The preceding 32 chapters of this volume reveal the healthy state of the current field of music and emotion. In this final chapter, we take the opportunity—benefiting from the overview that the volume is offering—to comment on the history of the field, summarize current trends, and propose future directions for research.

33.1 The Past

A comprehensive, multidisciplinary history of the field of music and emotion remains to be written and represents a major undertaking, which we leave for a possible chapter in a future volume. Budd (1985) offered a useful survey of some of the historical ideas in philosophical thought on the topic—which continue to resonate in current (p. 934 ) philosophical work (reviewed by Davies, this volume)—and Cook and Dibben (this volume) provide a whistle-stop tour of the key ideas in musicological thought since ancient times, especially since 1600. In this section, we will mainly restrict our comments to the development of music and emotion as a sub-field of music psychology. Even a cursory history of the field may be useful to explain the current state of the field and inform speculations about the future.

Despite the fact that the first studies of emotion in music coincided with the advent of psychology as an independent discipline in the late nineteenth century (e.g. Downey, 1897; Gilman, 1891; Weld, 1912), the emergence of music and emotion as a separate sub-field was not going to be easy: Apart from an early ‘peak’ of studies in the 1930s and 1940s, including the seminal work by Kate Hevner (1935), Melvin Rigg (1940), and Carl Seashore (1938), the area would soon be sidelined by other areas. The early work on musical emotions was mainly experimental, descriptive, and concerned with perception of emotion rather than induction of emotion (cf. Table 1.2, Chapter 1, this volume). Typical studies focused on self-report, asking subjects to match verbal labels to pieces of music—also relating such matching to individual differences (for reviews, see Gabrielsson & Juslin, 2003; and Gabrielsson & Lindström, this volume). However, the dominant trends in music psychology in the early twentieth century concerned more ‘basic’ psychophysical and perceptual processes, reflecting the regarding of the natural sciences as the ideal for ‘the new behavioural science’. The strive for experimental control and the ‘bottom-up’ approach to psychology were not beneficial for an understanding of how listeners actually experience music (e.g. emotionally).

The subsequent trends in psychology more generally (for a recent review, see Goodwin, 2008) did little to change this situation: After first having to endure the ‘emotion-banning’ era of Behaviourism (e.g. Skinner, 1953), and then the ‘cognitive revolution’ (Gardner, 1985), the field barely survived until the eighties, when the tide slowly began to turn. During most of the history of music psychology, musical emotion studies were conducted by a few pioneers, with little or no connection to the broader field of affect. To be sure, each of the following decades would see the publication of what are today regarded as ‘classic’ books on the topic (by Meyer, 1956; Berlyne, 1960, 1971; Clynes, 1977). However, these books were largely isolated efforts which did not succeed in bringing music and emotion research into the ‘mainstream’ of music psychology; for instance, Meyer’s work did not stimulate emotion-related work until the early 1990s (Sloboda, 1991) while Berlyne’s work was not revived until the 1980s (Konečni, 1982) and 1990s (Hargreaves & North, 1997). Well into the 1980s, the Zeitgeist in music psychology was mainly characterized by experimental perceptual and cognitive research (as saluted in the influential book by Sloboda, 1985), in keeping with the origins of music psychology.

As noted by Sloboda and Juslin (this volume), the breakthrough came in the late 1980s and early 1990s—driven by many of the authors featured in the present book. Several factors contributed to this trend. One crucial factor was the blossoming of another, related field—the social psychology of music (e.g. Hargreaves & North, 1997)—which, arguably, moved music psychology away from the typical 1980s paradigm of laboratory-based experiments regarding cognitive processes to a broader exploration of the manifold ways in which music is used and experienced in everyday life—which (p. 935 ) in turn would contribute to the use of a broader range of methods in studying music experience. The music-emotion field also received a ‘boost’ from unexpected quarters, as the influential books by Damasio (1994) and LeDoux (1996; see also Panksepp, 1998) convinced neuroscientists that affect was perhaps after all worthy of serious attention—leading also to the initial neuroimaging studies of music and emotion (Blood et al, 1999). This trend culminated in the ‘Geneva Emotion Week’ 1998—devoted entirely to musical emotions—and the publication of the book Music and emotion (Juslin & Sloboda, 2001). The research carried out since then forms the basis for any current evaluation of the state of the field.

