
Rhythm perception and production invoke poral organization represent an interactive sys-
regions in the cerebellum and basal ganglia tem of music processing (Peretz & Coltheart
(Ivry & Keele 1989; Janata & Grafton 2003), as 2003). Although the extent to which pitch and
well as several motor areas such as the premo- rhythm processing are separable or constitute
tor cortex and supplemental motor area (Hals- Fodorian modules (Fodor 1983) is not entirely
band et al. 1993). Timing, synchrony, and en- understood, double-dissociation evidence from
trainment may be subserved by a system of patient populations strongly suggests indepen-
hierarchically controlled oscillators in the cere- dence of pitch and rhythm processing (Ayotte
bellum (Ivry & Hazeltine 1995; Ivry & Schlerf et al. 2000; Di Pietro et al. 2004; Liégeois-
2008; Sternberg et al. 1982) that contribute Chauvel et al. 1998; Peretz 1990; Peretz &
to our sense of tempo (Levitin & Cook 1996). Kolinsky 1993; Piccirilli et al. 2000; Vignolo
One recent experiment (Grahn & Brett 2007) 2003). The prevailing view is thus that pitch,
investigated the perception and production of rhythm, and loudness are processed separately
both regular (small integer ratios) and irreg- and then come together later (where “later” in
ular (complex integer ratios) rhythmic group- neural processing time may be 25–50 ms later)
ings in monotonal (same pitch) sequences. The to give us the impression of a fully realized mu-
perception of an accented tone, or beat, was sical object or phrase.
216 Annals of the New York Academy of Sciences

The perception of certain tonal and rhyth- cal culture (Trehub et al. 1999). The ability that
mic structures (such as octave equivalence) children have in detecting changes in key and
appears to be innate, pointing to a possible evo- harmony in their native music appear between
lutionary function (as mentioned earlier). Dur- the ages of 5 and 7 years (Trainor & Trehub
ing the first year of life, infants prefer pleasing 1994). The processes of acquiring knowledge of
to displeasing musical intervals (Trainor et al. one’s musical culture can be viewed as involving
2002). Infants are also capable of perceiving vi- statistical learning (Saffran et al. 1996; Saffran
olations in complex meter, a feature that char- et al. 1999). Native listeners of their musical sys-
acterizes much non-Western music, a capacity tem, after exposure to thousands of tonal
that declines at the end of the first year of life sequences, implicitly learn which tones and
(Hannon & Trehub 2005). This suggests that chords are mostly likely to complete a mu-
maturation involves becoming sensitive to the sical sequence. Composers sometimes reward
music of one’s culture during the first year of and sometimes violate listener expectations
receiving musical input, and that humans may (Narmour 1990) but do so within this system
be born with the capacity to learn any of the of legal tones for their culture’s music. Rarely
world’s music forms. By the age of 5 years chil- if ever does one encounter a tone from out-
dren demonstrate an adultlike electrophysio- side one’s musical system. Even upon hearing a
logical response called the “early right anterior relatively complex piece for the first time—say
negativity” response or ERAN, and a negative by Schoenberg or Reich—one would not sud-
voltage response approximately 500 ms after denly encounter a tone from an Indian Raga
the event, known as the N5, to violations of or a pygmy scale.
musical syntax for the music of their culture Experiments that introduce stimuli violating
(Jentschke et al. 2008). the rules of musical grammar have been em-
The now well-known idea put forth by ployed to investigate how the human brain pro-
Chomsky (1965) was that humans enter the cesses musical structure. When presented with a
world equipped with a “language acquisition violation in a chord sequence, an ERP response
device” (LAD). That is, given proper input, called ERAN is elicited (Koelsch et al. 2007).
we have the cognitive equipment to automati- This response is not due to acoustic variation
cally acquire language. Because of the presence in the chord sequence, but to the irregularity of
of precocious perceptual abilities of infants, a its musical grammar or violation of expectation
nativist position is held by the majority of re- (the so-called oddball paradigm). We return to
searchers in this field. Our intrinsic capacity this topic under Music and Language below,
for music is seen as leaning to nature, while as numerous studies using this paradigm have
the learning of specific musical forms relies on been conducted to discern possible associations
nurture—specifically on exposure during a sen- and dissociations between grammar processing
sitive or critical period of development (Trainor in music and language.
2005). This strong nativist position in the liter- Musical context has been found to be cru-
ature on early musical capacities suggests that cial to music perception. For example, in clas-
a counterpart in music to the LAD indeed ex- sic probe-tone studies (cf. Krumhansl & Kessler
ists, which we will call the “music acquisition 1982) participants judged the perceptual con-
device”. gruence of chords after being primed by partic-
The way in which pitches and, to a lesser de- ular musical scales, creating in them a percep-
gree, rhythms may be lawfully strung together tual space for chord stability that was found to
constitutes the grammar of a given musical style be substantially the same as Western music the-
or culture. It has been shown that infants at ory would predict; in other words, the average
the age of 9 months are sensitive to particular listener implicitly internalizes the rules of West-
characteristics of the scales of their own musi- ern tonal music. In a recent study investigating
Levitin & Tirovolas: Cognitive Neuroscience of Music 217

the neuromagnetic response to the probe-tone active during the early part of each transition,
method (Otsuka et al. 2008), the contextual and a dorsal fronto-parietal network, including
modality (major or minor mode) as well as the dorsolateral prefrontal cortex (dlPFC, BA9)
perceptual priming affected the perception of and posterior parietal cortex (PPC, BA 40), was
chord tonality, as measured by activation at the active during the later part. These activations
level of the auditory cortex. were predominantly right lateralized.
Internalizing the rules of one’s musical cul- Prediction and anticipation are truly at the
ture naturally makes the processing of tonal heart of the musical experience. Even nonmu-
structures in that music more automatic and sicians are actively engaged, at least subcon-
efficient. This doesn’t mean that one can’t en- sciously, in tracking the ongoing development
joy music from outside one’s culture, but sug- of a musical piece and forming predictions
gests that doing so may carry additional cog- about what will come next. Typically in music,
nitive costs. Nan et al. (2008) studied this by when something will come next is known, due
examining the differences in neural activation to music’s underlying pulse or rhythm (what
for native and non-native music. They found musicians call the ‘‘tactus’’), but less known
that native music engages the ventro-medial is what will occur next. There is an impor-
prefrontal cortex (VMPC), an area known to tant link between such predictive processes
activate when an individual processes informa- and the formation of event boundaries: In mu-
tion with ease, or a “feeling of knowing,” as sic the VLPFC has been consistently implicated
well as motor regions typically found to acti- in the detection of violations in musical ex-
vate during music listening. Moreover, the same pectancies or predictions (such as violations in
study investigated the presence and absence of chord and harmonic expectancies), even in mu-
phrase boundaries in native and non-native sically untrained listeners.
music. This implicated the planum temporale
(PT), an integratory mechanism in the tempo- Music, Movement,
ral lobe just posterior to Heschl’s gyrus in the and Synchronization
center of Wernicke’s area. The PT was increas-
ingly activated as phrase boundaries became Humans are the only species capable of syn-
more difficult to identify in native music. PT has chronizing movement to sound (Patel 2007;
also been implicated in studies of absolute pitch Sacks 2007; although see Patel et al. in press,
(a larger leftward asymmetry is associated with for new data that suggest such behavior may be
AP possession, Keenan et al. 2001; Schlaug found in Cacatua galerita eleanora). Although other
et al. 1995) and it has been likened to a “com- animals (chimpanzees, elephants) can keep a
putational hub” (Griffiths & Warren 2002). steady tempo, when one animal is doing so, a
The extraction of phrase boundaries is an es- conspecific that joins in will not be able to keep
sential preparatory operation for memory en- the beat or play in time.
coding: In order for an event to be stored it The well-known association between music
needs to be temporally segmented into a be- and movement both behaviorally and neurally
ginning and end. The neural basis for such (across cultures and throughout history) sug-
event segmentation in musical phrase transi- gests an ancient evolutionary connection be-
tions was investigated using fMRI and found tween music and dance, or more generally, be-
to involve distinct, dissociable dorsal and ven- tween sound and movement. In fact, the motor
tral fronto-temporal structures (Sridharan et al. theory of speech of speech perception (Liber-
2007). In particular, a ventral fronto-temporal man 1982; Liberman & Mattingly 1985) argues
network, including the ventrolateral prefrontal that we learn to speak by observing the mouth
cortex (vlPFC, BA 47, and BA44/45) and pos- and lip movements of others. The recent dis-
terior temporal cortex (pTC, BA 21/22), was covery of mirror neurons (Rizzolatti et al. 1996)
218 Annals of the New York Academy of Sciences

