Music and Empathy are psychological neighbours. Empathetic people process music differently – Stephen Johnson * Music, Empathy, and Cultural Understanding – Eric Clarke, Tia DeNora, Jonna Vuoskoski * Current Disciplinary and Interdisciplinary Debates on Empathy – Eva-Maria Engelen, Birgitt Röttger-Rössler * Neurophysiological Effects of Trait Empathy in Music Listening – Zachary Wallmark, Choi Deblieck, Marco Iacaboni. * Interpersonal Reactivity Index (IRI).

“The ‘other’ need not be a person: it can be music.” Clifton (1983)

There are two pathways when it comes to understanding each other: thinking or mind reading and feeling or empathy.

Empirical investigations have shown that people who have a tendency to be more empathic experience more intense emotions in response to music.

Listening to sounds, even outside of a musical context, significantly activates empathy circuits in the brains of high empathy people. In particular, sounds trigger parts of the brain linked to emotional contagion, a phenomenon that occurs when one takes on the emotions of another.

Musical engagement can function as a mediated form of social encounter, even when listening by ourselves. Recent research has shown that trait empathy is linked to musical preferences and listening style. If we consider music through a social psychological lens, it is plausible that individuals with a greater dispositional capacity to empathize with others might also respond to music’s social stimulus differently on a neurophysiological level by preferentially engaging brain networks previously found to be involved in trait empathy.

Music can be conceived as a Virtual Social Agent… listening to music can be seen as a socializing activity in the sense that it may train the listener’s self in social attuning and empathic relationships. In short, musical experience and empathy are psychological neighbors.

Esenherg et al (1991) define empathy as, “an emotional response that stems from another’s emotional state or condition and is congruent with the other’s emotional state or condition.”

Who or what do we empathize with when listening to music?

For some people music is able to represent a virtual person with whom to empathize, and whom they can experience as empathizing with their felt emotions. Studies that have investigated people’s reasons for listening to sad music when they already feel sad have found that some listeners can experience the music itself as providing empathy and understanding for the feelings that they are going through, functioning as a surrogate for an empathic friend.

Mirror neurons may be as much a consequence of a culture of inter-subjective engagement as they are a foundation for it.

It is quite conceivable that people who are inclined to imagine themselves from others’ perspectives also tend to take up the physical actions implied by others’ musical sounds, whether a smooth and gentle voice, a growled saxophone, or any other musical sound reflecting human actions.

It’s no surprise our level or empathy impacts how we process social interactions with other people. But how might empathy affect the way we process music?

That’s the question addressed in a first of-its-kind study published in Frontiers in Behavioral Neuroscience. The results showed that high empathy people not only got more pleasure from listening to music, but also experienced more activity in brain regions associated with social interactions and rewards.

The implication is that empathy can make you interact with music as if it were a person, or a “virtual persona,” as described in a 2007 study:

“Music can be conceived as a virtual social agent… listening to music can be seen as a socializing activity in the sense that it may train the listener’s self in social attuning and empathic relationships.”

The researchers conducted two experiments to examine how empathy impacts the way we perceive music. In the first, 15 UCLA students listened to various sounds made by musical instruments, like a saxophone, while undergoing an fMRI scan.

Activation sites correlating with trait empathy (IRI subscales) in selected contrasts.

Some of the instrument sounds were distorted and noisy. The idea was that the brain might interpret these sounds as similar to the “signs of distress, pain, or aggression” that humans and animals emit in stressful scenarios, and these “cues may elicit heightened responses” among high-empathy people. Participants also completed the Interpersonal Reactivity Index, a selfreported survey commonly used by scientists to measure one’s level of empathy.

The results confirmed what the team had hypothesized: listening to the sounds, even outside of a musical context, significantly activated empathy circuits in the brains of high-empathy people. In particular, the sounds triggered parts of the brain linked to emotional contagion, a phenomenon that occurs when one takes on the emotions of another.

But how does empathy affect the way we listen to a complete piece of music?

To find out, the researchers asked students to listen to music that they either liked or disliked, and which was either familiar or unfamiliar to them. They found that listening to familiar music triggered more activity in the dorsal striatum, a reward center in the brain, among high-empathy people, even when they listened to songs they said they hated.

Familiar music also activated parts of the lingual gyrus and occipital lobe, regions associated with visual processing, prompting the team to suggest that “empathic listeners may be more prone to visual imagery while listening to familiar music.”

In general, high-empathy people experienced more activity in brain regions associated with rewards and social interactions while listening to music than did low-empathy participants.

“This may indicate that music is being perceived weakly as a kind of social entity, as an imagined or virtual human presence,” said study author Zachary Wallmark, a professor at SiViU Meadows School of the Arts. “If music was not related to how we process the social world, then we likely would have seen no significant difference in the brain activation between high-empathy and low empathy people.”

We often conceptualize music as an abstract object for aesthetic contemplation, Wallmark said, but the new findings could help us reframe music as a way to connect others, and to our evolutionary past.

“If music can function something like a virtual “other,” then it might be capable of altering listeners’ views of real others, thus enabling it to play an ethically complex mediating role in the social discourse of music,” the team wrote.

Music, Empathy, and Cultural Understanding

Eric Clarke, Tia DeNora, and Jonna Vuoskoski

In the age of the internet and with the dramatic proliferation of mobile listening technologies, music has unprecedented global distribution and embeddedness in people’s lives. It is a source of intense experiences of both the most individual (personal stereos) and massively communal (large-scale live events, and global simulcasts) kind; and it increasingly brings together or exploits a huge range of cultures and histories, through developments in world music, sampling, the re-issue of historical recordings, and the explosion of informal and ‘bedroom’ music making that circulates via YouTube. For many people, involvement with music can be among the most powerful and potentially transforming experiences in their lives.

To what extent do these developments in music’s mediated and mediating presence facilitate contact and understanding, or perhaps division and distrust, between people? This project has pursued the idea that music affords insights into other consciousnesses and subjectivities, and that in doing so may have important potential for cultural understanding.

The project:

1) brings together and critically reviews a considerable body of research and scholarship, across disciplines ranging from the neuroscience and psychology of music to the sociology and anthropology of music, and cultural musicology, that has proposed or presented evidence for music’s power to promote empathy and social/cultural understanding through powerful affective, cognitive and social factors, and to explore ways in which to connect and make sense of this disparate evidence (and counter-evidence);

2) reports the outcome of an empirical study that tests one aspect of those claims demonstrating that ‘passive’ listening to the music of an unfamiliar culture can significantly change the cultural attitudes of listeners with high dispositional empathy.

Researchers and Project Partners

Eric Clarke, Faculty of Music, University of Oxford

Tia DeNora, Sociology, Philosophy & Anthropology, Exeter University

Jonna Vuoskoski, Faculty of Music, University of Oxford


Music is a source of intense experiences of both the most individual (personal stereos, headphone listening) and massively communal (large-scale live events, and global simulcasts) kind; and it increasingly brings together or exploits an exceptional range of cultures and histories, through developments in ‘world music’, sampling, historical recording and hybridization. At a time when musicology, the social and cultural study of music, have become far more circumspect about essentializing and romanticizing claims, it is still not uncommon to find claims being made for music as a ‘universal language’ that can overcome (or even transcend) cultural difference, break down barriers of ethnicity, age, social class, ability/disability, and physical and psychological wellbeing.

There are widespread symptoms of this belief or claim, including the activities of the West-Eastern Divan Orchestra (founded by Edward Said and Daniel Barenboim, to bring together Israeli and Palestinian musicians); and the appointment by UNICEF of classical musicians to act as ‘goodwill ambassadors’, bringing their music to people in deprived, war-torn, or disaster-hit parts of the world so as to offer emotional support, solidarity, and a kind of communion.

An extract from the violinist Maxim Vengerov, who in 1997 was the first classical musician to be appointed a goodwill ambassador, reads: “1997, September: For Maxim Vengerov’s first official undertaking with UNICEF, he organized a musical exchange with children from Opus 118 a violin group from East Harlem, New York. The children of Opus 118, aged 6 to 13, came from three different elementary schools in this inner-city neighbourhood. This innovative programme has spurred a whole generation to learn ‘violin culture’. Along with the youths, Mr. Vengerov not only played Bach but also southern blues and tunes such as ‘Summertime’ and ‘We Shall Overcome’.”

And from the same webpage, beneath a picture showing the violinist in jeans and T-shirt playing as he leads a line of children in the manner of a latter-day Pied Piper is the caption: “In the remote village of Baan Nong Mon Tha, children from the Karen hill tribe ethnic group follow Maxim Vengerov, in a human chain, to a school run by a UNICEF-assisted NGO. Thailand, 2000.”

Similarly, the 1985 Live Aid, and 2005 Live 8, events were global pop music events intended not only to raise money (in the case of Live Aid) and popular pressure on politicians (in the case of Live 8) for the relief of famine and poverty, but also to galvanize a global consciousness and a united ‘voice’: as Bob Geldof, the prime mover of Live 8 put it: “These concerts are the start point for The Long Walk To Justice, the one way we can all make our voices heard in unison.”

And finally, the popular UK television series ‘The Choir’ (which has run to six series so far) documents the powerful ‘identity work’ and intense emotional experiences that accompany the formation of choirs in schools, workplaces, and military establishments out of groups of people who have had little or no previous formal musical experience, and who come from very varied walks of life (from bank executives to fire officers and military wives).

In all these very public examples of a much wider if less visible phenomenon, we see a complex mixture of implicit musical values, discourses about music’s ‘powers’, folk psychology and its sociological equivalent, and (in some cases) more or less grounded or unsupported claims about the impact of music on the brain (cf. Tame, 1984; Levitin, 2006). It would be easy to be hastily dismissive of some of these claims, but a considerable volume of research by highly regarded scientists and scholars, coming from disciplines that range from neuroscience and philosophy through psychology and sociology to anthropology and cultural studies has also made a significant case for the capacity of music and musicking (Small, 1998) to effect personal and social change (e.g. Becker 2004; DeNora 2013; Gabrielsson 2011; Herbert 2011). If music can effect change, and speak across barriers, it can also offer a means of intercultural understanding and identity work.

As Cook (1998: 129) puts it:

“If both music and musicology are ways of creating meaning rather than just of representing it, then we can see music as a means of gaining insight into the cultural or historical other. If music can communicate across gender differences, it can do so across other barriers as well. One example is music therapy… But the most obvious example is the way we listen to the music of other cultures (or, perhaps even more significantly, the music of subcultures within our own broader culture). We do this not just for the good sounds, though there is that, but in order to gain some insight into those (sub)cultures. And if we use music as a means of insight into other cultures, then equally we can see it as a means of negotiating cultural identity.”

In different ways, these (and other) claims seem to make use of a generalized notion of empathy. Empathy has recently seemed to gain considerable attention/currency in musicology, psychology of music, sociology of music and ethnomusicology as a way to conceptualize a whole range of affiliative, identity-forming, and ‘self-fashioning’ capacities in relation to music. But what is brought together or meant by the term ‘empathy’, and is it a useful and coherent way to think about music in relation to its individual and social effects?

Our project, and this report, arise from the disparate nature of the evidence for the claims about music’s transformative power, individually and socially, and the ‘scattering’ of the case across theories and findings in a huge disciplinary range: from research on music and mirror neurons (Overy and Molnar-Szakacs 2009) to the ethnomusicology of affect (Stokes 2010), the history of musical subjectivity (Butt 2010) and sociological studies of music and collective action (Eyerman and Jamieson 1998), the case has been made for different perspectives on music’s capacity to afford compassionate and empathetic insight and affiliation, and its consequent power to change social behaviour.

These diverse research strands all point to the crucial role that musicking plays in people’s lives, to its transformational capacity, and to the insights that it can afford. There is no single window onto ‘what it is like to be human’, but musicking seems to offer as rich, diverse, and globally distributed a perspective as any and one that engages people in a vast array of experiences located along dimensions of public and private, solitary and social, frenzied and reflective, technological and bodily, conceptual and immediate, calculated and improvised, instantaneous and timeless. The fact that music can be heard and experienced by large numbers of people simultaneously and in synchrony (orchestral concerts, stadium gigs, live simulcasts) means that the embodied experience of music can also be shared fostering entrainment and a sense of cosubjectivity. Indeed, some theories of the evolutionary significance of music highlight the importance of music’s empathy promoting aspects, suggesting that a fundamental adaptive characteristic of music is its capacity to promote group cohesion and affiliation (Cross & Morley, 2008).

While a whole range of studies has suggested that empathic interaction with other human beings is facilitated by musical engagement, the direct empirical evidence for this important possibility is scattered and disciplinarily disconnected. The aim of the project summarised in this report was to examine critically a substantial body of research evidence that relates to claims for music’s capacity to engender cultural understanding, primarily through the mediating construct of empathy; examine its consequences and significance, and provide a framework within which to connect its disparate elements and highlight points of interdisciplinary convergence and divergence; and carry out a focused empirical study that was designed to investigate a specific aspect of that complex case.

The report follows the general disciplinary outlines of the initial literature search, which revealed in excess of 300 items relating to the broad theme (‘Music, Empathy and Cultural Understanding’) of the project.


The word empathy has had currency in English for little more than 100 years, listed by the Oxford English Dictionary as being first used by the psychologist Edward Titchener in 1909, and defined by the OED as:

“a. Psychol. and Aesthetics. The quality or power of projecting one’s personality into or mentally identifying oneself with an object of contemplation, and so fully understanding or appreciating it.

b. orig. Psychol. The ability to understand and appreciate another person’s feelings, experience, etc.”

Titchener’s ‘empathy’ was his attempt to translate the term Einfühlung, coined by the philosopher Robert Vischer (1873) in a book on visual aesthetics. But it was Theodor Lipps (1903) who really championed the concept of empathy, developing it from an essentially aesthetic category (the ability to ‘feel into’ an artwork) into a much more general psychological/philosophical concept to account for the human capacity to recognize one another as having minds. Laurence (2007) gives an important account of the origin and development of the idea of empathy, tracing a line back to Adam Smith’s (1759) The Theory of Moral Sentiments, and Smith’s appeal to a notion of sympathy and ‘fellow feeling’ as the basis for understanding and living a moral life that is based on imagining how we would feel in the circumstances of others. The distinction between imagining how we would feel and simply identifying with how another feels is crucial, since it places Smith’s notion of sympathy in the domain of imaginative reason rather than blind contagion, and makes clear the role of cultural artefacts (paintings, literature, drama, music) as a means of socially learning that sympathetic attitude.