33.2 The Present

At the current stage, the topic of music and emotion has become a generally accepted field of research, as revealed by its inclusion in handbooks of music psychology (e.g. Hallam, Cross, & Thaut, 2009) and emotion psychology (e.g. Davidson, Scherer, & Goldsmith, 2003; Lewis, Haviland-Jones, & Barrett, 2008). (Even so, that ‘music’ is not taken quite as seriously as the other sub-topics is revealed by the fact that, although the other chapters in the last-mentioned handbook were written by experts in their respective fields, the chapter on music was written by two researchers who have not conducted any studies on emotion in music—Johnson-Laird & Oatley, 2008). At the time of writing, three conferences focusing on musical emotion were planned for the summer of 2009, testifying to the continuing high profile of the topic.¹

What is the field of music and emotion of 2009 like? One characterization is obviously provided by preceding chapters of the present book. Judging from the content of this volume, how does the field of today differ from that of the past? Apart from the fact that it has grown in recent years, the field is characterized by a broader range of methods and genres of music than previously (the latter is presumably a result of a new generation of researchers entering the field); a stronger focus on induction of emotions, relative to perception of emotions; and increasing links to the developments in the ‘affective sciences’ more generally. One can also discern the emergence of further subdivisions of the field into areas such as ‘measurement’, ‘performance’, ‘neuroscience’, ‘music experience’, ‘development’, ‘music in everyday life’, ‘music preference’, and ‘applications’.

However, the field is still mainly descriptive rather than hypothesis-driven, which may suggest that the field has not yet quite reached maturity. Theoretical work is less common in the music and emotion domain than in the emotion domain generally (Davidson, Scherer, & Goldsmith, 2003). As far as conceptualizations of emotion are (p. 936 ) concerned, both the dominant approaches (categorical and dimensional) continue to exist, side by side, in current research: for example, in the current volume, Hargreaves and North adopt a one-dimensional approach in terms of arousal; Thaut and Wheeler, and Schubert, adopt two-dimensional models; others such Zentner and Eerola, Sloboda, and Juslin et al adopt categorical models.² Furthermore, it can be noted that much current research is method-driven rather than issue-driven, suggesting the need to provide a broader methodological training to a new generation of musical emotion researchers; they need exposure to methods beyond their usual academic specialties by means of workshops and summer schools on specific techniques.

Moreover, despite calls for more cross-cultural research (see Sloboda & Juslin, 2001), there are still very few cross-cultural studies in the field; similarly, previous calls for further multidisciplinary research and integration have not yet been met—that is, there are still few clear cases of interdisciplinary collaboration. There are some examples of interdisciplinary influences, however, as in the adoption of Gibsonian ecological psychology in musicological and sociological work (Cook & Dibben, this volume; DeNora, this volume); the adoption of sociologically inspired approaches in British music psychology (Sloboda, this volume), and the increasing influence of neuroscientific research in music therapy (Thaut & Wheeler, this volume). However, intensely interdisciplinary research of the kind observed in neuroscience (Brehm, 2008) is still rare in the music and emotion field.

To be fair, multidisciplinary collaboration may not be quite as easy as it appears. This relates to the thorny issues of what counts as ‘theory’ and ‘evidence’ in different disciplines. For instance, Becker (this volume) is arguing forcefully for an integration of humanistic and scientific approaches to music and emotion; she notes that ‘while the styles of argument and the criteria for evidence may remain distinct, the conclusions need to be comparable and not incommensurable’ (p. 144). How the conclusions can ever be comparable and commensurable when the criteria for evidence remain distinct is unclear, however. Most of the ‘dichotomies’ or ‘dimensions’ that we used to characterize the field in a previous commentary (cf. Sloboda & Juslin, 2001) are still evident in the field; for example, the stimulus-driven biological view adopted by Peretz (this volume) is distant from the sociological listener-as-agent perspective adopted by DeNora (this volume). Still, Juslin et al (this volume) propose that some of these differences can be reconciled by noting that emotional reactions to music involve processing at multiple (and partly independent) levels of the brain (e.g. subcortical structures that enable ‘automatic’ processing as well as ‘higher’ cortical structures related to self-consciousness and imagination). On this view, different disciplines complement each other, by providing distinct parts of the overall puzzle. But the question remains just how far interdisciplinary integration is possible, or even desirable.