and the evidence of their presence in Broca’s When listening to music many people report
area (Heiser et al. 2003; Johnson-Frey 2003; that it is difficult to avoid moving their bodies,
Lametti & Mattar 2006) suggests a plausible whether it is a simple head nod to the beat,
neuroanatomical substrate for the motor the- a body sway or a foot tap. This movement is
ory of speech perception—and the connection processed via the medial geniculate nucleus,
between music and dance. Listening to music a subcortical auditory relay station (Brown &
may activate mirror neurons that cause us to Parsons 2008); the absence of communication
think (at least unconsciously) about those motor to cortical structures following automatic, syn-
movements that would be required to make the chronous movement to music can therefore be
music. Dance can be conceived as an extension interpreted as biologically (as well as behav-
or complementary correlate of the movements iorally) unconscious (cf. Levitin et al. 2003;
required to create music. It has been widely ob- Levitin & Menon 2003). When young adults
served that infants are readily able to sing back were prompted to describe activities associated
melodies that they hear—taking input from one with the songs of their past, one of the most
sense (hearing) and producing output with an- common activities recalled was dancing (Janata
other sense (vocal−motor) seamlessly. Broca’s et al. 2007).
area may well be the seat of this ability as well. The connection between music and move-
If so, the connection between music and dance ment shows up also in studies of visual percep-
can be thought of as an extension of the move- tion of musical performances. Watching a mu-
ments required for vocalizing simply applied to sical performance, even with the sound turned
other body regions. The voluntary motion of off, conveys a great deal of structural and emo-
the limbs to music, which characterizes danc- tional information, further supporting evolu-
ing, activates the precuneus, a region of the tionary connections between music and move-
parietal lobe (Brown & Parsons 2008). ment (Chapados & Levitin 2008; Davidson
It is worth noting that music cannot exist 1993; Vines et al. 2005, 2006).
without movement. Because sound is transmit- The connection between auditory and kines-
ted via vibrating molecules, some physical mo- thetic senses was explored in a series of studies
tion is required to set those molecules vibrat- with both infants and adults (Phillips-Silver
ing in the first place—hitting, plucking, bowing, & Trainor 2005, 2007, 2008). Participants ei-
blowing, or forcing air through the vocal cords ther bounced themselves (adults) or had them-
(Levitin et al. 2002). Even when lying perfectly selves bounced (infants) to an unaccented
still, listeners in fMRI studies show activation rhythm either in a duple (march) or triple
in those regions of the brain that would nor- (waltz) meter. The meter biased the percep-
mally orchestrate motor movement to music, tual representation and subsequent recall of
including the cerebellum, basal ganglia, and the sequences. In effect, the movement itself
cortical motor areas—it is as though movement created the (cross-modal) accented beat. This
is impossible to suppress (Levitin 2008; Levitin interactive process was found to be mediated
& Menon 2003). Tapping in synchrony to the by the vestibular system: Although full body
pulse of a musical sequence (by humans) en- movement is the most effective in engender-
gages the presupplementary motor area, the ing the movement−sound interaction, head
supplemental motor area, the dorsal premo- movement alone is capable of producing it,
tor cortex, the dorsolateral prefrontal cortex, while body movement alone is not (Phillips-
the inferior parietal lobule, and lobule VI of Silver & Trainor 2008). Additional data im-
the cerebellum, as measured by the BOLD re- plicate the dorsal premotor cortex (dPMC) in
sponse (Chen et al. 2008). rhythmic synchronization. Participants tapped
A generalized motor theory can account for to rhythmic sequences of varying levels of met-
this connection between sound and movement. ric difficulty; greater difficulty was correlated
Levitin & Tirovolas: Cognitive Neuroscience of Music 219

with increased dPMC activation (Chen et al. medial midbrain, and deactivation in the amyg-
2008). dala (Blood & Zatorre 2001). Opioid trans-
mission in the NAc has been associated with
dopamine release in the ventral tegmental area
Emotion (VTA) (Kelley & Berridge 2002), and together
they are involved in mediating the brain’s re-
Music represents a dynamic form of emo- sponses to reward. During music listening the
tion (Dowling & Harwood 1986; Helmholtz VTA mediates activity in the NAc, hypothala-
1863/1954; Langer 1951). The conveying of mus, insula, and orbitofrontal cortex; this net-
emotion is considered to be the essence if not work represents the neural and neurochemi-
the purpose of music (Meyer 1956; Nietzsche cal (via dopaminergic pathways) underpinnings
1871/1993) and the reason that most people re- of the anecdotal reports of pleasurable music
port spending large amounts of time listening (Menon & Levitin 2005). In addition, the hip-
to music (Juslin & Sloboda 2001). Somewhat pocampus has been found in positron emis-
paradoxically, the cognitive and structural as- sion tomography (PET) studies to activate dur-
pects of music have been the most extensively ing pleasant music, and the parahippocampul
studied, perhaps because methods of studying gyrus, also implicated in emotion processing,
them have been part of the standard cognitive has been found to activate during dissonant
psychology paradigms for decades. Advances in music (Koelsch et al. 2006). This network of
affective neuroscience as well as new links be- structures, which includes the amygdala and
tween neurochemistry and cognition have only the temporal poles, is thought to be the neu-
recently made it possible to study emotion in rological basis for the emotional processing of
music rigorously (Blood & Zatorre 2001; Blood music (Koelsch et al. 2006).
et al. 1999; Panksepp 2003). Complementary to the study of the neuro-
Historically, studies in affective neuroscience logical underpinnings of chills in response to
have focused almost exclusively on the process- music is a recent study on the physiological and
ing of negative emotions (LeDoux 2000). The psychological aspects, as well as the character-
few existing studies of positive emotions have istics of the music that engenders this emotion-
tended to use drugs of addiction to induce pos- ally driven response (Grewe et al. 2007). Psy-
itive emotions artificially (Berridge 2003), and chologically, individuals who experience chills
only recently have more naturalistic and eco- are not necessarily thrill-seekers; they tend to-
logically valid studies of positive emotion been ward greater sensitivity to sensory stimulation.
conducted (Kringlebach et al. 2003; Small et al. Those who experience chills are more likely to
2001). Listening to classical music is known depend on rewards from the environment, in a
to evoke strong emotions, including feelings of sense being more vulnerable to the response’s
pleasure (Krumhansl 1997; Sloboda & Juslin occurrence; are highly familiar with classical
2001). Further, this experience is often accom- music (a genre included as part of the stimuli
panied by physical responses (Panksepp 1995), in the experiment); identify strongly with their
such as thrills, chills, shivers, and changes in musical preferences; and often listen to music in
heart rate that can be blocked by nalaxone, isolation. Psychoacoustically, there was no spe-
a known opioid antagonist (Goldstein 1980). cific pattern that emerged in most of the chill-
The experience of pleasant, or consonant, mu- inducing excerpt, but a small portion included
sic activates orbitofrontal, subcallosal cingulate, peaks in loudness, sharpness, and fluctuation.
and frontal polar cortical areas (Blood et al. The contextual aspects of what induced chills
1999). Chills have been shown to correlate with were the entry of a voice, loudness, the entrance
activity in the left ventral striatum, an area “re- of a specific theme, and the auditory experience
sponsible for” approaching reward, the dorso- of two contrasting voices. These are the unique
220 Annals of the New York Academy of Sciences