Laurence also draws significantly on the work of Edith Stein (1917) a doctoral student of Edmund Husserl whose On the Problem of Empathy also engages with the problem of how it is that we can know or experience the mental states of others, whether this knowledge or experience is given in some direct and primordial sense, and Stein’s conclusion that empathy is dependent on the mediating role of similarity with the person (or even animal) with whom/which we attempt to empathize. Laurence ends up with definition of empathy that emphasizes empathy as both a process, and as a social and educable achievement:

“In empathizing, we, while retaining fully the sense of our own distinct consciousness, enter actively and imaginatively into others’ inner states to understand how they experience their world and how they are feeling, reaching out to what we perceive as similar while accepting difference, and experiencing upon reflection our own resulting feelings, appropriate to our own situation as empathic observer, which may be virtually the same feelings or different but sympathetic to theirs, within a context in which we care to respect and acknowledge their human dignity and our shared humanity.” (Laurence 2007: 24)

Finally, and in significant contrast to Laurence, Baron Cohen (2011) provides a wide ranging account of empathy that explicitly presents it as a psychometrically measurable trait, with a genetic and environmental basis, distributed in a particular network of brain regions, and manifested in seven ‘degrees’ ranging from the zero degrees of empathy of the psychopath or autistic person, to the six degrees of empathy of some ‘hyper empathic‘ individuals. Baron Cohen regards empathy as critically valuable human resource, and sees the erosion or loss of empathy as an issue of global importance that has the most serious consequences for social health at scales ranging from the family to international relations.

As this necessarily brief review has revealed, there is a significant range of perspectives on empathy, from which two in particular might be drawn. The first is the distinction between empathy as a skill or social achievement acquired, educable, and in some sense fundamentally collective; and empathy as a trait relatively fixed, individual, and with a genetic component. The second concerns the extent to which different perspectives emphasize the involuntary and inter subjective character of empathy (sometimes expressed through the metaphor of contagion), involving identification with the other and a loss of self; as opposed to a more cognitive and deliberate view in which empathy depends upon an imaginative projection into the circumstances of the other (closer to what Smith called sympathy).

These differences in perspective affect the scope and reach of the term empathy, and are an issue to which we return towards the end of this report in the specific context of music.

Music and Empathy across Different Fields

This section critically reviews the existing literature on music and empathy under a number of different conceptual and disciplinary headings.

1. Neuroscience

An increasing body of neuroscientific evidence indicates the very close coupling of perceptual and motor functions in the central nervous system, strongly suggesting that one way to account for the human capacity to adopt the perspective of another (sometimes referred to as ‘theory of mind’ or even ‘mind reading’) is on the basis of the way in which a person’s experience of their own actions is entangled with their perception of the actions of others. At the level of brain anatomy, it has long been recognized that there are suggestive parallels between the organization of sensory and motor cortices of the human brain and this might provide at least superficial evidence for the close relationship between perception and action.

More recently, however, and particularly in the wake of the discovery of mirror neurons in the early 1990s (e.g. Di Pellegrino, Fadiga, Fogassi, Gallese, and Rizzolatti G., 1992), there has been a surge of interest in the ways in which perception action relationships at the level of the central nervous system might provide a powerful way to explain a variety of intersubjective and empathic phenomena. Freedberg and Gallese (2007: 197) have argued that the activation of a variety of embodied neural mechanisms underlie a range of aesthetic responses, proposing that “a crucial element of esthetic response consists of the activation of embodied mechanisms encompassing the simulation of actions, emotions and corporeal sensation, and that these mechanisms are universal.” Freedberg and Gallese are primarily concerned with the embodied and empathic qualities of visual art, but Overy and co-authors (Molnar Szakacs & Overy 2006; Overy & Molnar Szakacs 2009; McGuiness & Overy 2011) have developed a persuasive model of how the embodied, emotive and empathic effects of music might be understood from a mirror neuron perspective.

(A mirror neuron is a neuron that fires both when an animal acts and when the animal observes the same action performed by another.
Thus, the neuron “mirrors” the behavior of the other, as though the observer were itself acting. Such neurons have been directly observed in primate species. Birds have been shown to have imitative resonance behaviors and neurological evidence suggests the presence of some form of mirroring system.
In humans, brain activity consistent with that of mirror neurons has been found in the premotor cortex, the supplementary motor area (SMA), the primary somatosensory cortex and the inferior parietal cortex.)

In simple terms, mirror neurons (or mirror systems as they are often called) are neurons in a motor area of the brain that become active when an individual merely observes an action of the kind that these neurons are usually responsible for controlling.

These ‘as if body loops’, as Damasio (1999) has called them, provide a direct identification with the actions of another, and constitute the fundamental building blocks of what Gallese (2001; 2003) has termed the ‘shared manifold’. The shared manifold is understood as a three-leveled mechanism for inter-subjective identification: i) a phenomenological level that is responsible for our sense of similarity with others which Gallese equates with an expanded notion of empathy; ii) a functional level characterized by models of self other interaction; and iii) a subpersonal level, instantiated by the activity of mirror matching neural circuits (Mirror Neuron Systems). The aim of the shared manifold hypothesis is to ground a sense of empathy and self other identity without suggesting that human experience and neuroscience can simply be collapsed into one another: hence the distinction between phenomenological, functional and subpersonal levels.

Gallese is also at pains to point out that self other identity is not all that there is to inter subjectivity: mirror systems do not allow us to experience others exactly as we experience ourselves, since to do so would (ironically) preclude the possibility of experiencing others as such at all. Our capacity to experience an external reality with content and behaviours that we can understand is made possible by “the presence of other subjects that are intelligible, while preserving their alterity character.” (Gallese 2003: 177)

At times the mirror neuron idea has been presented as if it were a hardwired feature of the brain that acted rather like a magic bullet. But as Heyes (2010) has argued, while one way to see mirror neurons is as an evolutionary adaptation (and therefore present at the species level), an alternative is to see the development of mirror systems as acquired through the operation of associative processes through the lifetime of individuals. From this perspective, mirror processes originate in sensorimotor experience, much of which is obtained through interaction with others. Thus, the mirror neuron system is a product of social interaction, as well as a process that enables and sustains social interaction. One rather specific example of this kind of plasticity is the finding by Bangert et al. (2006) that trained pianists listening to the sound of piano music showed significantly more neural activity in the motor areas of their brains than did a matched group of non-musicians.

2. Perception-action coupling, Empathy and Embodiment

Mirror systems are one way to understand inter-subjective interaction and identity, with direct relevance to music, at a neural level.

At the behavioural level there is another extensive literature that has revealed the significance of mimicry and synchronization in mediating human relationships in general, and music in particular. In a review of the extensive literature Chartrand and Dalton (2008; see also Chartrand & Bargh 1999) make the case for the importance of mimicry in social life, ranging from postural and facial to vocal and syntactic mimicry (people unconsciously mimicking one another’s accents and sentence structures) as both manifestations of existing social bonds and affiliations as well as the means by which such social bonds may be established (e.g. Inzlicht, Gutsell & Legault, 2012). As Heyes (2011) has argued such imitative behaviours may be automatic and insuppressible, and constitute a fundamental embodied basis for a critically important domain of human social interaction.

At a similarly general level, a number of authors (e.g. Valdesolo and DeSteno 2011) have demonstrated the power of synchronization in inducing altruistic and compassionate behaviours, this synchronization in many cases serving to entrain people’s behaviours upon one another.

With this general psychological literature in mind, it is easy to see that music powerfully affords these kinds of cooperative and affiliative engagements. Music has long been associated with socially coordinated work, worship and celebration, where its rhythmically entraining attributes and opportunities for controlled mimicry and complementation (such as in the ‘call and response’ character of many vernacular musical cultures) play a central role (e.g. Clayton, Sager and Will 2005).

Hove and Risen (2009) demonstrated with a tapping task that the degree of synchrony between individuals tapping together predicted how affiliated those individuals rated one another, and in a more directly musical context both Kirschner and Tomasello (2009) and Rabinowitch, Cross & Burnard (2012) have shown that over both shorter and longer timescales children involved in rhythmically synchronized music activities subsequently behaved more cooperatively and empathically than did children who were involved in an equivalent but not synchronized activity. Music is a powerfully multi-sensory, and particularly kinaesthetic (see Stuart 2012) phenomenon whose embodied character draws people into fluid and powerful social groups at a range of scales and degrees of (im)permanence, and in doing so helps to enact a kind of empathy.

3. Dispositional empathy and music

As discussed above, some authors (e.g. Baron-Cohen 2011) have understood empathy as a trait, arguing that since some people have a tendency to experience empathy more readily than others, being more or less empathic can be understood as a personality trait or a disposition.

In its broadest sense, dispositional empathy can be defined as an individual’s general responsiveness to the observed experiences of others, involving both perspective-taking capabilities or tendencies, and emotional reactivity (e.g., Davis, 1980).

Davis (1980) has suggested that dispositional empathy is a multidimensional construct comprising at least four components:

Perspective-taking (PT) can be understood as the ability as well as the tendency to shift perspectives (i.e., to see and understand things from another’s point of view)

Fantasy can be described as the tendency to identify oneself with fictional characters in books and films, for example. By contrast,

Empathic Concern (EC) taps into the tendency to experience feelings of compassion and concern for observed individuals, whereas

Personal Distress is associated with the individual’s own feelings of fear, apprehension and discomfort in response to the negative experiences of others.

Empathic Concern (EC) and Personal Distress are associated with the more emotional side of empathy.

Theories of music-induced emotions suggest that some form of empathy may be involved in the emotional responses induced by music (e.g., Scherer & Zentner, 2001; Juslin & vastfjall, 2008; Livingstone & Thompson, 2009). The proposed mechanisms range from pre-conscious ‘motor resonance’ with musical features that resemble vocal and motor expression of emotion (Molnar-Szakacs & Overy, 2006; Livingstone & Thompson, 2009) and emotional contagion (Juslin & Vastfjal, 2008) to empathizing with emotions and notions that are construed in the listener’s imagination (e.g., Scherer & Zentner, 2001). Indeed, empirical investigations have shown that people who have a tendency to be more empathic experience more intense emotions in response to music (Vuoskoski & Eerola, 2012; Ladinig & Schellenberg, 2011), providing indirect evidence for the role of empathy in music-induced emotions.

As people with high dispositional empathy are more susceptible to emotional contagion in general (Doherty, 1997), it may be that highly empathic people also experience emotional contagion from music more readily (Vuoskoski & Eerola, 2012). A complementary explanation is that empathic people may be more likely to engage in some form of reflective empathy during music listening, involving visual or narrative imagery, for example (e.g., Vuoskoski & Eerola, 2012; 2013).

Dispositional empathy has been associated with music-induced sadness in particular, as highly empathic people have been found to experience more intense sadness after listening to sad instrumental music (Vuoskoski & Eerola, 2012).

Interestingly, empathic individuals also tend to enjoy sad music more than non-empathic individuals, suggesting that empathically experienced negative emotions such as sadness can be enjoyable in the context of music (Vuoskoski et al., 2012; Garrido & Schubert, 2011).

Similar findings have been made in the context of films, where the experience of empathic distress while watching a tragic film has been associated with greater enjoyment of the film (De Wied et al., 1994).

It is not yet known what the mechanisms behind such enjoyment are, although the portrayal of more positive themes such as friendship, love, and human perseverance often present in tragic films have been proposed as one potential source (De Wied et al., 1994). However, it is not clear whether this explanation could also apply in the context of music. Nevertheless, these findings do suggest that there is something inherently enjoyable in empathic engagement in an aesthetic context even when the experienced emotions could be nominally characterized as negative.

4. Music as a virtual person, music and subjectivity

People tend to describe music in terms of attributes commonly used to describe psychological attributes of people (Watt & Ash, 1998). Indeed, it has been suggested that music is capable of creating a ‘virtual person’ of sorts (Watt & Ash, 1998; Livingstone & Thompson, 2009). The musical expression of emotion bears a close resemblance to human vocal and motor expression of emotion, involving similar auditory and gestural cues (for a review, see Juslin & Laukka, 2003), and it has been proposed that listeners may respond to music as they would to the perceived emotional state of a conspecific (e.g., Livingstone & Thompson, 2009). However, music’s capacity to represent a virtual person seems to go beyond acoustic and gestural cues that resemble vocal and motor expression of emotion.

An example is provided by studies that have investigated people’s reasons for listening to sad music when they already feel sad. These studies have found that some listeners can experience the music itself as providing empathy and understanding for the feelings that they are going through, functioning as a surrogate for an empathic friend (Lee, Andrade & Palmer, 2013; Van den Tol & Edwards, 2013).

The participants in Van den Tol and Edwards’s study felt that:

“The music was empathizing with their circumstances and feelings, supporting them, making them feel understood, or making them feel less alone in the way they were feeling” (Van den Tol & Edwards, 2013, p. 14).

Thus, it appears at least for some people that music is able to represent a virtual person with whom to empathize, and whom they can experience as empathizing with their felt emotions.

There has been considerable interest in the musicological literature in the relationship between music and human subjectivity (e.g. Cumming, 2000; McCIary, 2004), pursuing the idea that music has attributes either of an idealized person, or of an idealized collection or community of people. Lawrence Kramer (e.g. 2001; 2003) has written extensively about music as the instantiation of a kind of imagined subjectivity not associated specifically with the composer, performers, or anyone else explicitly and literally engaged with the making of the music, nor simply as the mirror of a listener’s own subjectivity, but in a more abstracted and generic manner. Likewise, the philosopher and violinist Naomi Cumming, in a paper that focuses on the violin introduction to the aria ‘Erbarme Dich’ from J. S. Bach’s St. Matthew Passion, writes of how the listener does not just find her or his own subjectivity passively reflected back, but reconfigured:

“The pathos of Bach’s introduction, and its elevated style, are quite unmistakable, and recognition promotes empathy. Once involved with the unfolding of the phrase’s subjectivity, the listener does not, however, find a simple reflection of his or her own expectancies. The music forms the listener’s experience, and in its unique negotiation of the tension between striving and grief, it creates a knowledge of something that has been formerly unknown, something that asks to be integrated in the mind of the hearer.” (Cumming, 1997: 17)

And in a still more explicitly psychological manner, DeNora (2000; 2003; 2013) has written of the ways in which music acts as a technology that affords a listener the opportunity to structure and organize their identity in long-term ways, and as a way of managing their immediate emotional states and sense of identity. Writing of one of her informants, ‘Lucy’, DeNora points out how she (Lucy) uses music as a medium in which she can draw a connection between the musical material, her own identity, and a kind of social ideal. As Lucy herself expresses it, she ‘finds herself’, the ‘me in life’ within musical materials, in a manner that allows her to reflect on who she is and how she would like to be, a process that DeNora points out is not just private and individual:

“Viewed from the perspective of how music is used to regulate and constitute the self, ‘solitary and individualistic’ practices may be re-viewed as part of a fundamentally social process of self-structuration, the constitution and maintenance of self. In this sense then, the ostensibly private sphere of music use is part and parcel of the cultural constitution of subjectivity, part of how individuals are involved in constituting themselves as social agents.” (DeNora, 2000: 47-8)

Music and musicking, then, can be viewed as a rich environment in which more or less active participants (listeners and makers) can engage with the real and virtual subjectivities of other real and virtual participants, and in doing so come to experience (and perhaps increasingly understand) the cultural perspective that those others (real or virtual) inhabit. Music is in this way both a medium for empathic (and antagonistic) engagement with others, and an environment in which to explore and experiment with a range of more or less projected, fantasized and genuinely discovered subject positions.