Another way to capture the nature of the field is in terms of the articles and books cited most widely in the field. Thus, Table 33.1 shows a subset of the most frequently (p. 937 )

cited papers, as indicated by a Google Scholar search on 12 May 2009. We emphasize that such citation counts are imperfect measures of scientific impact, and that they may well tend to favour older publications and publications in relatively more ‘active’ sub-fields. Yet they offer some sense of which papers influence the field, one way or another. The articles and books listed in Table 33.1 could be—and probably already are—used as ‘key reading materials’ in courses on music and emotion. (These could be augmented by the ‘recommended further readings’ provided by the authors of the previous chapters in this book.)

33.3 The Future

33.4 Concluding Remarks

We finished our last commentary (Sloboda & Juslin, 2001) with the hope that a future book would report progress in a rather more self-conscious and self-confident manner (p. 462). In some ways, at least, we think this is true of the present volume: the music and emotion field of 2009 is more advanced than it was in 2001. But make no mistake, the field is still fraught with disagreement about many issues, such as which emotions music induces, whether there are uniquely musical emotions, and whether a categorical or dimensional approach provides the best account of such emotions. Because the topic of music and emotion is so close to the heart of why most of us engage with music, one sometimes gets the feeling that scholars are motivated by what they want musical emotions to be—rather than what they actually are for most people. Such motives may partly underlie oversimplified debates about whether music evokes ‘basic’ or ‘refined’ emotions. We predict that the field will slowly move away from such basic issues, similarly to what has happened in the ‘affective sciences’ more generally (Davidson et al, 2003). Perhaps we do not need to resolve the age-old question of whether emotions are discrete or dimensional, but rather can focus on under what circumstances they manifest as one or the other, or are most usefully measured in terms of one or the other. We believe that music-emotion researchers will soon begin to focus on more limited, but more precisely defined, topics which will, eventually, help to resolve some of the ‘grander’ issues that currently define (or perhaps stifle) the field.

One consideration, which could guide us more explicitly away from ‘time-honoured’ questions, is the notion of benefit: which groups or constituencies stand to benefit from the results of specific pieces of research? There is a tendency, which is by no means confined to music-emotion research, for researchers to define the goals in terms of their own professional worlds—the need to be innovative, rigorous, peer-appreciated, and understandable within the institutional and professional constraints and incentives that determine how research resources are allocated. There is of course nothing (p. 946 ) unacceptable about any of these goals, but if research is governed mainly by these goals, it will tend to be of relevance or interest primarily to other scholars. Perhaps music and emotion researchers could frame more research programmes in light of their potential benefit to a particular constituency within the world of music or within the wider worlds of education, health and healing, business and industry, culture and religion. Although there are some notable exceptions (particularly in the areas of health and therapy), the general impression is that applicability is either not considered at all, or is bolted on at the end—in the form of a rather tokenistic afterthought.

Sloboda (2005) argued that applicability should be a more explicit factor in strategic research planning—both at an individual and an institutional level. Is there music-emotion research that could be framed around specific practical concerns of composers, performing musicians, audiences, or music broadcasters? Can music-emotion research address important issues of human and cultural development? As Sloboda (2005) noted, the situation is not so much that disputed answers have been given to such questions. Rather, it is the case that most researchers do not even bother to pose or answer them in a serious way.

Retrospectively, it may turn out that the second half of the twentieth century was a rather privileged moment in the history of research, where many researchers were left free to ‘follow their noses’ wherever their individual interests took them. There are signs that the world may be entering a phase of its history when the opportunities and resources for rigorous research become significantly more constrained. As this happens, stakeholders (including taxpayers and governments) may ask more searching questions about the value of specific areas of research. Music-emotion researchers perhaps need to be ready with better answers than are currently apparent from most of their outputs!

Despite all this, the fact that our emotional experiences with music can be so incredibly rich, multi-faceted, and difficult to capture in any straightforward fashion is perhaps partly the reason why the field of music and emotion remains endlessly fascinating— and why we can be quite certain that this book will not be the last on this topic.

1. 1

   1. A. Strategies for bringing together scientific and cultural approaches to music and emotion.

   2. B. Music and altered states.

   3. C. Music and spirituality.

2. (p. 947 ) 2

   1. A. For understanding the role of music on emotion in the film/video context, developmental studies and longitudinal studies would be important in sorting out effects of declining plasticity with age versus effects of particular critical periods for acquiring representations of certain kinds of information.