musical contexts considered by the researchers example, the amygdala shows increased acti-
to represent increases in attention, bringing on vation when music is presented concurrently
an emotional experience, of which the physi- with an audio-visual stimulus providing context
cal reaction of a chill is a consequence (Grewe (Eldar et al. 2007). No such activation is found
et al. 2007). The listener is thus considered to when positive or negative music is presented
be an active participant in not only responding, alone, suggesting that real-world context aids
but creating an emotional experience with mu- in building a more meaningful emotional rep-
sic through attention, leading to a chill as an resentation, capable of differentially engaging
induced side effect. the amygdala. Presumably, the adaptive qual-
Many listeners report using music for mood ity of the amygdala (central to the mammalian
regulation and may find comfort in sad music fear and avoidance, fight-or-flight network) is
(Chamorro-Premuzic & Furnham 2007). One increased by the corroboration of a potential
might assume that sad people would be up- danger from another sensory modality.
lifted by happy music, but this is not always The neuroanatomical substrates of emo-
the case. Huron (2006) offers an explanation. tion regulation in music were studied in a
Prolactin, a tranquilizing and consoling hor- group of postoperative epileptics (Khalfa et al.
mone, is produced by the anterior pituitary 2008) with temporal lobe resections (including
gland when we’re sad (Panksepp 2006). The the amygdala). Patients with right-hemisphere
evolutionary purpose of sorrow is to aid in en- resection showed reduced recognition of sad
ergy conservation and allow for reassessment music (and overidentification of happy music)
of priorities for the future following a traumatic while patients with left-hemisphere resections
event. Prolactin is released after orgasm, after showed reduced recognition of both happy and
birth, and during lactation in females. A chem- sad music. These findings must be interpreted
ical analysis reveals that prolactin is not always with caution because the experiment did not
present in tears—it is not released in tears of evaluate the preservation of lower level percep-
lubrication of the eye, or when the eye is irri- tual function in the patients following surgery;
tated, or in tears of joy; it is only released in that is, pitch or contour deficits could con-
tears of sorrow. Huron speculates that sad mu- ceivably underlie the participants’ judgment of
sic allows us to “trick” our brain into releasing emotion in the music.
prolactin in response to the safe or imaginary There exists a widespread belief in Western
sorrow induced by the music, and the prolactin culture that major keys are intrinsically associ-
then reverses our mood. Aside from the neu- ated with positive affect while minor keys are
rochemical story, there is a more psychological related to negative affect. This turns out to
or behavioral explanation for why we find sad be largely a product of exposure and learning
music consoling. When people feel sad or suffer and is thus culturally dependent. It has been
from clinical depression, they often sense being shown that other musical systems (e.g., Mid-
cut off from other people, feeling as though dle Eastern, Indian) do not share these asso-
no one understands them. Happy music can ciations (Balkwill & Thompson 1999; Balkwill
be especially irritating because it makes them et al. 2004; Trehub 2003).
feel even less understood. Sad music is consol- Consistent with findings on state-dependent
ing because it connects the listener to others memory (see Bower 1981), mood affects mem-
who seem to be experiencing a similar affective ory for music played in different modalities.
state. It has been reported that when induced with a
As the field of music cognition advances, its positive mood, Western listeners are more likely
investigators are acquiring deeper, empirically to recognize a melody played in a major key
driven understanding of the complexity of emo- than a minor key (Houston & Haddock 2007),
tion, manifested as a contextual process. For indicating that strong associations are made to
Levitin & Tirovolas: Cognitive Neuroscience of Music 221

music that is congruent with (culturally defined) ceptions to auditory transmission include that
transient mood states. both can be written down, felt through bone
Happy classical music (mentioned above conduction or other tactile means, and that
in the context of strong feelings of pleasure lip readers can understand speech without an
and chills) has been associated with activity auditory signal.) In both music and language
in the bilateral ventral and left dorsal stria- the sensory input evolves over time in a co-
tum, left anterior cingulate, left parahippocam- herent structure. Theories of structure in both
pul gyrus, and auditory association areas in domains are in fact theories of temporal co-
listeners unselected for musical background herence and how elements are grouped over
(Mitterschiffthaler et al. 2007). In contrast, the time (Cooper & Meyer 1960; Cooper & Paccia-
emotional state induced by sad music activates Cooper 1980; Krumhansl 1990, 1991; Lerdahl
the hippocampus/amygdala and auditory as- 2001; Lerdahl & Jackendoff 1983; Patel 2007;
sociation areas; emotionally neutral classical West et al. 1985; Zatorre et al. 2002). A sec-
music is processed in the insula and auditory ond parallel concerns the specific order of con-
association areas. stituents in revealing meaning. The sentences
Electrophysiological data confirm that mu- of all human languages (spoken and signed) are
sicians process the emotional content of mu- composed of words in a certain linear order
sic, as indexed by its mode, differently than (Akmajian et al. 1980). Although some lan-
do nonmusicians. Specifically, when perceiving guages display considerable freedom of word
melodies that are similar in length, tempo, and order, in no human language may the words
rhythm, but different in mode, musicians dis- of a sentence occur in a random order. This is
play a late positive component ERP (P3) to also the case with music: Musical phrases are
the onset of a note in minor mode melodies composed of notes and/or chords, but these
(Halpern et al. 2008). Interestingly, neither mu- are not randomly ordered and a reordering of
sicians nor nonmusicians showed a late positive elements produces a different melody (Lerdah
component to music in the major mode. The 2001; Patel 2003).
authors argue that the absence of this effect in Based on the existence of such commonali-
musicians is likely due to an enculturation ef- ties, Patel (2003) introduced the “shared syntac-
fect of the preponderance of music in the major tic integration resource hypothesis” (SSIRH),
mode. As a consequence, the minor mode acts which proposes that syntax in language and
as an oddball stimulus, which requires addi- music share a common set of circuits instanti-
tional information processing. ated in frontal brain regions. SSIRH is based
as well on empirical findings implicating frontal
regions in the processing of harmonic structure
Music and Language (Janata et al. 2002; Tillman et al. 2003) and, in
particular, the processing of harmonic anoma-
The last several years have seen an increased lies (Koelsch et al. 2002; Maess et al. 2001).
focus on studies of music and spoken language, Evidence for the SSIRH comes from sev-
due in part to advances in digital recording eral studies that co-locate musical and lin-
and signal processing technology, and to the guistic operations. When musical structure is
increased recognition that music and language disrupted, areas of the brain implicated in lin-
both represent complex, higher-order cogni- guistic syntax—Brodmann area (BA) 47 and
tive processes that invoke a large number of the adjoining anterior insula—play a role in the
subsystems including attention, categorization, perception of that disruption (Levitin & Menon
memory, and feature detection. Music and lan- 2003, 2005). Violations of musical expectations
guage share many attributes. Both are primarily also invoke BA 47 (Koelsch et al. 2000) and
auditory-based forms of communication. (Ex- Broca’s area (Koelsch et al. 2002).
222 Annals of the New York Academy of Sciences

An additional link between language and meaning in music and reinforces the idea that
music comes from an experiment with chil- language and music are based on shared neural
dren with specific language impairment (SLI) underpinnings.
(Jentschke et al. 2008). Four- and 5-year olds Most spoken languages employ pitch vari-
with SLI presented a particular ERP pattern ation (as a component of linguistic prosody)
when they listened to the final chord in a se- to convey meaning and to disambiguate utter-
quence that violated harmonically lawful musi- ances. In tonal languages (such as Thai, Man-
cal syntax. This response pattern deviates from darin, and Cantonese) pitch variation within
that of children and adults who develop lan- a word can completely alter the meaning of the
guage typically, in that the SLI response elicited word (e.g., different pitch trajectories for the
did not include an early right anterior negativ- word /ma/ in Cantonese can give the word
ity (ERAN) or the N5. Jentschke and colleagues the meanings “mother” or “gun powder”). The
(2008) further suggest that musical training may extent to which bona fide musical operations
be a means of early intervention for children at are involved in processing tonal languages are a
risk for developing SLI. topic of current interest. Western, nontonal lan-
Music theorists and philosophers at least guage speakers were played excerpts from tonal
from the time of Nietzsche have struggled with languages. The processing of pitch information
the question of whether music has meaning in those tonal language utterances was found
apart from the formal, syntactic structures that to be more accurately coded by musicians than
constitute the rules of a musical system or style. nonmusicians, as measured by pitch encoding
To invoke parallels with distinctions made by at the subcortical level of the inferior colliculus
linguists for the analysis of natural languages (Wong et al. 2007), perhaps associated with mu-
(e.g., Fromkin & Rodman 1993), it has been sicians increased usage or attention to the pitch
suggested that all musical semantics are insep- attributes of an auditory signal. This is the first
arable from considerations of musical pragmat- study to demonstrate superior subcortical pitch
ics (Ashley 2007). In language, pragmatics refers processing in musicians (although the direction
to the level of analysis concerned with how peo- of causality is unknown—do people with supe-
ple actually use sentences and the intentions of rior subcortical pitch processing become mu-
the speaker apart from the actual words used sicians? or does musical experience exert this
(irony and sarcasm fall in this domain). Ash- effect?). A conceptually related study examined
ley argues that in effect musical semantics or the ability of French musicians versus nonmu-
meaning derives from musical pragmatics or sicians to detect prosodic variation in a struc-
intentions. Nevertheless, a recent study found turally similar and related language, Portuguese
neural evidence for two distinct processes, one (Marques et al. 2007). Although musicians and
syntactic and the other semantic (which Ashley nonmusicians were equally capable of detect-
would interpret as semantic−pragmatic). ing strong prosodic incongruities in foreign sen-
The N5 electrophysiological response is as- tences, musicians were significantly better at
sociated with failures of a musical sequence identifying the more subtle incongruities.
to meet harmonic expectations (Koelsch et al. A number of studies have examined mu-
2000). Steinbeis and Koelsch (2008) reasoned sic and language by studying children during
that if the N5 were in fact elicited by processes language acquisition. There is evidence that
governing meaning, a simultaneously presented low-level auditory processing at the level of the
linguistic−semantic anomaly ought to reduce brain stem is related to literacy skills in chil-
the N5 but not the ERAN. This was in fact dren; those individuals who respond to speech
what they found (see also Koelsch et al. 2007). sounds in an early or intermediate fashion dis-
This result indicates that the structure of mu- play higher achievement in reading than those
sic itself can be one path to the construction of individuals who are delayed in their responses
Levitin & Tirovolas: Cognitive Neuroscience of Music 223