5. Sociological Perspectives

Turning from the rather individualistic accounts that have dominated the previous sections, towards understandings of music and empathy that take an explicitly social stance, sociological perspectives that enhance understanding of empathic processes derive from what may be termed the ‘new sociology of art’ (de la Fuente 2007). This sub-disciplinary paradigm investigates aesthetic materials for the ways that they may be seen to frame, shape or otherwise have an impact in social life. It is linked in turn to perspectives within sociology that cluster around the so-called, ‘strong’ program of cultural sociology (Alexander 2008) in which cultural materials are understood as active mediators of psycho-social and subjective processes and in which arts are not understood to be ‘about’ society or shaped ‘by’ society but rather ‘in’ society and constitutive of social relations (Hennion 2007).

These ‘new’ sociologies of art and culture are in turn linked to a ‘meso’ perspective in sociology devoted to groups of actors understood as networks of people, practices (conventions, operations, activities with histories of use) and things (Fine 2010). They focus on interaction orders (Fine 2012), or local actions that produce forms of ordering. The interaction order is the place where meanings are created, validated and reproduced in ways that travel to other networks.

Within this meso perspective there is no macro-micro divide since both macro and micro are mutually produced within scenes and settings of activity. The focus on this concerted activity in turn offers considerable scope for examining the question of just how cultural forms, including musical forms, actually enter into action and experience (DeNora 2003).

The impetus for these perspectives comes from various distinct but complementary developments in sociology since the middle 1980s that describe the ways that aesthetic and symbolic materials ‘anchor’ action (Swidler 2002) by presenting actors with orientational materials that can inform, focus and specify styles and trajectories of action in real time. The concern with how aesthetic materials ‘get into’ action (Acord and DeNora 2012) is one that has been associated with other developments in sociology, most notably the turn from a focus on the cognitive components of action, and models of social actors as calculating beings, to a focus on embodiment and feeling (Witkin 1994). These developments resonate well with, and are further illuminated by, developments in the philosophy of consciousness that begin with notions of the ‘extended mind’ (Clark and Chalmers 1998) and draw out that concept to embrace ‘the feeling body’ (Colombetti 2013) in which embodied conditions and sensations are understood both to take shape in relation to things outside of individuals and to inform cognitive appraisal.

Insofar as feeling and embodiment can be understood to take shape through encounters with aesthetic materials and can be understood to cultivate sensibilities or predispositions in favour of some social scenarios (and thus, contrarily, away from others), aesthetic materials have been highlighted within sociology as sources of social order. In this respect, the ‘new’ sociology of art harks back to Adam Smith’s ideas of sympathy and the capacity for fellow feeling discussed above, in which Smith suggests that the capacity for fellow feeling and being able to imagine the other is a lynchpin of mutual orientation and, thus, social stability. While Smith makes it clear that sympathy (the capacity to imagine the other) is not empathy (the capacity to feel what the other is feeling, literally to share their experience), Smith’s focus on the prerequisites for achieving sympathy highlight the importance of bodily processes. Specifically, Smith describes how, if sympathy is to be achieved, it is necessary for actors to moderate their passions (tamp down, raise up levels of intensity or ‘pitch’ as Smith calls it) so as to encourage mutual engagement through shared modalities of feeling (Smith 1759: I. I. 36-39). In this respect, Smith’s interest in mutual emotional calibration, understood as a prerequisite of mutual understanding, resonates with Alfred Schutz’s concept of attunement, understood as the prerequisite for ‘making music together’ (his example is the performance of a string quartet used as a case in point of social action writ large and the need for mutual orientation, entrainment, calibration and the gestalt to which they give rise, namely, shared feeling forms.

Classical sociology can, in short, be read as offering important leads for the study of empathy, understood as emotional and embodied mutual orientation, predisposition and preference, and in this sense it can also be read as offering an excellent basis for appreciating ‘art in action’ and the role of the arts in underwriting communicative action or how we bind ourselves together in time, whether in conversation, with its prosodic and timing patterns (Scollon 1982), or more generally, as Trevarthen puts it, as the dynamic sympathetic state of a human person that allows co-ordinated companionship to arise’ (cited in Ansdell et al 2010). As such the arts, and in the case of this project, music, can be conceptualised as offering materials for shaping up the feeling body from infancy to old age, in a wide range of roles and guises.

But if the arts and music more specifically ‘get into’ action, the question, as stated earlier, remains: how does this happen and can we trace that process? And in relation to empathy, this question can be posed in terms of how shared feeling states, sensibilities and predispositions come about, and how they can be cultivated and thus also how they may be more problematically controlled (Hesmondhalgh 2013; Born 2012; DeNora 2003). Within sociology the most fruitful paradigms have focused upon learning, mostly informal situated learning, among which the classic work on this topic is Howard S. Becker’s ‘Becoming a Marijuana User’ (1953).

Becker’s piece has been used by subsequent scholars to develop new (grounded) theories of how culture gets into action from comparisons of how one learns to respond to musical ‘highs’ (Gomart and Hennion 1999) to how one learns to feel and respond sexually (Jackson and Scott 2007; DeNora 1997) and how one learns to respond in various workplaces and forms of occupation (Pieslack 2009; DeNora 2013) and how one manages and modifies emotions and energy levels as part of everyday self-care (DeNora 2000; Batt-Rawden, DeNora and Ruud 2005; Skanland 2010) or in scene-specific settings such as retail outlets (DeNora 2000). Specifically these studies have followed the ways that individuals and groups engage in processes of modelling, adjustment, tutoring and directing and attempted alignment with musical materials in ways that draw out emotional and embodied sensation and experience in musically guided ways. This work helps to highlight just how deeply culture can come to penetrate embodied processes and experiences, and thus dovetails with more recent work on the culturally mediated experience of health and wellbeing.

6. Music Therapeutic and Wellbeing Perspectives

The focus on music, health and wellbeing is a growing area (Koen et al 2008; MacDonald et al 2012; MacDonald 2013). It encompasses music therapeutic perspectives, community music, psychotherapeutic perspectives and more overtly medical applications as well as the history of medicine and healing. At the level of the individual, and in overtly medical contexts, research in these areas has documented music’s potential for the management of pain (Edwards 2005; Hanser 2009), anxiety (Drahota et al 2012), palliative care (Aasgaard 2002; Archie et al 2013; DeNora 2012), and immunology (Fancourt et al 2013; Chandra and Levitin 2013) all of which emphasise mind-bodyculture interaction. At the broader level at which music connects with and can be seen to contribute to wellbeing, music has been described ecologically as part of salutogenic (health-promoting) space (DeNora 2013).

In all of this work, there are excellent resources for the study of empathy, in particular for investigating empathy understood as sensibility, perception and orientation as musically mediated. Specifically, the focus on the malleability of consciousness and selfperception (Clarke and Clarke 2011) points to a human capacity for entering into different modes of awareness, ones that are simultaneously sensitising (aesthetic) and desensitising (anaesthetic), and in so doing indicates the importance and power of the cultural technologies through which altered states can be achieved. The case of music and pain management illustrates many aspects of this theme.

As Hanser has described it, recent theoretical understandings of pain have moved toward a multi-dimensional conception of pain perception, one in which pain is not unmediated but rather comes to be experienced in relation to cultural and situated interventions, including music. In part, musical stimuli simply compete with neural pain messages. But more interestingly, music stimulates both oxytocin and embodied sympathetic responses (Grape 2002; Hurlemann et al 2010). Recent interdisciplinary perspectives highlight how the music, in tandem with other biographical and contextual factors, may lead a person in pain into alternative situations, ones in which she/he becomes sensitised to musically inspired associations and desensitised to the former situation of being in pain. Thus, music cannot necessarily address the cause of the pain but it can redirect the sensation of pain by capturing consciousness in ways that recalibrate it (DeNora 2013). So too, in the Bonny Method of Guided Imagery and Music (Bonde 2012) music may provide a grid or template against which knowledge-production (memory, self-and mutual understanding, historical accounts) can be elaborated and scaffolded in ways that can be used to diminish ‘negative’ emotions and associations, effectively recalibrating perception and, in this case, the self-perception of pain.

Methodologically, the music therapy index (Nordoff & Robbins 2007), a highly detailed real-time log of musical and para-musical action, can be used to display key or pivotal moments of musically instigated or musically guided action (movement, shifts in comportment, utterances). So too, the ‘musical event’ scheme can be used to track with some precision the ways in which the musical permeates the paramusical and vice versa across time and in keeping with the meso focus described earlier networks (DeNora 2003; Stige and Aara 2012).

More generally, and in ways that draw music therapy and music and conflict resolution into dialogue, musical engagement may be used to transform psycho-social situations, again leading the actor or actors away from the perception of distressing features of body/environment and toward more positive features and scenarios, and in ways that may also contribute to hope, patience and general mental wellbeing (Ansdell et al 2010; Ansdell 2014) as well as broader forms of cross-cultural and interactional accord, linked to music and guided imagery (Jordanger 2007). Community Music Therapy has perhaps most notably described music’s role in the production of communitas, through joint improvisation and as a means of generating proto-social capital (Procter 2012).

Within the growing field of music and conflict transformation studies (Laurence 2007), a key theme has focused on the importance of shared practice and actual grass-roots (bottomup) musicking as a prerequisite for enduring forms of change (Bergh 2011; 2010; Robertson 2010). In particular, as Bergh has described, if music is to contribute to enduringly altered practice, or altered consciousness of the other, that endurance requires continued and repeated practice, continued and repeated participation in musical activity. And as we have already indicated, music is by no means an unmitigated ‘good’ within the conflict transformation literature: as Bergh has observed (Bergh 2011), music can be and has been used to inculcate feelings of animosity, or for purposes of oppression and torture (Cusick 2008); and historically has been incorporated into military culture through drill, march music and, more recently, through psych-op motivational techniques (Gittoes 2004; Pieslack 2009). Indeed Laurence (2007: 33), even while writing of the potential for music in conflict resolution, argues that inculcating peaceful values is one of music’s rarest uses, and that “of music’s purposes, many and probably most, serve the ongoing ends of power relationships one way or another.”

7. Cross-cultural perspectives

The final category of literature that we consider in this report touches on the potentially vast question of cultural and cross-cultural understanding. Within the psychology of music there has been an interest in the relationship between possibly ‘universal’ and culturally specific aspects of musical communication dating back to the very beginnings of both the psychology of music and ethnomusicology in the work of Carl Stumpf (Stumpf & Trippett 1911/2012).

Among other more recent empirical studies, Balkwiil, Thompson, & Matsunaga (2004) have shown that music can successfully communicate emotional meanings across different cultures, but ethnomusicologists, perhaps rightly suspicious of simplistic notions of inter-cultural communication, have pointed to issues of representation, and of the incommensurability of concepts (or in this case emotional meanings) across cultural contexts as factors that might undermine the validity of a naively empirical approach (Stock 2014). A number of authors have recently proposed the value of a ‘relational musicology’ that might tackle issues of inter-cultural understanding, including Cook (2012: 196) who argues for relational musicology as “a means of addressing key personal, social and cultural work that is accomplished by music in today’s world.”

One specific kind of ‘cultural work’ that has recently been addressed in ethnomusicology that is of direct relevance to this project is the affective and social work that is accomplished by/within modern ‘sentimental’ cultures. Martin Stokes (2007; 2010) has provided vivid accounts of the emotional, intimate and affiliative character of contemporary sentimental musical cultures in Egypt and Turkey, and Butterworth (2014) in relation to Peruvian huayno music, in which something very much like empathy (though Stokes relates it more directly to a ‘Smithian’ as in Adam Smith notion of sympathy) is understood as a cultural construct or condition. In Stokes’s words (2010: 193), one might view “sentimentalism as a kind of civic project, a way of imagining affable relations of dependence on strangers in modern society.” This is a very different perspective on empathy one that sees it as a social achievement, rather than personality trait; a collective skill, rather than the expression of a circuit of ten interconnected brain regions (cf. Baron-Cohen 2011). As Cook (2012) argues in relation to the relational understanding of musical methods from one domain applied to another (Schenkerian analysis and Chinese music; Nineteenth century Western transcriptions of Indian melodies that were presented as ‘authentic Hindostannie airs’), such encounters conceived within an appropriate relational conceptual framework offer a domain of shared experience and cross-cuitural understanding.

Figure 1. Mean IAT scores (D-values) ±standard error of the mean, grouped by condition. Positive D-values indicate an unconscious preference for West African (relative to Indian) people, and negative values indicate an unconscious preference for Indian (relative to West African) people.

To investigate the effect of listening to Indian vs. West African music on participants’ Dvalues as well as the hypothesized moderating effect of dispositional empathy on the effects of music we conducted an ANCOVA with the Type of Music (Indian or West African) as a factor, and Dispositional Empathy (global IRI scores) as a covariate. We also included an interaction term of Type of Music and Dispositional Empathy in the model. There was no significant main effect of Type of Music; F(1,54) = 2.59, p = .11, although the trend was in the anticipated direction with participants exposed to Indian music displaying a slight preference for Indian (relative to West African) people, and participants exposed to West African music displaying no apparent preference.

The mean D-values of the two groups are displayed in Figure 1. As might have been expected, Dispositional Empathy was not significantly related to IAT scores when examined across the two conditions; F(1,54) = 0.20, p = .89. However, there was a significant interaction between Type of Music and Dispositional Empathy; F(1,54) = 5.51, p = .023, n(2/p) = .09, suggesting that dispositional empathy indeed moderated participants’ susceptibility to the musical manipulations. The relationship between dispositional empathy and D-values in the two groups is displayed in Figure 2. We also investigated the potential contributions of musical training, sex, and subjective responses to the music (ratings of liking and felt emotional impact) to the D-values, but no statistically significant relationships were found.