   2. B. Emotion research benefits from fMRI, but the problem of the noise of the magnet and the small claustraphobia-inducing chamber provoke an unnatural emotional baseline. Emotion research will benefit when these two problems are eliminated. When this happens (which I expect it might in 5 years) the basic fMRI studies will need to be redone to see if in fact the results are the same. There are no (or very few) studies of the effects of music and video under fMRI, and these need to be done (some may be currently in progress, which would begin to pave the way).

   3. C. Longitudinal studies (that one as a researcher could begin early in life) would assist developmental models, and one should view every study as possibly a developmental study that someone else, if not oneself, could carry out again on the same participants several years, or even decades, down the road.

   4. D. The development of a system (digital database) for sharing stimulus resources and data through the Internet will also benefit progress on emotion research.

3. 3

   1. A. How far emotions are verbally (or even critically/musicologically) constructed.

   2. B. If emotional interpretations are constructed in the course of performance, the extent to which they are constrained by structural/referential/other features of the composition.

   3. C. How far emotions have a history, and the extent to which the study of music (scores, recordings, written responses, etc.) might shed light on that history (another way to put this: how far the study of music and emotion is a psychological or a musicological enterprise).

4. 4

   1. A. In terms of research I would like to see done: I’d be interested in more genuinely cross-cultural studies of expressiveness, involving cultures with no prior exposure to each other’s music (so excluding Westerners). The examples would have to be endorsed as clearly expressive by composers/performers of the music’s home culture. The questions would be about what if any emotions the music expresses (not what participants feel or what it makes them think of). And as follow-up, there could be studies about if and how rapidly they improve through increased exposure or training from those of the music’s home culture.

   2. B. I also have a reservation about the methodology of your (Juslin’s) experiments that this excludes the contribution of dynamic/structural elements in music’s expressiveness by holding the tune constant. So I’d be interested in experiments that reversed yours, by holding many of the prosodic elements constant but varying the dynamic/structural features of the music in appropriate ways.

5. 5

   1. A. Music, emotions, and collective action.

   2. B. Music, emotions, and health.

   3. C. Music, emotions, and social integration/disintegration.

6. 6

   1. A. Research should focus more on ‘vitality affects’ in music, that is, dynamic features such as crescendo/diminuendo, accelerando/ritardando, glissando, changes of timbre, and other types of sudden or gradual changes, further on the effects of formal features such as repetition, variation, condensation or transposition of the musical material. This requires further development of efficient techniques for continuous recording of perceived and/or felt emotion as well as refinement of the meaning of the concept ‘vitality affects’ in music.

   2. (p. 948 ) B. Free phenomenological (‘subjective’) description should be used much more as it often provides much more depth, insight, and explanation of the listener’s experience than crude techniques such as choice among adjectives, simple rating scales, etc.

   3. C. Research has so far mainly dealt with ‘basic’ emotions, but many ‘feelings’ are much more complex and include perceptual and cognitive components. The concept of ‘feelings’ has to be further analysed and refined.

7. 7

   1. A. The ethics and public policy implications of using music as a negative stimulus in situations ranging from crowd control to torture.

   2. B. More cross-cultural comparison of how the brain processes ‘consonance’ and ‘dissonance’ in societies that use scales with different musical intervals and have different sonic expectations.

   3. C. More research on the effects of different kinds of mediation in shaping the emotional response to the same piece of music.

8. 8

   1. A. A study on the long-term affective outcomes of active engagement with music during compulsory education through making or listening to music with particular reference to well-being, self-esteem, social skills, and a range of transferable skills including concentration, self-discipline, and being able to manage motivation.

   2. B. A study exploring the perceptions of pupils and teachers of the role of emotion in classroom music and the ways in which this can be used to promote greater enjoyment of music and relevance to experiences outside the classroom.

   3. C. Further work on the teaching of emotional communication in music performance, also extending it to include improvisation and composition.

9. 9

   1. A. Studies of neurological changes as a function of engaging in musical experiences.

   2. B. Studies of immunologial changes as a function of engaging in musical experiences.

   3. C. Analyses of correlations between psychological and physiological measurement of emotions as they are affected by music.