(Abrams et al. 2006; Banai et al. 2005; Kraus music (Shuter-Dyson 1999). There is consid-
& Banai 2007). Although it stands to reason erable debate among scientists, musicians, and
that auditory processing in the linguistic do- the population at large as to whether musicality
main would be related to literacy, there are also is based on talent, experience, or some combi-
recent findings that link musical discrimination nation of both (Howe et al. 1998). A compli-
abilities to reading ability. In a series of stud- cating factor is that “musicality” can manifest
ies with school-aged children, Forgeard et al. itself in diverse—and sometimes nonoverlap-
(2008) found that the ability to discriminate ping forms: One can be expert in composi-
melodies is predictive of phonological skills, tions, performances, improvisations, listening,
particularly phonemic awareness (sensitivity to editing, etc. Within a given subdomain, exper-
the sounds of language), which is a prerequisite tise can exist primarily for rhythm, pitch, or
for reading ability. Moreover, they find that chil- timbre. Despite these diverse definitions, cer-
dren with specific reading disability (dyslexia) tain individuals describe themselves as “musi-
are impaired in both melodic and rhythmic cal,” while others do not.
tasks, indicating impairment that extends be- As we described above, musical processes
yond melodic discrimination to a more perva- can be parsed into different components which
sive deficit in music processing. are subserved by different areas of the brain.
Second language pronunciation ability in This leads to the prediction that brain damage
children was positively associated with musi- should selectively impair only affected compo-
cal aptitude test. In addition, greater ERP ac- nents. Dozens of case studies have borne this
tivation to mistuned variants of musical chords out (e.g., Ayotte et al. 2000; Di Pietro et al.
was observed in those with better pronunci- 2004; Peretz et al. 2004; Piccirilli et al. 2000).
ation skills (Milovanov et al. 2008). Although Sensitivity to pitch and time in music are
common neural underpinnings between music considered fundamental to musical adeptness.
and language constitute a likely explanation of Traditionally, tone deafness was conceived as a
the congruence between linguistic and musical selective impairment in fine-grained pitch per-
skills in the domains of language and music, ception held by an estimated 4% of the popu-
other factors possibly accounting for the rela- lation (Cox 1947; Joyner 1968). One definition
tionship noted by Milovanov and colleagues states that individuals who are tone deaf are not
(2008) are executive functions (such as atten- able to discriminate pitches less than one semi-
tion), as well as the maturity level of the tem- tone apart, but do not show associated deficits
poral lobes, sensitivity to the musicality of lan- in time perception (Hyde & Peretz 2004). A
guage, and basic sound processing. more nuanced definition of amusia/tone deaf-
ness was proposed to include selective impair-
ments in perception (of rhythm or melody, and
Amusia perhaps not both), in production, and in song
identification, arising from several distinct eti-
The term amusia is generally applied to indi- ologies (Levitin 1999). A dissociation between
viduals with a supposed deficit in one or more the perception and production supports this
aspects of music processing. The lay term tone (Loui et al. 2008). Also consistent with the ex-
deafness is seen as an equivalent. Scientists now panded definition is recent evidence that amu-
make a distinction between acquired amusias sia is not isolated to difficulties in pitch per-
(typically following brain injury) and congenital ception alone. Compared to a control group,
amusia. The term musicality has been described individuals with amusia perform significantly
in several ways, from a universal human at- worse on a mental rotation task, when con-
tribute to the ability to attain high levels of musi- trolling for sex differences in spatial processing
cal expertise, or even simply the ability to enjoy (Douglas & Bilkey 2007). Similarly, enhanced
224 Annals of the New York Academy of Sciences

spatial capacities were found in orchestral mu- Patel et al. 2008) finds impairment in differ-
sicians (Sluming et al. 2007). The mechanism entiation of statements and questions among
driving improved spatial ability is argued to be amusic participants. A follow-up study to clarify
related to experience in the sight reading of these inconsistencies by the 2008 study of Pa-
music (musical notation requires greater spatial tel et al. confirmed that approximately 30% of
acuity than text reading). individuals in a British and French-Canadian
The Montreal Battery of Evaluation of Amu- sample had difficulty in sentence−statement
sia (Peretz et al. 2003) is a test designed to screen differentiation. A possible confound in this
for acquired amusia (resulting from brain in- study is that the stimuli presented to the British
jury) or congenital amusia (tone deafness). The listeners were recorded in American English
tasks in the battery include tests of pitch, time while the stimuli for the French-Canadian lis-
and memory such as scale discrimination, con- teners were recorded in continental French,
tour, rhythm, and meter. Recently, Peretz et al. neither stimulus representing ecologically valid
(2008) created an online version of the MBEA, stimuli for the participants.
which takes only 15–30 minutes to complete, Individuals with congenital amusia also re-
composed of less than half the trials as the orig- port fundamentally different experiences with
inal MBEA. The test is seen as a reasonably music in their daily lives. In particular, con-
precise and efficient diagnostic tool for congen- trols listened to approximately 3 times as much
ital amusia, and the authors claim only a 7% music per week than individuals with amusia,
error rate in detection. Particular subtests of less than half of an amusic sample claims to
the online version have been used to screen for like or love music or use music in the context
individuals with potential amusia before bring- of setting the mood in a romantic encounter
ing these individuals into the laboratory for a (McDonald & Stewart 2008). The most com-
full diagnostic with the original MBEA, as in mon psychological state induced by music in
McDonald and Stewart (2008). the amusics was nostalgia; however the major-
Neurologically, there do appear to be differ- ity attributed the nostalgia to cultural or lyrical
ences in the amusic brain, relative to controls. associations to the music, rather than the music
Individuals with congenital amusia have thicker itself. Some individuals with amusia report an
cortex in the right inferior frontal gyrus (IFG) extremely negative form of arousal, described
and the right auditory cortex (Hyde et al. 2007). as aural irritation.
This morphological difference at the cortical
level is attributed to atypical cortical develop-
ment, affecting the right frontotemporal tract Expertise
known to play a role in the processing of musi-
cal pitch. Research in the domain of expertise com-
Evidence is mixed as to whether individuals prises studies of how musical experience affects
with amusia have deficits that extend beyond practice and performance, as well as evidence
the musical into the linguistic domain. For ex- for skill transfer from music to other cogni-
ample, Ayotte et al. (2002) found that individu- tive domains. Several studies attempt to shed
als with congenital amusia suffer from domain- light on the similarities and differences between
specific musical impairment, consistent with the cognitive and auditory−perceptual pro-
the view of modularity in music processing cessing capabilities of trained musicians versus
(Peretz & Coltheart 2003), as their participants nonmusicians.
were unimpaired in a variety of linguistic tasks, Historically, it was believed that auditory per-
including the processing of prosody in human ception was largely the result of automatic,
speech. However, a study of French Cana- bottom-up processes. It is now believed that au-
dian amusics by Lochy and colleagues (cited in ditory object formation and music perception
Levitin & Tirovolas: Cognitive Neuroscience of Music 225