Figure 2. The relationship between dispositional empathy and IAT scores (D value), grouped by condition. Positive D values indicate an unconscious preference for West African (relative to Indian) people, and negative values indicate an unconscious preference for Indian (relative to West African) people.

2 Conclusions

The empirical study has provided preliminary evidence for the hypothesis that listening to music without any explicit semantic content (such as comprehensible lyrics) can evoke empathy and affiliation in listeners with high dispositional empathy. This interpretation is supported by the significant interaction between Type of Music and Dispositional Empathy, which revealed that people with high dispositional empathy scores were more likely to display an unconscious preference for the ethnic group to whose music they were exposed than those with low dispositional empathy scores. The fact that high dispositional empathy made participants more susceptible to the musical manipulations suggests that the observed findings cannot be explained in terms of priming or knowledge activation effects, such as those observed in the case of background music and purchasing decisions (e.g. North, Hargreaves & McKendrick, 1999). The lack of a statistically significant relationship between the IAT scores and liking ratings also indicates that our findings cannot be accounted for by a simple preference effect (cf., Nantais & Schellenberg, 1999).

Instead, we propose that the more empathic participants may have been more open to the music, and more likely to entrain with the music involving internal mimicry and emotional contagion; and may also have been more likely to engage in reflective empathy, in the form of visual and/or narrative imagery, and/or semantic elaboration. In the context of music, entrainment comprises both temporal as well as affective components (see e.g., Phillips-Silver & Keller, 2012), and in general imitation and entrainment have been found to both reflect and elicit affiliation (Chartrand & Bargh 1999; Hove & Risen 2009). Since people with high dispositional empathy have been found to exhibit stronger motor and sensory resonance to observed actions, and the pain of others (Gazzola, Aziz-Zadeh 8i Keysers, 2006; Avenanti et al., 2008), it is possible that empathic people are also more likely to resonate with the acoustic and gestural features of music. This stronger resonance could explain why empathic individuals are more susceptible to emotional contagion from music (cf. Vuoskoski & Eerola, 2012), and why they also appear to be more sensitive to the affiliation-inducing effects of music listening.

However, further investigation is required in order to better understand the phenomenon, and to distinguish between the potential contributions of pre-reflective motor and affective resonance, and the more reflective empathy involving imagery, perspective-taking, and other extra-musical associations. As dispositional empathy comprises both emotional reactivity and cognitive perspective-taking attributes, either or both of these components may contribute to the observed affiliation-inducing effects of music listening. A possible way to investigate this would be to implement a nondemanding distractor task during the music listening, which would limit participants’ capacity to conjure up imagery and other extra-musical associations. Furthermore, the failure to find a statistically significant main effect of Type of Music on participants’ implicit associations could either be due to the fact that the variation in participants’ pre-existing preferences for Indian vs. West African people was too great in relation to our sample size, or that the participants with low dispositional empathy were simply not affected by the music.

Future studies could attempt to investigate this issue by implementing pre as well as post-manipulation measures of implicit associations, although there may be other, more problematic issues associated with exposing participants to Indian and West African images prior to the musical manipulations.

5. General Discussion, Implications, Prospects

The result of our empirical study provides some evidence for the capacity of music even when encountered in arguably the most passive circumstances (solitary headphone listening in a ‘Iaboratory’ setting) to positively influence people’s unconscious attitudes towards cultural others. Specifically, people with higher dispositional empathy scores show more differentiated positive associations with images of people from two different cultural groups after listening to music explicitly belonging to that cultural group than do people with lower dispositional empathy scores.

This is a striking result, and provides what might be characterized as narrow but ‘hard-nosed’ evidence for music’s positive inter-cultural potential, and we have speculated on the broad psychological mechanisms (including entrainment, mimicry, emotional contagion, and semantic elaboration) that may be responsible.

But a number of notes of caution also need to be sounded. We have no evidence for the robustness or duration of the effects that we have observed: it may be that this is a very temporary shift that is easily disrupted, casting doubt on the practical efficacy of music as an agent of change in cultural understanding. And in the light of the interaction with dispositional empathy, the result suggests that any practical efficacy might be confined to those individuals who are already predisposed to be empathic towards others arguably those people who are (to put it simplistically) the least urgent cases.

Are we then forced to conclude that music has little or no power to change attitudes among those people who are most resistant? Perhaps more seriously, music as we have already indicated is arguably as capable of distinguishing (Bourdieu 1984/1979), dividing and alienating people as it is of bringing them together. Hesmondhalgh (2013: 85) points out that “Music can reinforce defensive and even aggressive forms of identity that narrow down opportunities for flourishing in the lives of those individuals who adhere to such forms of identification”, and provides a vivid anecdotal example of just such a defensive/aggressive encounter with or through music. He describes a Friday night out with friends at a pub where an Elvis impersonator is performing. Having at first dreaded the performance, Hesmondhalgh and his friends, along with a large number of strangers who are also in the pub for a night out, are quickly won over and join with one another, and the performer, with increasing intensity. The chorus of the final song “elicits an ecstasy of collective singing, women and men, all at the top of our voices. There are smiles and laughter, but there’s melancholy too. It seems that bittersweet lines from the Elvis repertory are invoking thoughts about relationships, past and present… [We] stagger out of the pub feeling we’ve had a great night, and that the working week has been obliterated by laughter and bittersweet emotion. Unwittingly, I brush against a man’s drink as I’m leaving, and he follows me out demanding an apology for his spilt beer… The power of Elvis’s music, it seems, has brought strangers and acquaintances together, and with a formidable intensity. But my pursuer has reminded me unpleasantly that there are those who feel excluded from such collective pleasures. If music-based gatherings answer to our need for sociality and attachment, and combat loneliness, might they also evoke envy when others miss out?” (Hesmondhalgh 2013: 103-4)

Are we to regard music’s affiliative and divisive attributes as two sides of the same coin, or as a more fundamental incompatibility between emancipatory and oppressive qualities? Indeed, rather than considering how music might help to make a bridge between apparently pre-existent cultural ghettos, should we not be asking in what ways music is already implicated in the establishment and maintenance of those very ghettos in the first place? These are significant challenges to the potentially starry-eyed representation of music that an uncritical attitude might project; but as Hesmondhalgh, again, puts it: “Music’s ability to enrich people’s lives [and expand their empathic understanding] is fragile, but I believe it can be defended better if we understand that fragility, and do not pretend it floats free of the profound problems we face in our inner lives, and in our attempts to live together.” (Hesmondhalgh 2013: 171)

Part of understanding that ‘fragility’ is considering what, if anything, is special about music as a force for (compromised) cultural benefit. Why not football, or food, both of which can lay claim to mass engagement and global reach? Is there anything about music that affords either particular, or particularly powerful or efficacious kinds of intercultural engagement? One way to tackle these questions is consider what the mechanisms for empathy and cultural understanding might be, and in what ways those mechanisms are engaged by different cultural manifestations whether those are music, food or football. As our critical review of the literature reveals, this is a fascinating but considerable challenge, and one that turns in part on how broad or narrow a conception of empathy is entertained.

One approach might be to admit a considerable range of inter-subjective engagements as occupying different positions on an empathy spectrum, from conditions of seIf-other identity in the context of what might be called ‘deep intersubjectivity’ (perhaps emblematically represented by that pre-Oedipal oneness between mother and infant); through powerful experiences of compassionate fellow-feeling; to the operation of much more controlled and deliberate rational, imaginative projection into the circumstances of others. Some (such as Adam Smith, Felicity Laurence and Colwyn Trevarthen) would want to make firm distinctions between, say, empathy and sympathy. But an alternative might be to agree on an umbrella term (and empathy might do), and then focus on what distinguishes different positions under the umbrella, and what the implications (practical, functional, conceptual) of those differences might be.

A common thread that runs through most of these positions is the central role of embodiment in empathy. From the most neuroscientifically reductionist approach (e.g. a ‘fundamentalist’ mirror neuron perspective) to the position of Smith or Stokes, a capacity to feel the situation of another underpins the inter-subjective character of empathy/feIlow-feeling/sympathy. And arguably it is in this respect that music has ‘special properties’ properties of enactment, of synchronization and entrainment in situations ranging from a single individual alone with their music (the solitary headphone listener ‘Lost in music’ (Clarke 2014) to massively social contexts (pop festivals, simulcasts) where enormous numbers of peopie can participate in collective, synchronized, embodied engagement.

As others have pointed out (e.g. Cross 2012), music is a uniquely widespread, emotionally and physically engaging, social, participatory and fluidly communicative cultural achievement, a powerful (cultural) ecological niche that affords extraordinary possibilities for participants, and which both complements and in certain respects surpasses those other global cultural achievements in which human beings participate (language, religion, visual culture, craft). There is little, perhaps, to be gained by attempting to set any one of these up on a uniquely high pedestal but equally it is important not to flatten the terrain by failing to recognize music’s particular combination of affordances in this rich cultural mix: cognitive and emotional complexity, from solitary to mass-social engagement, compelling embodiment, floating intentionality (Cross 2012), synchronization/entrainment, flexible mimicry, temporal and ambient character, and digitaI-analog mix.

As our critical review of the literature has revealed, the empathy-affording character of this mix of affordances has been explored and theorized across an astonishing range of disciplines invoking mechanisms that range from mirror neurons to semiotics and the cultural history of sentimentalism. Are these kinds of explanation in any way compatible with one another, and is there a way to avoid a simplistic and potentially reductionist ‘layers of an onion’ approach in which supposedly ‘fundamental’ biological attributes (whether those are genetic in the case of a narrowly ‘trait’ perspective on empathy; or neurological in the case of sensorimotor contingency theory) underpin progressively more ramified and arbitrary cultural constructs? We have already seen (e.g. Heyes 2010) that from within the scientific literature itself, as well as from outside it, there is ample evidence for the plasticity of so-called fundamental properties, and for the reciprocal relationship between biology and culture. Mirror neurons may be as much a consequence of a culture of inter-subjective engagement as they are a foundation for it. But it clearly remains a considerable challenge to develop in detail the more flexible and relational approach that we point towards in this report.

Finally, there is the question of the utility of the concept or term ‘empathy’ itself. Perhaps rather like the word ‘meaning’, it both enables and suffers from the capacity to bring together a wide range of phenomena, which critics may find unhelpful in its heterogeneity. We share the concern not to confuse chalk with cheese, but against a drive to compart-mentalize (sic) we are persuaded of the long-term value of sticking with a word and its associated conceptual field which, although still just a century old, offers a rich and powerful way to try to understand a central element of human sociality. The debates about whether to understand empathy as a genetic predisposition, a personality trait, an emergent attribute of perception-action coupling, a skill, or a social achievement are symptomatic of the conceptual reach of the term.

Engelen and Rottger-Rossler (2012), in a brief overview of a speciai issue of the journal Emotion Review devoted to empathy, declare in their first sentence that “there is no accepted standard definition of empathy, either among the sciences and humanities or in the specific disciplines”, but nonetheless emphatically endorse the importance of continuing to develop better understandings of that fundamentally social capacity to “feel one’s way into others, to take part in the other’s affective situation, and adopt the other’s perspective“ to grasp the other’s intentions and thus to engage in meaningful social interaction.”

We, too, are committed to the value of that enterprise, and to the specific role that music may play in understanding empathy, and as itself a ‘medium’ for empathy. In addressing the complex network of relationships between neighbouring terms (sympathy, compassion, contagion, entrainment, ‘theory of mind‘, attunement…) we see the prospect of a more nuanced and differentiated understanding of what Baron Cohen (2011: 107) has characterized as “the most valuable resource in our world” and “an important global issue related to the health of our communities.”

Current Disciplinary and Interdisciplinary Debates on Empathy

Eva-Maria Engelen, Birgitt Röttger-Rössler, University of Konstanz, Department of Philosophy & Freie Universitat Berlin, Social and Cultural Anthropology

Almost anybody writing in the field would declare that there is no accepted standard definition of empathy, either among the sciences and humanities or in the specific disciplines. However, even when accepting that there can be no all time and universally valid definition, one can still try to clarify some aspects and establish a few landmarks that will help to ensure that the phenomenon with which various researchers are dealing is the same or has at least important features in common.

Although there is no established concept, several topics and discussions have proved to be crucial for the phenomenon that was once given this specially made up label empathy by Edward Titchner who introduced this word into English at the beginning of the 20th century in order to translate the German term Einfühlung.

The idea behind this special issue on empathy is to present a range of the currently most lively topics and discussions to be found not only within several disciplines but also across several disciplinary boundaries. This makes it interdisciplinary. Authors from different disciplines were asked to contribute to the field in a style that would be accessible for a broader range of interested readers. These contributions come from the following disciplines in which empathy is either an ongoing or an upcoming topic of academic interest: neuropsychology, developmental psychology, philosophy, literary studies, and anthropology. The commentators giving their views on the articles are sometimes experts on empathy from the same discipline as the authors and sometimes from adjoining ones. We tried as far as possible to introduce crossovers, but these did not always fit.

Points of Discussion and Open Questions

Roughly speaking, there are two pathways when it comes to understanding each other: thinking or mind reading and feeling or empathy. Nonetheless, one of the ongoing debates in psychology and philosophy concerns the question whether these two abilities, namely, understanding what the other is thinking and “understanding“ what the other is feeling, are separate or not.

Other debates refer to the best theoretical model for empathy and ask whether it makes sense to assume just one kind of empathy or whether one should differentiate between at least two kinds: cognitive and affective.

Further questions are: Does a living being have to be able to make a self-other distinction in order to be empathic? How far do emotional contagion or sympathy and pity differ from empathy? Is empathy necessarily an affective ability and does it have to be conscious? Does it occur in face to face relationships between two persons or more? And can it also occur between a reader and a fictive character in a novel (Coplan 2004)?

These are just some of the questions currently being discussed. But before addressing them in detail in the following six articles and twelve commentaries, we shall survey the different definitions of empathy presented and defended in this special issue.

A Starting Point for the Discussions

We start off with the concept of empathy in the social cognitive neurosciences. The major growth of interest in empathy is largely due to a recent debate in this field. Previously, in the late nineteenth and first half to middle of the twentieth century, it was an important term in psychology, hermeneutics, and phenomenology. Later on, interest in the concept spread to developmental psychology as well. But the currently ongoing debate received its initial impetus from the question how far mind reading and empathizing are different faculties and how far they may not be completely separable (Singer 2006).

Basically speaking, both faculties are about understanding the other, either cognitively or emotionally. What are the intentions of the other? What are his or her wishes, beliefs, or deductions? These questions belong to the mind reading side, whereas understanding the other‘s emotional state belongs to the other side: the capacity of empathy.

Nonetheless, despite these clear cut definitions, there are also concepts such as the affective theory of mind that is also called cognitive empathy. The rationale for this distinction is that empathy is based on understanding the affective states of others.