10. 10

    1. A. Examining relationships among specific music emotions and overarching terms such as Konečni’s ‘aesthetic awe’ or ‘being moved’. Could one be moved by the music without being able to identify specific emotions and, if so, is that still an emotional response? Embedded in this is a call for a movement away from simpler emotion words such as happy/sad toward more nuanced, complex emotions.

    2. B. Examining music emotions in longer pieces of music that may have varying expressive content (e.g. a Mahler symphony instead of a 30-second excerpt). If one can ‘track’ the various emotions in a complex piece as they shift suddenly or even overlap, does that support a cognitivist position, since it would be unlikely that a listener would actually experience a series of short-lived, changing emotions? Or, as above, could one be moved in general while being aware that many other emotions are being expressed?

    3. C. Increasing the naturalistic context of the listening situation.

11. 11

    1. A. Collaborative cross-cultural research. Too little emotion research addresses questions of cultural difference and similarity. Emotion researchers need to forge links with sympathetic anthropologists and ethnomusicologists in order to carry out studies in a wide range of non-Western cultures. Granting agencies need to be sympathetic to requests for travel funds and ignore the objections of those anthropologists/ethnomusicologists who are ideologically opposed to cross-cultural experimentation.

    2. (p. 949 ) B. Better non-animal behavioural research methods. Whether one thinks ill or well of animal research, the field of emotion has benefited enormously from the study of rats, mice and cats (e.g. Panksepp, 1998). Invasive neurosurgery and neurochemistry simply cannot be used with human participants. Unfortunately, animals cannot tell us how they are feeling, so researchers must rely on explicit overt behaviours (such as licking, stratching, sniffing, nursing, fleeing, etc.) in order to infer the affective state of the animal. As a consequence, animal research has addressed only a handful of ‘major’ emotions such as fear, panic, anger, hunger, social bonding, seeking, etc. Especially if we want to avoid research with primates, we simply can’t use animals to study emotions like jealousy, envy, pride, loyalty, joy, gratitude, sadness, schadenfreude, optimism, humour, etc. Considerable progress has been made with imaging technology, but we need to be a lot cleverer if we want to address the neurophysiology and neurochemistry involved in the sorts of subtle emotions that are found in music listening and creation.

    3. C. Attending to qualia: Better connection with phenomenology. Feelings are the proximal reason people listen to music. Emotion research will benefit by working more closely with those engaged in detailed description of the phenomenology of musical experience.

12. 12

    1. A. Resolving problems concerning terminology and distinctions.

    2. B. Improving the measurement of emotional responses to music.

    3. C. Focusing more systematically on the underlying mechanisms (i.e. on how, exactly, emotions are evoked by music).

13. 13

    1. A. Psychological and physical care of young musicians (e.g. avoidance of early and/or excessive exposure; suitable repertoire, etc.).

    2. B. Development of evidence-based treatments for music-performance anxiety.

    3. C. Music as an adjunctive tool in the management of emotional disorders (i.e. anxiety and depression).

14. 14

    1. A. Identify the brain signatures underlying different emotions.

    2. B. Shed more light on the temporal dynamics of emotions (also with regard to the activity if the brain structures underlying these emotions).

    3. C. Investigate how we can use the emotion-modulating effects of music for the treatment of disorders, and to improve both psychological and physiological health.

15. 15

    1. A. A concerted return is overdue of both laboratory and field-research attention to the effects of emotion and mood on the choice among music-listening alternatives.

    2. B. What is required is a complex comparative examination of causal paths and mediation that are involved in the effects of music and other art stimuli, as well as a routine multi-component measurement of emotion.

    3. C. A detailed investigation of the states/concepts of being moved and aesthetic awe has been initiated in several laboratories; such redirection of some of the research effort that is currently invested in the alleged effects of music on the fundamental emotions, and on ‘states’ that exist only as linguistic ‘emotional’ labels in verbal reports, would be advantageous to the field.

16. 16

    1. A. Integration of theories concerning music with more general theories of emotion.

    2. B. Integration of theories concerning emotional response to music with those concerning ‘like-dislike’ responses.

    3. (p. 950 ) C. (From a more selfish perspective) Integration of theories of emotional response to music with theories concerning responses to other aspects of the environment.