are the consequence of dynamic, interrelated scious control represents overlearned and au-
processes involving cortical and subcortical re- tomatic processes characteristic of professional
gions, reflecting both bottom-up and top-down improvisers.
influences (Kraus & Banai 2007). In particu- The most striking and obvious example of
lar, the auditory system has shown itself to be top-down influences on auditory perception
plastic: What we think of as “auditory” cor- comes from the phoneme acquisition trajec-
tex can become remapped for visual input in tory in human babies. Born with the ability
the congenitally deaf (Neville et al. 1998; Pe- to discriminate all possible speech sounds, they
titto et al. 2000). Musicians’ better sensitivity to eventually retain only those distinctions that
foreign tonal languages (as mentioned above, are necessary for the language to which they
Wong et al. 2007) bolsters the argument that are exposed during a certain critical or sen-
experience can exert an influence on low-level sitive period (Kuhl 2004). Additional support-
processing, even as far downstream as the infe- ing evidence for the role of top-down influence
rior colliculus and brainstem. in auditory perception comes from the finding
Automaticity does not necessarily have that musicians’ brains show greater sensitivity
to entail purely bottom-up processing, how- to the sounds of their own instruments versus
ever. A recent ERP study (Brattico et al. others (Pantev et al. 2001). An electrophysio-
2006) found evidence that pitch processing— logical study determined that when musicians
specifically recognition of “in-tuneness” and (violinists in particular) listen to sounds of their
“in-keyness”—are automatic, preattentive pro- own instrument, gamma band activity specific
cesses reflecting overlearning of culturally de- to the timbre, or sound quality, of that instru-
pendent knowledge. In other words, attention ment is elicited (Shahin et al. 2008). The re-
is not required for recognizing violations of cer- sponse also occurs for piano timbre in children
tain tonal expectations, as indexed by an early after only one year of piano lessons.
frontal negativity in the ERP signal. Automatic Musical practice also enhances phase-
processing of pitch relations in the diatonic locking in the brain stem to the fundamental
scale may mean that neural networks have ac- frequencies of both musical and linguistic stim-
culturated themselves to that scale. uli (Musacchia et al. 2007). When participants
Another example of automatic processing in were presented with both audio and audiovi-
music is found in expert improvising musicians. sual stimuli in the domains of music and speech,
Jazz musicians creating spontaneous musical musicians had an earlier onset and larger am-
performances, or improvisations, were studied plitude evoked brainstem response than non-
using fMRI (Limb & Braun 2008). One might musicians. Because this response is a function
naı̈vely assume that improvisation requires fo- of the amount of training as opposed to musical
cal activation in that region of the brain that aptitude or basic pitch discrimination tasks, this
is uniquely developed in humans—the pre- finding gives empirical support for the saying,
frontal cortex (Gazzaniga 2008). In fact, strong “practice makes perfect.” A number of studies
patterns of deactivation were observed there, show regional changes in brain volume and in
suggesting that conscious thought and volition gray-to-white matter density as a function of
needed to be suppressed. Activation was ob- musical practice (see Münte et al. 2002 for a
served in neocortical sensory−motor areas that review).
mediate organization and execution of musical Musicians show additional activation in mo-
performance. This pattern of activation con- tor areas compared to nonmusicians during
forms to subjective reports by musicians that rhythm perception (Grahn & Brett 2007).
improvisation relies on sub- or preconscious Rhythm production in musicians produces
processes that are outside the domain of con- greater activation in the dorsolateral right pre-
scious control and awareness. The lack of con- frontal cortex (dLPFC) as well as right inferior
226 Annals of the New York Academy of Sciences

frontal gyrus, regions shown to subserve work- References

ing memory for music (Chen et al. 2008). This
difference is thought to represent a more ad- Abrams, D. A., Nicol, T., Zecker, S. G., et al. (2006). Audi-
vantageous cognitive strategy for reproducing tory brainstem timing predicts cerebral asymmetry
rhythm, which relies on top-down as opposed for speech. J. Neurosci., 26, 11131–11137.
Akmajian, A., Demers, R. A., & Harnish, R. M. (1980)
to bottom-up processing.
Overcoming inadequacies in the message-model of
A compelling recent finding in the area of linguistic communication. Communication and Cogni-
transfer effects is the relatively high perfor- tion, 13, 317–336.
mance on a three-dimensional mental rota- Ashley, R. (2007). Extracting information from musical scores.
tion task by orchestral musicians, compared to Paper presented at the Annual Meeting of the Society
others who were highly practiced at that task for Music Perception and Cognition, Montreal, QC.
Ayotte, J., Peretz, I., & Hyde, K. (2002). Congenital amu-
(Sluming et al. 2007). Behaviorally, the mu- sia: A group study of adults afflicted with a music-
sicians attempted more trials, achieved more specific disorder. Brain, 125, 238–251.
correct answers (at par with individuals who Ayotte, J., Peretz, I., Rousseau, I., et al. (2000). Patterns of
have had extensive practice), and displayed a music agnosia associated with middle cerebral artery
response time profile uncharacteristic of the infarcts. Brain, 123, 1926–1938.
Balkwill, L. L., & Thompson, W. F. (1999). A cross-cultural
typical profile for the task. Neurofunctionally,
investigation of the perception of emotion in music:
the orchestral musicians displayed greater acti- Psychophysical and culture. Music Perception, 17, 43–
vation in Broca’s area (BA 44/45) compared to 64.
the control group during task performance, an Balkwill, L. L., Thompson, W. F., & Matsunaga, R. (2004).
area known to subserve music sight reading. Recognition of emotion in Japanese, Western, and
Hindustani music by Japanese listeners. Japanese Psy-
cholog. Res., 46, 337–349.
Banai, K., Nicol, T., Zecker, S. G., et al. (2005). Brain-
Summary stem timing: Implications for cortical processing and
literacy. J. Neurosci., 25, 9850–9857.
Bendor, D., & Wang, X. Q. (2005). The neuronal repre-
The study of music cognition, and more re- sentation of pitch in primate auditory cortex. Nature,
cently of music and the brain, has a rich and 436, 1161–1165.
old intellectual history. We now know that mu- Berridge, K. C. (2003). Pleasures of the brain. Brain and
sical operations involve disparate regions of the Cognition, 52, 106–128.
brain, including all lobes of the brain, and both Bever, T. G., & Chiarello, R. J. (1974). Cerebral domi-
nance in musicians and nonmusicians. Science, 185,
cortical and subcortical structures. In particu- 537–539.
lar, the roles of the cerebellum and amygdala Blood, A. J., & Zatorre, R. J. (2001). Intensely pleasurable
are becoming increasingly appreciated. The responses to music correlate with activity in brain
components of music, including pitch, rhythm, regions implicated in reward and emotion. Proc. Nat.
contour, and timbre, are subserved by distinct Acad. Sci. USA, 98, 11818–11823.
Blood, A. J., Zatorre, R. J., Bermudez, P., et al. (1999).
and separable neural processing units. Music
Emotional responses to pleasant and unpleasant mu-
processing shares some circuitry with spoken sic correlate with activity in paralimbic brain regions.
language processing yet also involves distinct Nat. Neurosci., 2, 382–387.
neural circuits. The study of the neuroanatomi- Bower, G. H. (1981). Mood and memory. Amer. Psychologist,
cal underpinnings of musical emotion is an area 36, 129–148.
of particularly recent focus, as are the putative Brattico, E., Tervaniemi, M., Näätänen, R., et al. (2006).
Musical scale properties are automatically processed
evolutionary origins of the music faculty. in the human auditory cortex. Brain Res., 1117, 162–
Bregman, A. S. (1990). Auditory Scene Analysis: The Perceptual
Conflicts of Interest Organization of Sound. Cambridge, MA: MIT Press.
Brown, S., & Parsons, L. M. (July 2008). The neuroscience
The authors declare no conflicts of interest. of dance: Recent brain-imaging studies reveal some
Levitin & Tirovolas: Cognitive Neuroscience of Music 227