Another question that one might consider before reading the assembled articles on empathy is whether empathy has to be a process leading to a conscious state. We advise the reader to bring to mind the definition of empathy in his or her own research perspective before reading the articles presented here. Whether one agrees or disagrees with many of the arguments exchanged and discussed in the following articles and commentaries will depend on which definition of empathy one already has in mind. Hence, a reflection on one’s own implicit or explicit definition might lead one to reconsider one’s initial assumptions. Whatever the case, it will certainly help one to understand how different disciplines take divergent approaches to the subject.

One might also bear in mind that the notions of understanding and empathy to be found in the long lasting philosophical hermeneutic tradition have been used to differentiate between the sciences and the humanities. Explaining was considered to be the method of the sciences, whereas understanding and empathy were the methods of the humanities. This involves the assumption of a deep dualism, and one should be cautious about claiming a particular term for one or the other discipline and tradition without thoughtful reflection if one wishes to avoid stepping into the footprints of such dualisms.

Empathy as Embodied Capacity for Social Orientation

Coming from the humanities, we propose the following definition for empathy:

Empathy is a social feeling that consists in feelingly grasping or retracing the present, future, or past emotional state of the other; thus empathy is also called a vicarious emotion. (Vicarious: experienced in the imagination through the feelings or actions of another person.)

As a social feeling empathy is always shaped through cultural codes, which differently emphasize, modulate and train the capacity to “feel into“ another person’s emotions. The main function of this feelingly grasping is, we assume. orientation in social contexts. This can mean taking part in the precise emotional state that the other is in at a certain moment, namely: being happy when she is happy, scared when she is scared. and so forth.

But this does not have to be the case. Grasping the other’s emotional state, that is. adopting the other‘s emotional perspective, could also produce a different feeling or emotion in me than the one currently being experienced in the other. And even when the empathic adoption of the other’s perspective produces in me the same emotion as the other is having (or is fictively experiencing) at that very moment, it would not be the same emotion, because the self-other differentiation has not been overcome.

We want to make sure that we do not take empathy to mean the same as sympathy or pity. Both are, in our opinion, special forms of empathy that cover only a certain aspect of empathic processes. Whereas pity is the mode of feeling sorry for the other, sympathy is the mode of being in favor of the other. Both these feelings are ways of adopting an emotional perspective (as empathy is), but they cover only a special form of emotional perspective taking that is structured by the social bond or relation between the persons involved. Thus in social life, pity and sympathy are most likely to occur toward persons one is related to or who belong to one’s own ingroup, but less often toward outgroup members who are mostly perceived as being totally different, strange, or even malevolent, in short, as persons one can scarcely identify with.

Pity and compassion as particular kinds of empathy are deeply connected to social attachment. Frans de Waal (2009) conceives empathy as an evolved concern for others that is triggered through identification with these others. “Empathy’s chief portal is identification,” he argues, meaning that close social bonds increase, in a quasi-automatic way, the emotional responsiveness to others and thus the readiness to help and support fellow beings (de Waal. 2009, 2l3).

Continuing his line of argument, he stresses that empathy also needs a “turn off switch,” a mechanism to override and regulate automatic empathic responses. He considers that what constitutes this turn off switch of empathic processes is a lack of identification. What becomes evident here is that de Waal is implicitly equating pity and compassion with empathy, or he is conceiving them as the evolutionary basis of empathy. lf fellow beings harm or violate each other, as it is often the case in social reality, they must, according to de Waal’s model, have switched off their empathic capacity.

We deliberately take another position here: We conceive empathy as an evolutionarily grounded capacity to adopt an emotional perspective, to implicitly “feel into” the other regardless of the behavioral outcome. This may be directed toward ingroup members and be prosocial and supporting, or toward outgroup members and be destructive and harming.

We make a point of affectively grasping the emotional state of another, but that does not mean to draw a definite line between cognitive understanding and emotional grasping. There are good reasons to stick to a narrower notion when it comes to defining empathy as a “feelingly grasping” if one wants to make sense of notions such as vicarious emotion or of the history of the notion that started with Einfühlung (feeling into). However, the specific conceptual perspective one takes depends very strongly on one’s research traditions and research interest.

When it comes to the relation between empathic perspective taking and the cognitive perspective taking that is related to theory of mind (TOM), we cannot judge the discussions amongst neuropsychologists regarding whether or not these are completely different kinds of perspective taking, and whether or not these processes take place in different brain areas. However, defining the term according to an established tradition, we take empathy to be the emotional perspective taking; and mind reading (in TOM), to be the cognitive perspective taking. Nonetheless, on a purely conceptual level, one might have to admit that the two faculties cannot be separated altogether, because in cognitive perspective taking, the subject who is taking the perspective of another being has to be at least interested in the other being, and that means to care for the other in some way. First, you have to consider the other as an equal in a certain way, as a fellow human being, for instance, or at least as a creature able to feel. Second. you have to consider the other and the other‘s actions as relevant to yourself. You have to be somehow interested in order to be either emotionally involved or curious about the other’s intentions. Therefore, both cases, empathy and Tom, start with the same precondition:

You have to consider the other as being the same as you and of being your counterpart in a particular situation; there has to be a tacit analogy between the subject adopting the other‘s perspective and the other whose perspective is being taken, be it emotional or cognitive.

When specifying what we meant by empathy, we wrote of feelingly grasping or feelingly retracing something; this already suggests that the processes of feeling and of comprehending cannot always be separated clearly. And this makes empathic acts particularly interesting, because they resist the artificial dualisms in the philosophy of mind that still emboss philosophical, scientific, and everyday speech.

To recap briefly, empathy, as the embodied (or bodily grounded) capacity to feel one‘s way into others, to take part in the other’s affective situation, and adopt the other‘s perspective, is a fundamentally social capacity. It allows one to grasp the other’s intentions and thus to engage in meaningful social interaction. Empathy is a crucial means of social communication. It is not just an emotional contagiousness: in which one remains concentrated on oneself.

However, this definition of empathy fails to specify whether this comprehension involves a kind of simulation or imitation of the minds of others. In many of the following contributions, we shall see what important role simulation plays in the debates on a theoretical model of empathy.

Outline of the Contributions

The following six articles are written by distinguished scholars on empathy who come from five different disciplines. Each contribution presents recent research findings and theoretical reflections about the phenomenon of empathy within the respective discipline and simultaneously gives an insight into some currently ongoing debates on the subject within as well as across disciplinary boundaries. The following outline might already give a first impression about this.

Social Cognitive Neuroscience: Cognitive and Affective Empathy

The neuropsychologist Henrik Walter (2012) places his accent on understanding the emotional or affective states of another human being. Furthermore, he views understanding as a purely cognitive concept in this context that suggests making deductions and reasoning. Because Walter concentrates on this approach to understanding the affective states of others, conceptions such as affective theory of mind or cognitive empathy are also highly relevant for his ideas on the capacities for understanding other human beings. Whether this empathy is due to a cognitive faculty or an affective one is not the focus of this distinction. Empathy is, in this case, defined only by the understanding of the emotional state of the other and not by whether the process of understanding is either an affective one or a cognitive one. If it is a cognitive one, it is called cognitive empathy or affective theory of mind; if it is an affective one, it is called affective empathy.

Walter presents this conceptual analysis before linking it both to findings in empirical research investigating the neural basis of empathy and to data on the possible neurogenetic basis of empathy. The tradition followed by Walter when differentiating between TOM, cognitive empathy, and affective empathy is one developed in psychology since the late 1950s. It defined empathy as an emotional or affective phenomenon, and introduced the notion of cognitive empathy as a cognitive faculty or “intellectual or imaginative apprehension of another‘s condition or state of mind” (Hogan, 1969, 308). The main topic within this research tradition is the accuracy of our ability to conceive the other’s condition. Cognitive empathy is not defined in terms of shared emotions but in terms of knowing another’s state of mind by inferential processing (Ickes, 1997).

Social Cognitive Neuroscience again: Neural Overlap and Self-Other Overlap Stephanie Preston and Alicia Hofelich’s contribution (2012) comes from one of the most rapidly growing research fields on empathy, namely, the social neuroscience of empathy. Preston and Frans de Waal (2002) are well known in this field for having developed the perception-aclion model of empathy. This proposes that observing an emotion in someone else generates that emotion in the observer. Preston and Hofelich use this model to argue in favor of a neural overlap in the early stages of processing all cases of social understanding such as cognitive empathy, empathic accuracy, emolion contagion, sympathy. and helping behavior. The self-other overlap in empathy occurs only at a later state of processing. They offer some criteria for differentiating between neural overlap. subjeclive resonance, and personal distress. Because the self-other overlap is crucial for the definition of empathy. this represents an important attempt to seek empirical support for a theoretical differentiation. In addition, it offers a taxonomy of the different cases of social understanding that are supposed to be highlighted by a biological view of empathy.

The academic challenge of this undertaking lies not least in the attempt to show that there is some such thing as a self-other overlap on the neural level, and that it is not just to be found on the subjective level on which the conceptual capacities of a human being are already“at work.”

In order to engage in an empathic process, the empathic subject has to be able to differentiate between his or her own affective states and those of the being he or she is being empathic with, be this a conscious process, as is quite often the case on the subjective level, or a subconscious process on the neural level. This is also a necessary precondition for cognitive empathy and sympathy, but not for emotional contagion. Scientific research on the subjective overlap, that is, the sharing of an emotion, is the task of psychology. But in order to grasp this point on a biological level one has to avoid the subjective perspective. This is done by defining the self-other overlap via the notion of the activation of a personal representation in order to experience an observed state or action, and not via the notion of the activation of a personal representation when acting oneself or being in the state oneself. The overlap in representation on the neural level has to be reflected by a spatial overlap of brain activation between imitation and observation of facial emotional expression (on the subjective level, one is speaking about “sharing another’s emotional or intentional state”).

The process of observing or imagining someone else in a situation might therefore be crucial for determining whether a neural representation of an emotion is the representation of the emotion in somebody else, and therefore an empathic reaction, or whether it is the neural representation of one‘s own emotional process.

Developmental Psychology: The Self-Other Distinction

The developmental psychologist Doris Bischof Kohler (20l2) concentrates on the subjective level of empathy. She defines empathy as understanding and sharing the emotional state of another person. This definition implies not only that an empathic capacity is linked strongly to cognitive capacities, but also that the self-other distinction is crucial for the notion of empathy.

Bischof Kohler’s investigations on empathy are therefore related to her research on the symbolic representation of the self in imagination (self recognition). Her findings reveal that only children who are able to recognize themselves exhibit empathic behavior. This does not imply that self recognition leads to empathic behavior, but that it is a necessary precondition for empathy. And as her data show, this mode of self recognition does not have to be a kind of metarepresentation or conscious self reflection that the theory of mind predicts to first emerge only in 4 year olds. This can explain not only why empathy is already observable in 2 year old children but also why the mere recognition of a mark on one’s cheek while looking in a mirror is a transitional state to self recognition that is not linked to empathy. Her conclusion from these results is that “the capacity to empathize is an effect of maturation rather than socialization.”

Philosophy: Empathy and Simulation Theory

The philosopher Karsten Stueber (20l2) presents a model of the cognitive and afective understanding and knowledge of another human being’s mind, and demonstrates the importance of empathy for social cognition. He is well known as a representative of simulation Iheory-an approach that fits quite well with empirically based theories on empathy. In this article, he extends this basic approach by replying to some narrativist criticism. His main focus is on the cognitive mechanisms that allow us to gain knowledge of other minds and therefore on social cognition and on our understanding of individual agency. One challenge for such an approach is to give a theoretical account of resonance phenomena and projection mechanisms that does not presuppose some kind of Canesian subject who remains in a solitary state of skepticism about the existence of other minds. While insisting on the imponance of our sensitivity to differences between ourselves and other human beings, he introduces the importance of the other on the two levels distinguished in the simulation approach. The first level is the basic level of neuronal resonance phenomena. It is activated automatically by observation of the bodily activities and the accompanying bodily and facial expressions of other beings (basic empathy). The second level is the more developed stage, namely, the re enactment of the thoughts and reasonings of another human being as a rational agent (re enactive empathy). On this level, Stueber admits that in order to understand the actions of another person, we do not necessarily have to appeal to his or her beliefs and desires, but that the knowledge of the other‘s character traits or the other’s role in various social contexts could be equally important. By accepting this possibility, he opens up his model not only to some narrativist proposals for understanding the actions of others but also to the social, historical, or cultural contexts that one might have to consider in order to understand the actions of another human being. He insists, however, that this information would make neither the re enactment nor the simulation superfluous, because pretend beliefs and pretend desires are at the core of the imaginative perspective taking that is necessary for empathy.

Anthropology: The Cultural Embedednass of Empathy

The opening up of simulation theory toward an integration of personal, historical, and cultural information makes a philosophical approach like Stueber‘s attractive for a cultural and social anthropologist such as Douglas Hollan (2012). He takes up the distinction between basic empathy and re enactive empathy, although calling the latter complex empathy instead. This allows him not only to accept embodied forms of imitation and attunement as biologically evolved capacities, but also to concentrate on the more language bound evaluations and adjustments that have evolved culturally and historically. Hollan emphasizes that one has to be acquainted with the latter and with the personal background of a person in order to understand why he or she is in a certain emotional state. And, as he points out, this is necessary in order to be able to be empathic, because one has to understand not only that a person is in a certain emotional state but also why. In other words, one needs to have a cenain amount of knowledge about the normative and moral standards of a culture or society before one can evaluate the meaning of social situations and forms of behavior and comprehend another’s feeling state within the context of social circumstances. In short, empathic processes cannot be detached from the social and cultural contexts in which they are embedded. One way to narrow down the range of the meaning of the definition of empathy is to delete the need to understand why the person is in the state from the definition, leaving only the understanding that a person is in a certain emotional state.

The heuristic differentiation between basic empathy and complex empathy is in line with the ability to determine that another person is in a cenain emotional state and to understand the experience of the other. By reporting important research results on empathy in social anthropology, Douglas Hollan demonstrates not only how far some of the main features of empathy seem to be, by some means, universal, but also how far the studies on empathy need to be refined in light of some findings from anthropological research.

Intercultural findings on empathy reveal that the blending of feelingly perspective taking and cognitive perspective taking is one of the constant features of empathy. whereas the differentiation into “me” and “the other“ seems to be less distinct in empathic like responses in many non Western societies. Another finding of Hollan‘s research is that in the Pacific region, empathy is not a neutral engagement in the understanding of the emotional state of the other, but more like a sympathy that is linked very frequently to a positive attunement with that other person. And this positive attunement is expressed as an active doing rather than a passive experience.

Alongside these research results, he has noticed another, rather opposite tendency: a widespread fear that an empathic like knowledge could be used to harm others. This is why in many parts of the world-from the lndo Pacific to Latin America or Nonhem Canadapeople try to mask their faces. that is, to not express their inner feelings and thoughts but always show a “bright” face and not disclose their vulnerabilities. This phenomenon points to the fact discussed above that empathy is not linked automatically to compassion and helping attitudes, but might also be used by enemies or individual psychopaths as a way to find out how to harm the other.

Among the most challenging research desiderata that result from anthropological findings is the call for more studies on the complex interrelationship between the culture specific moral and situational contexts mediating the expression of empathy on the one side and the dispositions (or traits) that individuals develop to experience and display empathy on the other. Put succinctly, all cultures have some people who are likely to empathize more and others who are likely to empathize less. Hollan considers one of the most demanding tasks facing future research is to investigate how far personality traits interact with the culturally different modes of conceptualizing empathy.

Literary Studies: A Three-Step Model of Human Empathy

The findings on empathy filters introduced by the ethologist and primatologist Frans de Waal might well have been one of the starting points for the theory on empathy proposed by Fritz Breithaupt (2012), a scholar of German studies. As already mentioned. de Waal (2009, 213) has argued that “empathy needs both a filter that makes us select what we react to, and a turn off switch.” Breithaupt shares the hidden agenda for this approach, namely, that human beings are hyperempalhic, without equating pity and compassion with empathy. He has developed a three step model of human empathy that should account for the individual and cultural variety in empathy that also interests Douglas Hollan. According to Breithaupt’s theory, individual and cultural diferences are due to the control functions of blocking and channeling empathy.

These blucking mechanisms are important for a hyperempathic being (Step I) because of the costs accompanying such a social hyperactivity. As well as requiring energy, the danger of self loss might be another cost of empathy in this approach. This possibly ongoing activity therefore needs to be blocked (Step 2). Neurobiologists such as Marco lacobini (2008) have therefore proposed some kind of “super mirror neurons“ that control the mirror neurons. But, because Breithaupt is dealing with more conscious processes, he is hinting at cultural techniques and learning without excluding the possible existence of evolutionarily evolved mechanisms as well. Once the blocking mechanisms are in action, a third step is needed in order to be able to experience empathy at all (Step 3). This step consists in the techniques to circumvent the blocking mechanisms.

The technique to unblock the empathy inhibition on which Breithaupt is concentrating is side raking in a three person setting of empathy. The reason why he turns to a three person instead of a two person model is linked to the observation that hyperempathy in human beings goes hand in hand with hypersociabiliry, and a two person model might be too narrow to encompass this. The side taking process is deliberate: A person decides who’s side to take. After making this decision, empathy emerges (or returns), and it maintains and strengthens the initial choice, because empathy allows emotions to be released that confirm the decision. Breithaupt points out explicitly that the side taking is not involved in empathy itself (as it is in sympathy), but that it is rather “external” to it. The advantage of this model lies in the ability to combine cognitive elements in perspective taking with a caring attitude that might evolve when the side taking decision is followed by empathy.

The ambition of this special issue with its six articles from several disciplines is to give an overview on recent research on empathy. The twelve commentaries not only contribute greatly to achieving this aim but also help significantly to identify the hotspots in ongoing disciplinary and interdisciplinary debates.

Neurophysiological Effects of Trait Empathy in Music Listening

Zachary Wallmark, Choi Deblieck and Marco Iacaboni.

The social cognitive basis of music processing has long been noted, and recent research has shown that trait empathy is linked to musical preferences and listening style.

Does empathy modulate neural responses to musical sounds?

We designed two functional magnetic resonance imaging (fMRI) experiments to address this question. In Experiment 1, subjects listened to brief isolated musical timbres while being scanned. In Experiment 2, subjects listened to excerpts of music in four conditions (familiar liked (FL)/disliked (FD) and unfamiliar liked (UL)/disliked (UD).

For both types of musical stimuli, emotional and cognitive forms of trait empathy modulated activity in sensorimotor and cognitive areas: in the first experiment, empathy was primarily correlated with activity in Supplementary motor area (SMA),

Inferior frontal gyrus (IFG)

and Insula;

In Experiment 2, empathy was mainly correlated wth activity in prefrontal,


and reward areas, Taken together. these findings reveal the interactions between bottom-up and top-down mechanisms of empathy in response to musical sounds, in line with recent findings from other cognitive domains.


Music is a portal into the interior lives of others. By disclosing the affective and cognitive states of actual or imagined human actors, musical engagement can function as a mediated form of social encounter, even when listening by ourselves. It is commonplace for us to imagine music as a kind of virtual “persona,” with intentions and emotions of its own: we resonate with certain songs just as we would with other people, while we struggle to identify with other music.

Arguing from an evolutionary perspective, it has been proposed that the efficacy of music as a technology of social affiliation and bonding may have contributed to its adaptive value, As Leman indicates: “Music can be conceived as a Virtual social agent… listening to music can be seen as a socializing activity in the sense that it may train the listener’s self in social attuning and empathic relationships.” In short, musical experience and empathy are psychological neighbors.

The concept of empathy has generated sustained interest in recent years among researchers seeking to better account for the social and affective valence of musical experience; it is also a popular topic of research in social neuroscience. However, the precise neurophysiological relationship between music processing and empathy remains unexplored. Individual differences in trait empathy modulate how we process social stimuli, does empathy modulate music processing as well?

(Valence, as used in psychology, especially in discussing emotions, means the intrinsic attractiveness/”good”-ness (positive valence) or averseness/”bad”-ness (negative valence) of an event, object, or situation. The term also characterizes and categorize specific emotions. For example, emotions popularly referred to as “negative”, such as anger and fear, have negative valence. Joy has positive valence. Positively valenced emotions are evoked by positively valenced events, objects, or situations. The term is also used to describe the hedonic tone of feelings, affect, certain behaviors (for example, approach and avoidance), goal attainment or nonattainment, and conformity with or violation of norms. Ambivalence can be viewed as conflict between positive and negative valence carriers.)

If we consider music through a social psychological lens, it is plausible that individuals with a greater dispositional capacity to empathize with others might also respond to music’s social stimulus differently on a neurophysiological level by preferentially engaging brain networks previously found to be involved in trait empathy.

In this article, we test this hypothesis in two experiments using functional magnetic resonance imaging (fMRI). In Experiment 1, we explore the neural correlates of trait empathy (as measured using the Interpersonal Reactivity Index) as participants listened to isolated instrument and vocal tones. In Experiment 2. excerpts of music in four conditions (familiar liked/disliked. unfamiliar liked/disliked) were used as stimuli. allowmg us to examine correlations of neural activity with trait empathy in naturalistic listening contexts.

Measuring Trait Empathy

Trait empathy refers to the capacity for empathic reactions as a stable feature of personality. Individual differences in trait empathy have been shown to correlate with prosocial behavior and situational “state” empathic reactions to others.

Trait empathy is commonly divided into two components: emotional empathy is the often unconscious tendency to share the emotions of others, while cognitive empathy is the ability to consciously detect and understand the internal states of others.

There are a number of scales to measure individual differences in trait empathy currently in use. including the Toronto Empathy Questionnaire (TEQ). Balanced Emotional Empathy Scale (BEES). Empathy Quotient (EQ). Questionnaire of Cognitive and Affective Empathy (QCAE) and Interpersonal Reactivity INDEX (IRI). Here we use the IRI, which is the oldest and most widely validated of these scales and frequently used in neurophysiological studies of empathy.

The IRI consists of 28 statements evaluated on a 5 point Likert scale (from “does not describe me well” to “describes me very well”). It is subdivided into four subscales meant to tap different dimensions of self reported emotional and cognitive empathy. Emotional empathy is represented by two subscales: the empathic concern scale (hereafter EC) assesses trait level “other oriented” sympathy towards misfortunate others. and the personal distress scale (PD) measures “self oriented” anxiety and distress towards misfortunate others. The two cognitive empathy subscales «most of perspedive taking (PT). or the tendency to see oneself from another’s perspective, and fantasy (PS). the tendency to imaginatively project oneself into the situations of fictional characters.

Music and Empathy

Theories of empathy have long resonated with the arts. The father of the modern concept of empathy, philosopher Theodor Lipps originally devised the notion of Emfühlung (“feeling into”) in order to explain aesthetic experience. Contemporary psychological accounts have invoked mirror neurons as a possible substrate supporting Lipps’s “inner imitation” theory of the visual and performing arts. However. the incorporation of psychological models of empathy in empirical music research is still in its early stages, Empathy remains an ambiguous concept in general. but applications to music can appear doubly vexed in an influential formulation.

Esenherg et al (1991) define empathy as. “an emotional response that stems from another’s emotional state or condition and is congruent with the other’s emotional state or condition.” Aspects of this definition. though. might seem incongruous when applied to music. which is inanimate and not capable of possessing an emotional “state”. To connect music processing to trait empathy. therefore. it is first necessary to determine the extent to which music comprises a social stimulus, who or what do we empathize with when listening to music?

Scherer and Lentner proposed that empathy toward music is often achieved via identification and sympathy with the lived experiences and expressive intentions of composers and performers. Corroborating this view, in a large web based experiment Egermann and McAdams found that “empathy for the musician” moderated between recognized and induced emotions in music: the greater the empathy. the more likely an individual was to exhibit a strong affective response when listening.

In a related study Wollner presented participants with video of a string quartet performance in three conditions audio/visual, visual only. and audio only and reported a significant correlation between trait empathy measures and perceived expressiveness in both visual conditions (music only condition was non significant), leading him to conclude: “since music is the audible outcome of actions, empathic responses to the performer‘s movements may enhance the enjoyment of music.“ Similarly, Taruiti et al found correlations between the EC and FS scales of the IRI and accuracy in emotion recognition relative to musicians’ self reported expressive encodings in an audio only task.

A music specific manifestation of trait empathy was proposed by Kreutz et al, who defined “music empathizmg“ as a cognitive style of processmg music that privileges emotional recognition and experience over the tendency to analyze and predict the rules of musical structure (or. “music systematizing”). Garrido and Schubert compared this “music empathy” scale alongside the IRI EC subscale in a study exploring individual differences in preference for sad music. They found that people who tend towards music empathizing are more likely to enjoy sad music; however. high trait empathy was not significantly correlated with enjoyment of sad music. This would seem to suggest that the music empathizing cognitive style differs from general trait empathy,

A number of other studies have investigated the relationship between trait empathy and enjoyment of sad music using the IRI. In a series of experiments. Vuoskoski and Eerola reported statistically significant correlations between EC and FS subscales and self reported liking for sad and tender music. Similarly, Kawakami and Katahora found that FS and PT were associated with preference for and intensity of emotional reactions to sad music among children.

There is evidence that musical affect is often achieved through mechanisms of emotional empathy. According to this theory, composers and performers encode affective gestures into the musical signal, and listeners decode that signal by way of mimetic, mirroring processes; musical expression is conveyed transparently as affective bodily motions are internally reenacted in the listening process. Shubert, in his Common Coding Model of Prosocial Behavior Processing, suggests that musical and social processing draw upon shared neural resources: music. in this account, is a social stimulus capable of recruiting empathy systems, including the core cingulate paracingulate supplementary motor area (SMA) insula network, along with possible sensorimotor, paralimbic and limbic representations. The cognitive empathy component, which can be minimal, is involved primarily in detecting the aesthetic context of listening, enabling the listener to consciously bracket the experience apart from the purely social. This model may help account for the perceived “visuality” of musical experience, whereby music is commonly heard as manifesting the presence of an imagined other.

In sum, trait empathy appears to modulate self reported affective reactions to music. There is also peripheral psychophysiological evidence that primed situational empathy may increase emotional reactivity to music. It is plausible that such a relationship is supported by shared social cognitive mechanisms that enable us to process music as a social stimulus; however, this hypothesis has not yet been explicitly tested at the neurophysiological level.

Neural Correlates of Trait Empathy

Corroborating the bipartite structure of trait empathy that appears in many behavioral models of empathy, two interrelated but distinct neural “routes” to empathy have been proposed, one associated with emotional contagion and the other with cognitive perspective taking. Emotional empathy is conceived as a bottom up process that enables “feeling with someone else” through perception action coupling of affective cues. Such simulation or “mirroring” models maintain that empathy is subserved by the activation of similar sensorimotor, paralimbic and limbic representations both when one observes another and experiences the same action and emotional state oneself. This proposed mechanism is generally considered to be pre reflective and phylogenetically ancient; it has also been linked behaviorally to emotional contagion, or the propensity to “catch” others’ feeling states and unconsciously co experience them. For example. several imaging studies have found evidence for shared representation of observed/experienced pain in anterior cingulate and anterior insula, as well as somatosensory cortex. Similarly, disgust for smells and tastes has been shown to recruit the insula during both perception and action, and insula has been proposed as a relay between a sensorimotor fronto parietal circuit with mirror properties and the amygdala in observation and imitation of emotional facial expressions. There is also evidence that insula functions similarly in music induced emotions, particularly involving negative valence.

In contrast to emotional empathy, trait cognitive empathy has been conceived as a deliberative tendency to engage in top down, imaginative transpositions of the self into the “other’s shoes,” with concomitant reliance upon areas of the brain associated with theory-of-mind, executive control, and contextual appraisal, including medial, ventral and orbital parts of the prefrontal cortex; somatomotor areas; temporoparietal junction; and precuneus/postenor cingulate. As implied in the functional overlap between certain emotional and cognitive empathy circuits, some have argued that the two routes are neither hierarchical nor mutually exclusive: cognitive perspective taking is premised upon emotional empathy, though it may, in turn, exert top down control over contagion circuits, modifying emotional reactivity in light of contextual cues and more complex social appraisals.

Brain studies have converged upon the importance of the human mirror neuron system in action understanding, imitation and empathy, and has been demonstrated in multiple sensorimotor domains, including the perception of action sounds. Mirror properties were initially reported in the inferior frontal gyrus (IFG) and the inferior parietal lobule (IPL); consistent with simulation theories of trait empathy. Moreover, activity in these and other sensorimotor mirror circuits has been found to correlate with IRI scales in a variety of experimental tasks, including viewing emotional facial expressions; and video of hands injected with a needle. That is, high empathy people tend to exhibit greater activation in mirror regions during the observation of others. Simulation mechanisms also appear to underpin prosocial decision making.