of the complex neural choreography behind our abil- ing in normal-reading children and children with
ity to dance. Scientific American, 78–83. dyslexia. Music Percep., 25, 383–390.
Carroll, J. (1998). Steven Pinker’s cheesecake for the mind. Fromkin, V., & Rodman, R. (1993). An Introduction to Lan-
Phil. Lit., 22, 478–485. guage. Fort Worth, TX: Harcourt Brace Jovanovich.
Chamorro-Premuzic, T., & Furnham, A. (2007). Person- Gazzaniga, M. S. (2008). Human: The science behind what
ality and music: Can traits explain how people use makes us unique. New York: Ecco.
music in everyday life? Brit. J. Psych., 98, 175–185. Goldstein, A. (1980). Thrills in response to music and
Chapados, C., & Levitin, D. J. (2008). Cross-modal other stimuli. Physiolog. Psych., 8, 126–129.
interactions in the experience of musical perfor- Grahn, J. A., & Brett, M. (2007). Rhythm and beat per-
mances: Physiological correlates. Cognition, 108, 639– ception in motor areas of the brain. J. Cogn. Neurosci.,
651. 19, 893–906.
Chen, J. L., Penhune, V. B., & Zatorre, R. J. (2008). Mov- Green, A. C., Baerensten, K. B., Stodkilde-Jorgensen, H.,
ing on time: Brain network for auditory-motor syn- et al. (2008). Music in minor activates limbic struc-
chronization is modulated by rhythm complexity and tures: A relationship with dissonance? NeuroReport,
musical training. J. Cogn. Neurosci., 20, 226–239. 19, 711–715.
Chomsky, N. (1965). Aspects of the Theory of Syntax. Cam- Grewe, O., Nagel, F., Kopiez, R., et al. (2007). Listening
bridge, MA: MIT Press. to music as a re-creative process: Physiological, psy-
Cooper, G. W., & Meyer, L. B. (1960). The Rhythmic Struc- chological, and psychoacoustical correlates of chills
ture of Music. Chicago: University of Chicago Press. and strong emotions. Music Percep., 24, 297–314.
Cooper, W. E., & Paccia-Cooper, J. (1980). Syntax and Griffiths, P. (2004). New Penguin Dictionary of Music. London:
Speech. Cambridge, MA: Harvard University Press. Penguin Books.
Cosmides, L., & Tooby, J. (1995). From function to struc- Griffiths, T. D., & Warren, J. D. (2002). The planum tem-
ture: The role of evolutionary biology and compu- porale as a computational hub. Trends in Neuroscience,
tational theories in cognitive neuroscience. In M. S. 25, 348–353.
Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 1199– Halpern, A., Martin, J. S., & Reed, T. D. (2008). An ERP
1210). Cambridge, MA: MIT Press. study of major-minor classification in melodies. Music
Cox, I. (1947). Tone deafness. Music Edu. J., 34, 62–63. Percep., 25, 181–191.
Cross, I., & Woodruff, G. E. (in press). Music as a com- Halsband, U., Ito, N., Tanji, J., et al. (1993). The role of
municative medium. To appear in R. Botha, & C. premotor cortex and the supplementary motor area
Knight (Eds.), The Prehistory of Language, Oxford: Ox- in the temporal control of movement in man. Brain,
ford University Press. 116, 243–266.
Davidson, J. (1993). Visual perception of performance Hannon, E. E., & Trehub, S. E. (2005). Metrical categories
manner in the movements of solo musicians. Psych. in infancy and adulthood. Psycholog. Sci., 16, 48–55.
Music, 21, 103–113. Hanson, H. M. (1959). Effects of discrimination training
Di Pietro, M., Laganaro, M., Leeman, B., et al. (2004). Re- on stimulus generalization. J. Exp. Psych., 58, 321–
ceptive amusia: Temporal auditory deficit in a pro- 334.
fessional musician following a left temporo-parietal Hauser, M., & McDermott, J. (2003). The evolution of
lesion. Neuropsychologia, 42, 868–977. the music faculty: A comparative perspective. Nat.
Douglas, K. M., & Bilkey, D. K. (2007). Amusia is associ- Neurosci., 6, 663–668.
ated with deficits in spatial processing. Nat. Neurosci., Heiser, M., Iacoboni, M., Maeda, F., et al. (2003). The
10, 915–921. essential role of Broca’s area in imitation. Eur. J.
Dowling, W. J. (1978). Scale and contour: Two compo- Neurosci., 17, 1123–1128.
nents of a theory of memory for melodies. Psycholog. Helmholtz, H. L. F. (1863/1954). On the sensations of tone
Rev., 85, 341–354. (A. J. Ellis, Trans.). New York: Dover Publications,
Dowling, W. J., & Harwood, D. L. (1986). Music Cognition. Inc.
San Diego: Academic Press. Houston, D., & Haddock, G. (2007). On auditing auditory
Eldar, E., Ganor, O., Admon, R., et al. (2007). Feeling information: The influence of mood on memory for
the real world: Limbic response to music depends on music. Psych. Music, 35, 201–212.
related content. Cerebral Cortex, 17, 2828–2840. Howe, M. J. A., Davidson, J. W., & Sloboda, J. A. (1998).
Fechner, G. T. (1860). Elemente der Psychophysik. Leipzig: Innate talents: Reality or myth? Behavioral and Brain
Breitkopf und Härtel. Sciences, 21, 399–442.
Fodor, J. (1983). The Modularity of Mind. Cambridge, MA: Huron, D. (2006). Sweet Anticipation: Music and the Psychology
MIT Press. of Expectation. Cambridge, MA: MIT Press.
Forgeard, M., Schlaug, G., Norton, A., et al. (2008). The Hyde, K. L., Lerch, J. P., Zatorre, R. J., et al. (2007).
relation between music and phonological process- Cortical thickness in congenital amusia: When less
228 Annals of the New York Academy of Sciences

is better than more. J. Neurosci., 27(47), 13028– serves the processing of music. Neuroimage, 17, 956–
13032. 966.
Hyde, K. L., & Peretz, I. (2004). Brains that are out of Koelsch, S., Jentschke, S., Sammler, D., et al. (2007). Un-
tune but in time. Psychological Science, 15, 356–360. tangling syntactic and sensory processing: An ERP
Ivry, R. B., & Hazeltine, R. E. (1995). Perception and study of music perception. Psychophysiol., 44, 476–
production of temporal intervals across a range of 490.
durations: Evidence for a common timing mecha- Koelsch, S., Maess, B., & Friederici, A. D. (2000). Musical
nism. J. Exp. Psych.—Hum. Percep. Perform., 21, 3–18. syntax is processed in the area of Broca: An MEG
Ivry, R. B., & Schlerf, J. E. (2008). Dedicated and intrinsic study. NeuroImage, 11(5 Supplement 1), S56.
models of time perception. Trends Cogn. Sci., 12, 273– Kohler, W. (1938). Simple structural function in the chim-
280. panzee and the chicken. In W. D. Ellis (Ed.), A Source-
Ivry, R. B., & Keele, S. W. (1989). Timing functions of book of Gestalt Psychology (pp. 217–227). New York:
the cerebellum. J. Cogn. Neurosci., 1, 136–152. Harcourt, Brace & World. (Original work published
Janata, P., Birk, J. L., Horn, J. D. V., et al. (2002). The in German in 1918).
cortical topography of tonal structures underlying Kraus, N., & Banai, K. (2007). Auditory-processing mal-
Western music. Science, 298, 2167–2170. leability. Curr. Dir. Psycholog. Sci., 16, 105–110.
Janata, P., & Grafton, S. T. (2003). Swinging in the brain: Kringelbach, M. L., O’Doherty, J., Rolls, E. T., et al.
Shared neural substrates for behaviors related to se- (2003). Activation of the human orbitofrontal cortex
quencing and music. Nat. Neurosci., 6, 682–687. to a liquid food stimulus is correlated with its subjec-
Janata, P., Tomic, S. T., & Rakowski, S. K. (2007). Char- tive pleasantness. Cerebral Cortex, 13, 1064–1071.
acterisation of music-evoked autobiographical mem- Krumhansl, C. L. (1997). An exploratory study of musical
ories. Memory, 15(8), 845–860. emotions and psychophysiology. Canad. J. Exp. Psych.,
Jentschke, S., Koelsch, S., Sallat, S., et al. (2008). Children 51, 336–353.
with specific language impairment also show impair- Krumhansl, C. L. (1990). Cognitive Foundations of Musical
ment of music-syntactic processing. J. Cogn. Neurosci., Pitch. New York: Oxford University Press.
20, 1940–1951. Krumhansl, C. L. (1991). Music psychology: Tonal struc-
Johnson-Frey, S. H. (2003). Mirror neurons, Broca’s area tures in perception and memory. Annual Rev. Psych.,
and language: Reflecting on the evidence. Behavior. 42, 277–303.
Brain Sci., 26, 226–227. Krumhansl, C. L., & Kessler, E. J. (1982). Tracing the
Johnsrude, I. S., Penhune, V. B., & Zatorre, R. J. (2000). dynamic changes in perceived tonal organization in a
Functional specificity in the right human auditory spatial representation of musical keys. Psycholog. Rev.,
cortex for perceiving pitch direction. Brain, 123, 155– 89, 334–368.
163. Kuhl, P. K. (2004). Early language acquistion: Cracking
Joyner, D. R. (1968). The montone problem. J. Res. Music the speech code. Nat. Rev. Neurosci., 5, 831–843.
Educ., 17, 115–124. Lametti, D. R., & Mattar, A. A. G. (2006). Mirror neurons
Juslin, P., & Sloboda, J. A. (2001). Music and Emotion: Theory and the lateralization of human language. J. Neurosci.,
and Research. Oxford, UK: Oxford University Press. 26, 6666–6667.
Keenen, J. P., Thangaraj, V., Halpern, A., et al. (2001). Langer, S. K. (1951). Philosophy in a New Key (2nd ed.). New
Absolute pitch and planum temporale. NeuroImage, York: New American Library.
14, 1402–1408. LeDoux, J. E. (2000). Cognitive-emotional interactions:
Kelley, A. E., & Berridge, K. C. (2002). The neuroscience Listen to the brain. In J. E. LeDoux (Ed.), Cognitive
of natural rewards: Relevance to addictive drugs. J. Neuroscience of Emotion (pp. 129–155). Oxford, UK:
Neurosci., 22, 3306–3311. Oxford University Press.
Khalfa, S., Guye, M., Peretz, I., et al. (2008). Evidence of Lerdahl, F. (2001). Tonal Pitch Space. New York: Oxford
lateralized anteromedial temporal structures involve- University Press.
ment in musical emotion processing. Neuropsychologia, Lerdahl, F., & Jackendoff, R. (1983). A Generative Theory of
46, 2485–2493. Tonal Music. Cambridge, MA: MIT Press.
Koelsch, S., Fritz, T., Cramon, D. Y., et al. (2006). Inves- Levitin, D. J. (1999). Tone deafness: Failures of musical
tigating emotion with music: An fMRI study. Human anticipation and self-reference. Inter. J. Comput. Anticip.
Brain Mapping, 27, 239–250. Sys., 4, 243–254.
Koelsch, S., Gunter, T. C., Friederici, A., et al. (2000). Levitin, D. J. (2006). This Is Your Brain on Music: The Science
Brain indices of music processing: “Nonmusicians” of a Human Obsession. New York: Penguin Group, Inc.
are musical. J. Cogn. Neurosci., 12, 520–541. Levitin, D. J. (2008). The World in Six Songs: How the Mu-
Koelsch, S., Gunter, T. C., von Cramon, D. Y., et al. sical Brain Created Human Nature. New York: Penguin
(2002). Bach speaks: A cortical “language-network” Group, Inc.
Levitin & Tirovolas: Cognitive Neuroscience of Music 229