Implication of inferior frontal and inferior parietal mirror neuron areas is not a universal finding in the empathy literature, and some have suggested that it may reflect specific socially relevant tasks or stimulus types, not empathy in and of itself. However, evidence for mirror properties in single cells of the primate brain now exists in medial frontal and medial temporal cortex, dorsal premotor and primary motor cortex, lateral intraparietal area, and ventral intraparietal area. This means that in brain imaging data the activity of multiple brain areas may potentially be driven by cells with mirror properties.

In addition to studies using visual tasks, auditory studies have revealed correlations between mirror neuron activity and trait empathy. Gazzola et al, for instance, reported increased premotor and somatosensory activity associated with PT during a manual action sound listening task. A similar link was observed between IFG and PD scores while participants listened to emotional speech prosody. To date, however, no studies have investigated whether individual differences in empathy modulate processing of more socially complex auditory stimuli, such as music.

Study Aim

To investigate the neural substrates underlying the relationship between trait empathy and music. we carried out two experiments using fMRI.

In Experiment 1, we focused on a Single low level attribute of musical sound timbre, or “tone color”, to investigate the effects of empathy on how listeners process isolated vocal and instrumental sounds outside of musical context.

We tested two main hypotheses:

First, we anticipated that trait empathy (measured with the IRI) would be correlated with increased recruitment of empathy circuits even when listening to brief isolated sounds out of musical context (Gazzola et al).

Second, following an embodied cognitive view of timbre perception (Wallmark et al), we hypothesized that subjectively and acoustically ”noisy” timbral qualities would preferentially engage the emotional empathy system among higher empathy listeners. Abrasive. noisy acoustic features in human and many non human mammal vocalizations are often signs of distress, pain, or aggression (Isai et al): such state cues may elicit heighted responses among people with higher levels of trait EC.

To explore the relationship between trait empathy and music processing, in Experiment 2 participants passively listened to excerpts of self selected and experimenter selected “liked” and “disliked” music in familiar and unfamiliar conditions while being scanned. Musical preference and familiarity have been shown to modulate neural response. Extending previous research on the neural mechanisms of empathy, we predicted that music processing would involve circuitry shared with empathic response in non musical contexts (Schubert).

Unlike Experiment 1, we had no a priori hypotheses regarding modulatory effects of empathy specific to each of the four music conditions However, we predicted in both experiments that emotional empathy scales (EC and FD) would be associated with regions of the emotional empathy system in music listening, including sensorimotor. paralimbic and limbic areas, while cognitive empathy scales (PT and FS) would primarily be correlated With activity in prefrontal areas implicated in previous cognitive empathy studies (Singer and Lamm).


Experiment 1 demonstrated that trait empathy is correlated With Increased activation of circuitry often associated with emotional contagion, including sensorimotor areas and insula, in the perception of isolated musical timbres. FS and EC also appear to be sensitive to the affective connotations of the stimuli. Timbre is arguably the most basic and quickly processed building block of music. Though sufficient to recruit empathy areas, these brief stimuli do not, however, constitute “music” per sci

In Experiment 2, we turned our focus to more naturalistic stimuli including excerpts of music selected in advance by participants in order to explore the effect of trait empathy on the processing of music.


The present study demonstrates that trait empathy is correlated with neurophysiological differences in music processing. Music has long been conceived as a social stimulus. Supporting this view, our study offers novel evidence that neural circuitry involved in trait empathy is active to a greater degree in empathic individuals during perception of both simple musical tones and full musical excerpts. Individual variances in empathy are reflected in differential recruitment of core empathy networks during music listening; specifically. IRI subscales were found to correlate with activity in regions associated with both emotional (e.g.. sensorimotor regions, insular and cingulate cortex) and cognitive empathy (cg, PFC. TPI) during passive listening tasks

Our main hypotheses were continued, though with an unexpected twist regarding the two putative empathy types (at least as structured by the MU), Both experiments seem to suggest interactions between bottom up and top down processes (indexed in our study by both IRI scores and activity in neural systems) in empathy modulated music listening. This is in line with recent findings in prosocial decision making studies. Stimulus type, however, seems associated with different patterns of neural systems engagement.

In Experiment 1, sensorimotor areas were more frequently modulated by trait empathy in the processing of musical timbre: conversely. in Experiment 2, cognitive areas were more frequently modulated by trait empathy in the processmg of (famillar) music. Together this suggests that, contrary to our initial hypothesis for Experiment 2, modulation of neural activity by empathy was driven more by stimulus type than by empathy type; that is. the emotional empathy subscale (BC) was no more selective to emotional contagion circuitry than cognitive empathy scales (PT and FS), and vice versa (the PD scale did not reveal any significant correlations with brain activity. In what follows, we interpret these results and discuss their implications.

Empathy-Modulated Sensorimotor Engagement in Timbre Processing

Using isolated 2-s instrument and vocal tones as stimuli, Experiment 1 found that the four IRI subscales modulated response to timbre. First, we found that cognitive perspective was correlated with activity in motor areas SMA for A0: and SI and anterior cingulate (ACC).

This finding is in line with numerous studies suggesting a role for ACC and SI in emotional empathy; it also replicates a result of Gazzola et al 2006. who reported a correlation of somatomotor activity and PT scores in an action sound listening task. Activity in these regions may suggest a sensorimotor simulation process whereby high PT individuals imitate internally some aspect of the production of these sounds, This result could be explained in light of Cox’s 2016 “mimetic hypothesis,” according to which music is understood by way of covert or overt motor reenactments of sound producing physical gestures. It is quite conceivable that people who are inclined to imagine themselves from others’ perspectives also tend to take up the physical actions implied by others’ musical sounds, whether a smooth and gentle voice, a growled saxophone, or any other musical sound reflecting human actions.

It is intriguing, however, that PT was not implicated in the processing of positive or negative valence. One might assume that perspective takers possess a neural preference for “good“ sounds: for example, one study reported activation of larynx control areas in the Rolandic operculum while subjects listened to pleasant music (but not unpleasant), suggesting subvocalization only to positively valenced music (K elsch et al) Our results, however. indicate that PT is not selective to valence in these sensorimotor areas.

FS also revealed motor involvement (SMA) in the task > baseline contrast. Unlike PT, FS appeared to be sensitive to both positive and negative valence of timbres: we found activity in left TH and Broca‘s area of the IFG associated with positively valenced timbres, and temporal, parietal and prefrontal activations associated with disliked timbres. TPI is an important structure for theory of mind. Together with Broca‘s area, a well studied language and voice specific motor region that has been implicated in emotional empathy. It is plausible to suggest that individuals who are prone to fantasizing may exhibit a greater tendency to attribute mental states to the virtual human agents responsible for making musical sounds, and that this attribution would be more pronounced for positively valenced stimuli.

As hypothesized. EC was correlated with activation in a range of areas previously implicated in empathy studies, including IPL, IPG and SMA, along With SI, STG. cerebellum and AIC. It was also sensitive to negative valence: noisy timbres were processed with greater involvement from SMA in individuals with higher EC. EC is an “other oriented” emotional scale measuring sympathy or compassion towards the misfortune of others. Since noisy, distorted qualities of vocal timbre are an index of generally high arousal, negatively valenced affective states, we theorize that individuals with higher trait EC exhibited greater motor attunement owing to the ecological urgency typically signaled by such sound events.

In short, we usually deploy harsh vocal timbres when distressed or endangered (e.g., screaming or shouting), not during affectively positive or neutral low arousal states, and high empathy people are more likely to pick up on and simulate the affective motor implications of others in distress. Though our sensitivity to the human voice is especially acute, researchers have hypothesized that instrumental timbre can similarly function as a “superexpressive voice” via acoustic similarities to emotional vocal expression. Our result would seem to support this theory, as motor response appears to encode the combined effects of noisy tones, both vocal and instrumental.

It is also worth noting, as might be expected given the above, that noisy voice produced a unique signature of activation among high FS and EC participants relative to the normal vocal stimuli: FS modulated processing of the noisy voice in SII and IPL, while EC was selective to noisy vocal sounds in the SMA and primary motor cortex. This result appears to be at odds with other studies of vocal affect sensitivity that report motor mimetic selectivity for pleasant vocalizations. It is likely that individual variances in empathy (plus other mediating factors) predispose listeners to differing orientations towards others’ affective vocalizations, with empathic listeners more likely to “catch” the motor affective implications of aversive sounds than low empathy people, who might only respond to sounds they find pleasant while tuning out negatively valenced vocalizations.

Cox (2016) theorizes that music can afford listeners an “invitation” for motor engagement, which they may choose to accept or decline, Seen from this perspective, it is likely that individual differences in empathy play an important role in determining how we choose to respond to music’s motor invitations.

Regarding motor engagement across IRI subscales it is apparent that SMA is the most prominent sensorimotor area involved in empathy modulated processing of timbre. SMA is a frequently reported yet undertheorized part of the core empathy network; it has also been implicated in internally generated movement and coordination of action sequences, and has been shown in a single neuron study to possess mirror properties. Most relevant to the present study, moreover, SMA contributes to the vividness of auditory imagery, including imagery for timbre. Halpern et al and Lima et al attributed SMA activity in an auditory imagery task in part to subvocalization of timbral attributes, and the present study would seem to partially corroborate this explanation. We interpret this result as a possible instance of sensorimotor integration: SMA activity could reflect a basic propensity to link sounds with their associated actions, which are internally mirrored while listening. In accordance with this view, we would argue that people do not just passively listen to different qualities of musical timbre, they enact some of the underlying physical determinants of sound production, whether through subvocalization, biography specific act sound associations.

To summarize, sensorimotor areas have been implicated in many previous studies of emotional empathy, including IFG and IPL; “pain circuit” areas in AIC and ACC; and somatomotor regions. Interestingly. these precise regions dominated results of the Experiment 1 timbre listening task. This is true, moreover, for both emotional and cognitive scales: PT and FS, though often implicated in cognitive tasks, were found in this experiment to modulate SMA, SI, primary motor cortex, IPL, MC and IFG, well documented motor affective AREAS. We theorize that the contextual impoverishment and short duration of the timbre listening task (2-s isolated tones) may have largely precluded any genuine perspective taking or fantasizing from occurring, it is much harder to put oneself in the “shoes” of an single isolated voice or instrument, of course, than it is an affectively rich piece of actual music. However, even in the absence of conscious cognitive empathizing, which presumably would have been reflected in engagement of the cognitive empathy system, individuals with high trait PT and FS still showed selective activations of sensorimotor and affective relay circuits typically associated with emotional empathy. This could be interpreted to suggest that the two “routes” to empathy are not dissociated in music listening: although conscious PT in response to abbreviated auditory cues is unlikely, people who frequently imagine themselves in the positions of others also exhibit a tendency toward motor resonance in this basic listening task, even when musical context is missing.

Prefrontal and Reward Activation During Music Listening

Experiment 2 used 16-s excerpts of self and experimenter selected music to explore the effect of dispositional empathy on the processing of music in four conditions. familiar liked (FL), familiar disliked (FD), unfamiliar liked (UL), and unfamiliar disliked (UD). Participants consisted of individuals who reported regularly experiencing intense emotional reactions while listening to music. Musical liking is associated at the group level (ie. no IRI covariates), with left basal ganglia reward areas, and disliking with activity in right AIC, primary auditory cortex and prefrontal areas (OFC and VLPFC). Musical familiarity is associated with activation across a broad region of the cortex, subcortical areas, and cerebellum, including IPL. premotor cortex and the core empathy network, while unfamiliarity recruits only the SFG.

This robust familiarity effect is even more acute among high empathy listeners: after adding empathy covariates to our analysis, there were no regions that demonstrated an affect specific response after controlling for familiarity. This result is consistent with the literature in showing a large neurophysiological effect of familiarity on musical liking; it appears that trait empathy, as well, modulates responses to familiar music to a greater degree than unfamiliar music.

Contrary to expectations, activation in regions primarily associated with emotional empathy (e.g., sensorimotor areas, ACC, AIC) was not a major component in empathy modulated music processing. Instead. the most prominent activation sites for PT and EC scales were prefrontal, including medial, lateral, and orbital portions of the cortex, as well as TPJ. These regions are involved in executive control, regulation of emotions, mentalizing, contextual appraisal, and “enactment imagination”, and have figured prominently in many studies on the neurophysiology of cognitive empathy. Additionally, FS and EC results were characterized by dorsal striatum when participants listened to familiar music. This basal ganglia structure has been frequently reported in empathy studies but not often discussed; it has also long been associated with musical pleasure.

basal ganglia

Replicating this association, our results suggest that empathic people experience a higher degree of reward and motivation when listening to familiar music compared to lower empathy people.

PT was associated with left TPI in the task > baseline contrast. Activation of this region among perspective takers is consistent with studies implicating TPI in theory of mind and the merging of self and other (Lawrence et a1, 2006). The TPI was joined by posterior cingulate, cerebellum and superior prefrontal areas when listening to familiar liked music (FL > FD), the former two of which were also identified in a study on the neural bases of perspective taking. Interestingly, these results differ substantially from the PT correlations in Experiment 1, which were entirely sensorimotor. In the context of isolated musical sounds, PT results were interpreted as a reflection of covert imitation (or, enactive perspective taking): in contrast. however, it appears here that PT may reflect a more cognitively mediated, mental form of perspective taking, which conceivably extends beyond action perception coupling of musicians‘ affective motor cues to encompass contextual appraisal, assessments of the affective intent embodied in the music, and other executive functions.

In contrast to the prominent TPI and prefrontal activation associated with PT, FS results revealed activation of dorsal striatum (caudaie and putamen) and limbic areas (thalamus, hippocampus and amygdala). Activation of reward and emotion centers may suggest that fantasizers also tend to exhibit heightened positive emotional reactions to familiar music. Indeed, we found a moderate correlation between FS and preference ratings for familiar liked music, which may tentatively corroborate this claim. Moreover, structural brain studies have found that FS is associated with increased gray matter volume in hippocampus, an important memory area, perhaps also indicating enhanced encoding of familiar liked music among fantasizers.

The contrast in activation between the two IRI cognitive empathy scales (PT and FS) is notable, and may be attributed to the different aspects of empathy they were designed to assess. PT taps the tendency to imagine oneself in other people’s shoes, whereas FS captures the tendency to imagine oneself from the perspective of fictional characters. With this distinction in mind, one could surmise that the two scales also tap different views regarding the ontology of the musical agent: in this reading, people with high trait PT are more likely to take music as a social stimulus, i.e., as if it was a real or virtual human presence (with theory of mind, goals, beliefs), while high FS listeners are more likely to hear it as “fictional” from a social perspective, i.e., as a rewarding sensory stimulus with an attenuated grip on actual social cognition. Further research is called for to explore possible explanations for the differences in cognitive scales as reflected in music listening.