Levitin, D. J., & Cook, P. R. (1996). Memory for musical Mitterschiffthaler, M. T., Fu, C. H. Y., Dalton, J. A., et al.
tempo: Additional evidence that auditory memory is (2007). A functional MRI study of happy and sad
absolute. Percep. Psychophys., 58, 927–935. affective states induced by classical music. Human
Levitin, D. J., McAdams, S., & Adams, R. L. (2002). Con- Brain Mapping, 28, 1150–1162.
trol parameters for musical instruments: A founda- Mizuno, T., & Sugishita, M. (2007). Neural correlates
tion for new mappings of gesture to sound. Organised underlying perception of tonality-related emotional
Sound, 7, 171–189. contents. Neuroreport: For Rapid Communication of Neuro-
Levitin, D. J., & Menon, V. (2003). Musical structure is science Research, 18, 1651–1655.
processed in “language” areas of the brain: A possible Münte, T. F., Altenmüller, E., & Jäncke, L. (2002). The
role for Brodmann Area 47 in temporal coherence. musician’s brain as a model of neuroplasticity. Nat.
NeuroImage, 20, 2142–2152. Rev. Neurosci., 3, 473–478.
Levitin, D. J.,, & Menon, V. (2005). The neural locus of Musacchia, G., Sams, M., Skoe, E., et al. (2007). Mu-
temporal structure and expectancies in music: Evi- sicians have enhanced subcortical auditory and au-
dence from functional neuroimaging at 3 Tesla. Mu- diovisual processing of speech and music. Proc. Nat.
sic Percep., 22, 563–575. Acad. Sci., 104, 15894–15898.
Levitin, D. J., Menon, V., Schmitt, J. E., et al. (2003). Nan, Y., Knosche, T. R., Zysset, S., et al. (2008). Cross-
Neural correlates of auditory perception in Williams cultural music phrase processing: An fMRI study.
Syndrome: An fMRI study. NeuroImage, 18, 74–82. Human Brain Mapping, 29, 312–328.
Liberman, A. M. (1982) On finding that speech is special. Narmour, E. (1990). The Analysis and Cognition of Ba-
American Psychologist, 37, 148–167. sic Melodic Structures: The Implication-Realization Model.
Liberman, A. M., & Mattingly, I. G. (1985). The motor Chicago: University of Chicago Press.
theory of speech perception revised. Cognition, 21, Neville, H. J., Bavelier, D., Corina, D. P., et al. (1998).
l–36. Cerebral organization for language in deaf and hear-
Liégeois-Chauvel, C., Peretz, I., Babai, M., et al. (1998). ing subjects: Biological constraints and effects of ex-
Contribution of different cortical areas in the tempo- perience. Proc. Nat. Acad. Sci., 95, 922–929.
ral lobes to music processing. Brain, 121, 1853–1867. Nietzsche, F. (1871/1993). The Birth of Tragedy (Whiteside,
Limb, C. J., & Braun, A. R. (2008). Neural substrates of Trans.). Penguin Books.
spontaneous musical performance: An fMRI study Ohnishi, T., Matsuda, H., Asada, T., et al. (2001). Func-
of jazz improvisation. PloS One, 3, e1679. tional anatomy of musical perception in musicians.
Loui, P., Guenther, F. H., Mathys, C., et al. (2008). Action- Cerebral Cortex, 11, 754–760.
perception mismatch in tone-deafness. Curr. Biol., 18, Otsuka, A., Kuriki, S., Murata, N., et al. (2008). Neu-
R331–332. romagnetic responses to chords are modified by
Maess, B., Koelsch, S., Gunter, T. C., et al. (2001). Musical preceding musical scale. Neurosci. Res., 60(1), 50–
syntax is processed in Broca’s area: An MEG study. 55.
Nat. Neurosci., 4, 540–545. Pallesen, K. J., Brattico, E., Bailey, C., et al. (2005).
Marques, C., Moreno, S., Castro, S. L., et al. (2007). Mu- Emotion processing of major, minor and dissonant
sicians detect pitch violation in a foreign language chords: A functional magnetic resonance imaging
better than nonmusicians: Behavioral and electro- study. Ann. NYAS, 1060, Special Issue: The Neuro-
physiological evidence. J. Cogn. Neurosci., 19, 1453– sciences and Music II: From Perception to Perfor-
1463. mance, pp. 450–453.
McDonald, C., & Stewart, L. (2008). Uses and functions Panksepp, J. (1995). The emotional sources of “chills”
of music in congenital amusia. Music Percep., 25, 345– induced by music. Music Percep., 13, 171–207.
355. Panksepp, J. (2003). At the interface of the affective, be-
Menon, V., & Levitin, D. J. (2005). The rewards of music havioural and cognitive neurosciences: Decoding the
listening: Response and physiological connectivity of emotional feelings of the brain. Brain and Cognition, 52,
the mesolimbic system. NeuroImage, 28, 175–184. 4–14.
Meyer, L. (1956). Emotion and Meaning in Music. Chicago: Panksepp, J. (2006). Emotional endophenotypes in evo-
University of Chicago Press. lutionary psychiatry. Prog. Neuro-Psychopharma. Biolog.
Miller, G. F. (2000). Evolution of human music through Psych., 30, 774–784.
sexual selection. In N. L. Wallin, B. Merker, & S. Pantev, C., Roberts, L. E., Schulz, M., et al. (2001).
Brown (Eds.). The Origins of Music. New York: MIT. Timbre-specific enhancement of auditory cortical
Milovanov, R., Huotilainen, M., Valimaki, V., et al. representations in musicians. NeuroReport, 12, 169–
(2008). Musical aptitude and second language pro- 174.
nunciation skills in school-aged children: Neural and Patel, A. D. (2003). Language, music, syntax, and the
behavioral evidence. Brain Research, 1194, 81–89. brain. Nat. Neurosci., 6, 674–681.
230 Annals of the New York Academy of Sciences