Turning finally to emotional empathy, we found that EC recruits prefrontal, reward and sensorimotor affective areas in music listening, and is likewise quite sensitive to familiarity. In the Familiar > Unfamiliar contrast, we found activation of cerebellum, IPL, DLPFC, IFG, DMPFC, amygdala, anterior paracingulate, dorsal striatum, OFC and lingual gyrus, and a variation on this general pattern for the Familiar liked > Unfamiliar liked and interaction contrasts. Activation of bilateral IPL and IFG is consistent with mirror accounts of empathy. Furthermore, the ACC, paracingulate, and areas that extend dorsally (SMA, DMPFC) have been proposed as the core of the empathy network: our result would seem to extend support for the primacy of this region using an experimental task that is not explicitly social in the manner of most empathy studies. Lastly, DLPFC is an important executive control area in cognitive empathy, and has been implicated in emotional regulation. Activation of this region may reflect top down control over affective responses to familiar music, both in terms of up regulation to liked music and down regulation to disliked (or possibly up regulation to negative stimuli, as Open minded empathic listeners try to “see something positive” in the disliked music), In further research, connectivity analysis between DLPFC and limbic/reward areas may help to specify the neurophysiological mechanisms underlying empathy modulated emotional regulation during music listening.

In addition to motor, cingulate and prefrontal activity, we found the recruitment of emotion and reward processing areas as a function of EC and musical familiarity: dorsal striatum (the whole extent of the caudate nucleus, plus thalamus) may reflect increased pleasure in response to familiar music among empathic listeners. It is not surprising that the reward system would show preferential activation to familiar music, as confirmed in the basic group Liked > Disliked contrast.

Prevalence of basal ganglia for both EC and FS suggests that trait empathy may effectively sensitize people to the music they already know. This even appears to be the case for disliked music, which showed dorsal striatum activation (along with OFC) in the Familiar disliked > Unfamiliar disliked contrast. This could be interpreted to indicate that empathic people may experience heightened musical pleasure even when listening to the music they self select as “hating,” provided it is familiar. By way of contrast. no striatum activation was found for any of the unfamiliar music conditions.

In concert with limbic circuitry, then, it is apparent that musical familiarity recruits a broad region of the affect reward system in high EC listeners.

Activation of inferior parts of the lingual gyrus and occipital lobe was another novel finding, and may also be linked to musical affect. These areas are associated with visual processing, including perception and recognition of familiar sights and emotional facial expressions, as well as visual imagery. It is reasonable to think that empathic listeners may be more prone to visual imagery while listening to familiar music. Visual responses are an important mechanism of musical affect more generally, and are a fairly reliable index of musical engagement and attention.

If high EC people are more susceptible to musical affect, as suggested by our results, they may also show a greater tendency towards visual imagery in music listening.

To be clear, we did not explicitly operationalize visual imagery in this study: in the future, it would be interesting to follow up on this result by comparing visual imagery and music listening tasks using the EC scale as a covariate.

The behavioral data resonate in interesting and sometimes contradictory ways with these imaging findings. We found that EC is strongly associated with preference for liked music and unfamiliar music, and negative responses to familiar disliked music. Results suggest that high EC people are more responsive to the affective components of music, as reflected in polarity of preference responses. EC was also associated with open mindedness to new music (i,e.. higher ratings for unfamiliar music), though imaging results for this contrast did not reach significance, and might appear to be contradicted by the clear familiarity effect discussed previously.

We must be cautious in the interpretation of these findings owing to the small sample size, but this resonance between behavioral and imaging evidence is nonetheless suggestive in demonstrating a role for EC in affective responsiveness to familiar music. This conclusion is broadly consistent With previous behavioral studies, especially regarding pleasurable responses to sad music.

In sum, the present results provide complementary neural evidence that involvement of prefrontal areas and limbic/basal ganglia in music listening covaries with individual trait differences in empathy, with sensorimotor engagement playing a smaller role.

How do we account for the prominence of cognitive, prefrontal areas in music listening but not musical timbre in isolation? It must be noted that a broad swath of the emotional empathy system was involved in the basic task > baseline contrast (used to mask all IRI covariates): in other words, it is clear that music in aggregate is processed with some level of sensorimotor, paralimbic, and limbic involvement, regardless the empathy level of the listeners or the valence/familiarity of the music. However, our results seem to suggest that empathic people tend to be more attuned to the attribution of human agency and affective intention in the musical signal. as indicated by preferential engagement of cognitive empathy networks including PFC (MPF and DLPFC) and TP), as well as reward areas.

In other words, what seems to best characterize the high empathy response to musical stimuli is the tendency to take an extra cognitive step towards identification with some agentive quality of the music, over and above the work of emotional contagion mechanisms alone.

Thus while patterns of neural resonance consistent with emotional contagion appear to be common to most experiences of music and were also found among high empathy participants in Experiment 1, activation of prefrontal cognitive empathy systems for the PT and EC scales may indicate the tendency of empathic listeners to try to “get into the heads” of composers, performers, and/or the virtual persona of the music. This top down process is effortful, imaginative, and self aware, in contrast to the automatic and pre reflective mechanisms undergirding emotional contagion. Accordingly, as suggested by Schubert, the involvement of cognitive systems may not strictly speaking be required for affective musical response, which can largely be accounted for by emotional contagion circuitry alone.

A number of studies have shown that mental imagery may be supported by sensorimotor and affective components without the contribution of prefrontal areas. Nevertheless, they could betoken a more social cognitive mode of listening, a deliberative attempt on the part of listeners to project themselves into the lived experience of the musical agent. This imaginative projection is more intense, understandably, for music that empathic people already know, and also appears to interact with musical preference.

General Implications

The present study has a number of implications for social and affective neuroscience, music psychology, and musicology. For neuroscientific empathy research, we demonstrate the involvement of the core empathy network and mirror neuron system outside of tasks that are explicitly social cognitive. Most studies use transparently social experimental tasks and stimuli to assess neural correlates of state and trait empathy; for example, viewing pictures or videos of other people.

This study demonstrates that musical sound. which is perhaps not an obvious social stimulus, can elicit neural responses consistent with theories of empathy. By domg so, this study highlights the potential value of operationalizing artistic and aesthetic experience as a window into social cognitive and affective processing, a perspective that is arguably the historical progenitor of contemporary empathy research.

For music psychology, this research has at least three main implications.

First, this study demonstrates that trait empathy may modulate the neurophysiology of music listening. Although there is mounting behavioral and psychophysiological evidence pointing to this conclusion, this is the first study to investigate the effects of empathy on the musical brain.

Second, this study confirms and extends empirical claims that music cognition is inextricably linked to social cognition. Our results suggest that aspects of affective music processing can be viewed as a specialized subprocess of general social affective perception and cognition. This may begin to explain the neural bases for how music can function as a “virtual social agent”.

Third, in demonstrating neural differences in music processing as a function of empathy, we highlight the possible significance of looking at other trait features when assessing the functional neural correlates of musical tasks and stimuli. Many neurophysiological music studies take only a few trait features into account in sampling procedures and analysis, most notably sex, age, and musical training: the latter has been well explored, but other factors such as personality factors and mood are not frequently addressed. Individual differences in music processing may relate to dispositional characteristics that can be captured by psychosocial questionnaires, indirect observational techniques, or other methods. Exploring the role of such trait variables in musical behaviors and brain processing could provide a more detailed and granular account of music cognition,

Finally, these results enrich the humanistic study of music in providing a plausible psychobiological account for the social valence of musical experience observed in diverse cultural and historical settings. As music theorist Clifton claims, “the ‘other’ need not be a person: it can be music.”

In a very rough sense, this study provides empirical support for this statement: areas implicated in trait empathy and social cognition also appear to be involved in music processing, and to a significantly greater degree for individuals with high trait empathy.

If music can function something like a virtual ”other,” then it might be capable of altering listeners‘ views of real others, thus enabling it to play an ethically complex mediating role in the social discourse of music. Indeed, musicologists have historically documented moments of tense cultural encounter wherein music played an instrumental role in helping one group to realize the other’s shared humanity.

Recent research would seem to provide behavioral ballast for this view: using an implicit association task,Vuoskoski et a showed that listening to the music of another culture could positively modulate attitudes towards members of that culture among empathic listeners. Though we do not in this study explicitly address whether music can alter empathic brain circuits, it is suggestive that certain attitudes toward musical sound may have behavioral and neural bases in individual differences in trait empathy.


A few important limitations must be considered in interpreting these results, First, this study was correlational: no causative links can thus be determined in the relationship between music and trait empathy. In the future, it would be interesting to use an empathy priming paradigm in an MRI context to compare neurophysiological correlates of trait empathy with primed “state” empathy in music listening; this could provide a powerful method for disentangling possible differences in processing between dispositional attributes of empathy and contextual factors (e.g., socially conditioned attitudes about a performer, mood when listening).

As a corollary to the above, moreover, this study does not address whether our results are specific to music listening: perhaps high empathy people utilize more of these areas when performing other non musical yet not explicitly social tasks as well (eg. viewing abstract art). Additionally. we do not explore whether there could be other mediating trait factors in music processing besides empathy and sex: personality and temperament, for instance, have been shown to modulate responses to music.

Finally, this study will need to be replicated with a larger sample size, and with participants who do not self select based on strong emotional reactions to music, in order to strengthen the statistical power and generalizability of the results.


In two experiments using fMRI, this article demonstrates that trait empathy modulates music processing. Replicating previous findings in the social neuroscience literature, isolated musical timbres are related to sensorimotor and paralimbic activation; in actual MUSIC listening, however, empathy is primarily associated with activity in prefrontal and reward areas. Empathic participants were found to be particularly sensitive to abrasive, “noisy” qualities of musical timbre, showing preferential activation of the SMA, possibly reflecting heightened motor mimetic susceptibility to sounds signaling high arousal, low valence affective states.

In the music listening task. empathic subjects demonstrated enhanced responsiveness to familiar music, with musical preference playing a mediating role. Taken together, these results confirm and extend recent research on the link between music and empathy, and may help bring us closer to understanding the social cognitive basis for music perception and cognition.



Davis, M. H. (1980). A multidimensional approach to individual differences in empathy. JSAS Catalog of Selected Documents in Psychology, 10, 85.

Description of Measure:

Defines empathy as the “reactions of one individual to the observed experiences of another (Davis, 1983).”

28 items answered on a 5 point Likert scale ranging from “Does not describe me well” to “Describes me very well”. The measure has 4 subscales, each made up of 7 different items. These subscales are (taken directly from Davis, 1983):

Perspective Taking, the tendency to spontaneously adopt the psychological point of view of others.

Fantasy taps respondents‘ tendencies to transpose themselves imaginatively into the feelings and actions of fictitious characters in books, movies, and plays

Empathic Concern assesses “other-oriented” feelings of sympathy and concern for unfortunate others

Personal Distress measures “self oriented” feelings of personal anxiety and unease in tense interpersonal settings

Abstracts of Selected Related Articles:

Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113126.

The past decade has seen growing movement toward a view of empathy as a multidimensional construct. The Interpersonal Reactivity Index (IRI; Davis, 1980), which taps four separate aspects of empathy, is described, and its relationships with measures of social functioning, self esteem, emotionality, and sensitivity to others is assessed. As expected, each of the four subscales displays a distinctive and predictable pattern of relationships with these measures, as well as with previous unidimensional empathy measures. These findings, coupled with the theoretically important relationships existing among the four subscales themselves, provide considerable evidence for a multidimensional approach to empathy in general and for the use of the IRI in particular.

Pulos, S., Elison, J ,, & Lennon, R. (2004). Hierarchical structure of the Interpersonal Reactivity Index. Social Behavior and Personality, 32, 355 360.

The hierarchical factor structure of the Interpersonal Reactivity Index (IRI) (Davis, 1980) inventory was investigated with the Schmid Leiman orthogonalization procedure (Schmid & Leiman, 1957). The sample consisted of 409 college students. The analysis found that the IRI could be factored into four first order factors, corresponding to the four scales of the IRI. and two second order orthogonal factors, a general empathy factor and an emotional control factor.


The following statements inquire about your thoughts and feelings in a variety of situations. For each item, indicate how well it describes you by choosing the appropriate letter on the scale at the top of the page: A, B, C, D, or E.

When you have decided on your answer, fill in the letter next to the item number.

READ EACH ITEM CAREFULLY BEFORE RESPONDING. Answer as honestly as you can. Thank you.







1. I daydream and fantasize, with some regularity, about things that might happen to me. (FS)

2. I often have tender, concerned feelings for people less fortunate than me. (EC)

3. I sometimes find it difficult to see things from the “other guy’s” point of View. (PT) (-)

4. Sometimes I don’t feel very sorry for other people when they are having problems. (EC) (-)

5. I really get involved with the feelings of the characters in a novel. (FS)

6. In emergency situations, I feel apprehensive and ill at ease. (PD)

7. I am usually objective when I watch a movie or play, and I don’t often get completely caught up in it. (FS) (-)

8. I try to look at everybody’s side of a disagreement before I make a decision. (PT)

9. When I see someone being taken advantage of, I feel kind of protective towards them. (EC)

10. I sometimes feel helpless when I am in the middle of a very emotional situation. (PD)

11. I sometimes try to understand my friends better by imagining how things look from their perspective. (PT)

12. Becoming extremely involved in a good book or movie is somewhat rare for me. (FS) (-)

13. When I see someone get hurt, I tend to remain calm. (PD) (-)

14. Other people’s misfortunes do not usually disturb me a great deal. (EC) (-)

15. If I‘m sure I’m right about something, I don’t waste much time listening to other people’s arguments. (PT) (-)

16. After seeing a play or movie, I have felt as though I were one of the characters. (FS)

17. Being in a tense emotional situation scares me. (PD)

18. When I see someone being treated unfairly, I sometimes don’t feel very much pity for them. (EC) (-)

19. I am usually pretty effective in dealing with emergencies. (PD) (-)

20. I am often quite touched by things that I see happen. (EC)

21. I believe that there are two sides to every question and try to look at them both. (PT)

22. I would describe myself as a pretty soft hearted person. (EC)

23. When I watch a good movie, I can very easily put myself in the place of a leading character. (FS)

24. I tend to lose control during emergencies. (PD)

25. When I’m upset at someone, I usually try to “put myself in his shoes” for a while. (PT)

26. When I am reading an interesting story or novel, I imagine how I would feel if the events in the story were happening to me. (FS)

27. When I see someone who badly needs help in an emergency, I go to pieces. (PD)

28. Before criticizing somebody, I try to imagine how I would feel if I were in their place. (PT)

See also:

Music and the Mind

by Anthony Storr

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s