Patel, A. D. (2007). Music, Language, and the Brain. New tural components of music perception: A functional
York: Oxford University Press. anatomical study. Brain, 120, 229–243.
Patel, A. D., Iversen, J. R., Bregman, M. R., et al. (in Poppel, E. (1997). A hierarchical model of temporal per-
press). Studying synchronization to a musical beat in ception. Trends Cogn. Sci., 1, 56–61.
nonhuman animals. Ann. NYAS. Reese, H. W. (1968). The Perception of Stimulus Relations. New
Patel, A. D., Wong, M., Foxton, J., et al. (2008). Speech in- York: Academic Press.
tonation deficits in musical tone deafness (congenital Rizzolatti, G., Fadiga, L., Gallese, V., et al. (1996). Pre-
amusia). Music Percep., 25, 357–368. motor cortex and the recognition of motor actions.
Patterson, R. D., Uppenkamp, S., Johnsrude, I. S., et al. Cog. Brain Res., 3, 131–141.
(2002). The processing of temporal pitch and melody Sacks, O. (2007). Musicophilia: Tales of Music and the Brain.
information in auditory cortex. Neuron, 36, 767– New York: Alfred A. Knopf.
776. Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996).
Peretz, I. (1990). Processing of local and global musical Statistical learning by 8-month-old infants. Science,
information by unilateral brain-damaged patients. 274, 1926–1928.
Brain, 113, 1185–1205. Saffran, J. R., Johnson, E. K., Aslin, R. N., et al. (1999).
Peretz, I., & Coltheart, M. (2003). Modularity of music Statistical learning of tone sequences by human in-
processing. Nat. Neurosci., 6, 688–691. fants and adults. Cognition, 70, 27–52.
Peretz, I., Champod, A. S., & Hyde, K. (2003). Varieties Schlaug, G., Jäncke, L., Huang, Y., et al. (1995). In vivo
of musical disorders: The Montreal battery of eval- evidence of structural brain asymmetry in musicians.
uation of amusia. Ann. NYAS, 999, 58–75. Science, 267, 699–701.
Peretz, I., Gosselin, N., Tillman, B., et al. (2008). On-line Sergeant, J. (1993). Mapping the musician brain. Human
identification of congenital amusia. Music Percep., 25, Brain Mapping, 1, 20–38.
331–343. Shahin, A. J., Roberts, L. E., Chau, W., et al. (2008). Music
Peretz, I., & Kolinsky, R. (1993). Boundaries of separabil- training leads to the development of timbre-specific
ity between melody and rhythm in music discrimina- gamma band activity. NeuroImage, 41(1), 113–122.
tion: A neuropsychological perspective. Quar. J. Exp. Shuter-Dyson, R. (1999). Musical ability. In D. Deutsch
Psych. A, Hum. Exp. Psych., 46, 301–325. (Ed.), The Psychology of Music, Second Edition (pp. 627–
Peretz, I., Kolinsky, R., Tramo, M., et al. (2004). Func- 651). San Diego: Academic Press.
tional dissociations following bilateral lesions of au- Sloboda, J. A., & Juslin, P. (2001). Psychological perspec-
ditory cortex. Brain, 117, 1283–1301. tives on music and emotion. In P. Juslin, & J. A.
Peretz, I., & Zatorre, R. J. (Eds.). (2003). The Cognitive Sloboda (Eds.), Music and Emotion: Theory and Research
Neuroscience of Music. Oxford, UK: Oxford University (pp. 71–104). Oxford, UK: Oxford University Press.
Press. Sluming, V., Brooks, J., Howard, M., et al. (2007).
Petitto, L. A., Zatorre, R. J., Gauna, K., et al. (2000). Broca’s area supports enhanced visuospatial cogni-
Speech-like cerebral activity in profoundly deaf peo- tion in orchestral musicians. J. Neurosci., 27, 3799–
ple processing signed languages: Implications for the 3806.
neural basis of human language. Proc. Nat. Acad. Sci., Sluming, V. A., & Manning, J. T. (2000). Second to fourth
97, 13961–13966. digit radio in elite musicians: Evidence for musical
Phillips-Silver, J., & Trainor, L. J. (2005). Feeling the beat: ability as an honest signal of male fitness. Evol. Hum.
Movement influences infant rhythm perception. Sci- Behav., 21, 1–9.
ence, 308, 1430. Small, D. M., Zatorre, R. J., Dagher, A., et al. (2001).
Phillips-Silver, J., & Trainor, L. J. (2007). Hearing what Changes in brain activity related to eating chocolate:
the body feels: Auditory encoding of rhythmic move- From pleasure to aversion. Brain, 124(Pt. 9), 1720–
ment. Cognition, 105, 533–546. 1733.
Phillips-Silver, J., & Trainor, L. J. (2008). Vestibular influ- Spence, K. W. (1937). The differential response in animals
ence on auditory metrical interpretation. Brain and to stimuli varying within a single dimension. Psycholog.
Cognition, 67, 94–102. Rev., 44, 430–444.
Piccirilli, M., Sciarma, T., & Luzzi, S. (2000). Modularity Sridharan, D., Levitin, D. J., Chafe, C. H., et al. (2007).
of music: Evidence from a case of pure amusia. J. Neural dynamics of event segmentation in music:
Neurol. Neurosurg. Psych., 69, 541–545. Converging evidence for dissociable ventral and dor-
Pierce, J. R. (1983). The Science of Musical Sound. New York: sal networks. Neuron, 55, 1–12.
W. H. Freeman. Steinbeis, N., & Koelsch, S. (2008). Shared neural re-
Pinker, S. (1997) How the Mind Works. New York: W.W. sources between music and language indicate seman-
Norton & Co. tic processing of musical tension-resolution patterns.
Platel, H., Price, C., Baron, J. C., et al. (1997). The struc- Cerebral Cortex, 18, 1169–1178.
Levitin & Tirovolas: Cognitive Neuroscience of Music 231

Sternberg, S., Knoll, J., & Zuchovsky, P. (1982). Timing Vines, B. W., Krumhansl, C. L., Wanderley, M. M., et al.
by skilled musicians. In D. Deutsch (Ed.), The Psychol- (2005). Dimensions of emotion in expressive musical
ogy of Music, First Edition. (pp. 181–239). San Diego: performance. Ann. NYAS, 1060, 462–466.
Academic Press. von Ehrenfels, C. (1988). On “Gestalt qualities.” In B.
Tervaniemi, M., & Huotilainen, M. (2003). The promises Smith (Ed.), Foundations of Gestalt Theory. Munich:
of change-related brain potentials in cognitive neu- Philosphia Verlag. (Original work published in 1890).
roscience of music. Ann. N.Y. Acad. Sci., 999, 29–39. Warren, J. D., Uppenkamp, S., Patterson, R. D., et al.
Tillmann, B., Janata, P., & Bharucha, J. J. (2003). Activa- (2003). Analyzing pitch chroma and pitch height in
tion of the inferior frontal cortex in musical priming. the human brain. Ann. N.Y. Acad. Sci., 999, 212–214.
Cog. Brain Res., 16, 145–161. West, R., Howell, P., & Cross, I. (1985). Modelling per-
Tooby, J., & Cosmides, L. (2001). Does beauty build adap- ceived musical structure. In P. Howell, I. Cross, & R.
tive minds? Toward an evolutionary theory of aes- West (Eds.), Musical Structure and Cognition (pp. 21–52).
thetics, fiction and the arts. SubStance, 30, 6–27. London: Academic Press.
Trainor, L. J. (2005). Are there critical periods for musical White, B. W. (1960). Recognition of distorted melodies.
development? Dev. Psychobiol., 46, 262–278. Amer. J. Psych., 73, 100–107.
Trainor, L. J., & Trehub, S. E. (1994). Key membership Wilson, S. J., & Saling, M. M. (2008). Contributions of
and implied harmony in Western tonal music: De- the right and left mesial temporal lobes to music
velopmental perspectives. Percep. Psychophys., 56, 125– memory: Evidence from melodic learning difficul-
132. ties. Music Percep., 25, 303–314.
Trainor, L. J., Tsang, C. D., & Cheung, V. H. W. (2002). Wong, P. C. M., Skoe, E., Russo, N. M., et al. (2007). Musi-
Preference for sensory consonance in 2- and 4- cal experience shapes human brainstem encoding of
month old infants. Music Percep., 20, 187–194. linguistic pitch patterns. Nat. Neurosci., 10, 420–422.
Tramo, M. J. (2001). Biology and music: Music of the Yasui, T., Kaga, K., & Sakai, K. L. (2007). Language
hemispheres. Science, 291(5501), 54–56. and music: Differential hemispheric dominance in
Tramo, M. J., Shah, G. D., & Braida, L. D. (2002). Func- detecting unexpected errors in the lyrics and melody
tional role of auditory cortex in frequency processing of memorized songs. Human Brain Mapping, 30, 588–
and pitch perception. J. Neurophysiol., 87, 122–139. 601.
Trehub, S. E. (2003). The developmental origins of musi- Zatorre, R. J. (1985). Discrimination and recognition of
cality. Nat. Neurosci., 6, 669–673. tonal melodies after unilateral cerebral excisions.
Trehub, S. E., Schellenberg, E. G., & Kamenetsky, S. Neuropsychologia, 23, 31–41.
B. (1999). Infants’ and adults’ perception of scale Zatorre, R. J. (1988). Pitch perception of complex tones
structure. J. Exp. Psych.: Hum. Percept. Perf., 25, 965– and human temporal-lobe function. J. Acous. Soc.
975. Amer., 84, 566–572.
Vignolo, L. A. (2003). Music agnosia and auditory ag- Zatorre, R. J., Belin, P., & Penhune, V. B. (2002). Structure
nosia. Dissociations in stroke patients. Ann. NYAS, and function of auditory cortex: Music and speech.
999, 50–57. Trends Cogn. Sci., 6, 37–46.
Vines, B. W., Krumhansl, C. L., Wanderley, M. M., Zatorre, R. J., Perry, D. W., Beckett, C. A., et al. (1998).
et al. (2006). Cross-modal interactions in the per- Functional anatomy of musical processing in listeners
ception of musical performance. Cognition, 101, 80– with absolute and relative pitch. Proc. Nat. Acad. Sci.,
113. USA, 95, 3172–3177.