17. 17

    I think the field is mature enough to conduct:

    1. A. cross-domain (e.g. music and language); and

    2. B. cross-cultural comparisons.

18. 18

    1. A. Assessment: I think more attention needs to be given to developing reliable and valid measures of music-related behaviour (from music preferences to uses of music).

    2. B. Methods: I think it is important for the field to conduct studies that capture people’s everyday involvement and experiences with music (studies that capture naturalistic behaviours that occur in the ‘real world’).

    3. C. Generalizability: One concern I have about some of the music research out there (especially my own) is that it provides a snapshot of the connections between music and psychology among young people born in Western societies that speak English. The findings may not generalize to younger or older generations or to people in other cultures. Collecting data from more diverse samples would be very helpful in this regard.

19. 19

    1. A. The Wittgensteinian tenet holds that language is our greatest problem. Misunderstandings and misinterpretations of words from one field that have a similar meaning in another, or with a different word, are perhaps our biggest challenges. But they always have been, especially as the various silos of research feel compelled to talk to one another without having genuine time to do so thoroughly. I don’t mean that in a derogatory way. We all have limited time, and we do our best. So, it is not really an issue for the next five years, but a continuing challenge. Authors and teachers simply need to keep reminding us of how to handle the problems. An example is this very volume: while you have asked authors to pay careful attention to the definitions you have provided, my reading of at least one other chapter, and indeed my own writing, cannot be said to precisely fit them.

    2. B. Continuous-response research is in need of tools that are easily available and easy to use. My suspicion is that many people interested in continuous-response methods are too timid to use it because it seems too new and complicated. By having some tools that help them collect and/or help them analyse such data, they may start feeling brave enough to explore continuous-response approaches. At the same time, those who have more experience in statistics and mathematics may start entering the field, providing more technically ‘high-end’ contributions.

    3. C. We also need a lot more reliability and validity testing of continuous-response approaches—for example, how do we know for how long a listener remains on task when making continuous emotional responses? We can’t or perhaps shouldn’t simply interrupt them and ask them if they were concentrating. Well, maybe we could. But the challenge will be to come up with non-intrusive ways of addressing this and similar issues. So I think the period of (rather exciting) exploration with continuous emotional response will continue. But sooner or later, the approach will be taken more seriously, and become more refined when these checks are made, repeated and compared.

20. 20

Wow! A big, even intimidating request.

From my standpoint, the number one question is: What is the relation between emotion and appreciation in music? This question includes several subsidiary questions. For example, is emotion essential for music appreciation or can a purely cognitive or intellectual experience (p. 951 ) be equally valid? And if emotion does make a significant contribution, whether separately or in conjunction with cognitive appraisals, then what kinds of emotions make that contribution? For instance, are the emotions generic, such as the ‘arousal potential’ of the old ‘new’ experimental aesthetics, or are they in some way music specific? We seem to have distinctive modules for the cognitive processing of music. Do we have similar modules for affective appreciation of musical compositions? Or is the emotional experience of listening to Beethoven’s Fifth comparable to that of reading a great work of literature or painting? And permeating all these questions is the precise form of the emotion-appreciation relation: is it linear or curvilinear? And does the relation depend on emotional content or type?

1. 21

   1. A. More research on psychological mechanisms.

   2. B. More research into emotional sequencing (in long pieces of music)—what is the emotional result of listening, say, to a symphony, during the course of which you experience many different emotions?

   3. C. Individual differences in emotional responses to music (e.g. are there some people who don’t experience much emotion in music—and why?).

2. 22

   1. A. Revisit experimental aesthetics and Berlyne’s early attempts at building a basis for an understanding of a biological aesthetics. We have a better neuroscience base now but many of his perceptual and philosophical insights were ahead of their time.

   2. B. More rigorous research into the role of musical emotion in neurorehabilitation and psychiatry—contrary to popular belief, a field with a weak research record.

3. 23

   1. A. Investigations that promote a greater sensitivity to cross-cultural similarities and differences in the relation between music and emotion.

   2. B. A focus on music experiences that do not fall clearly within the category of emotional responses.

   3. C. A focus on experiments designed to elucidate psychological mechanisms underlying emotional responses to music.

4. 24

   1. A. Gaining greater insight into the contribution of language, cognition, and culture to the development of musical emotions.

   2. B. Developing more refined techniques for separating the perception of emotion in music from its emotional effects on listeners.

   3. C. Exploration of the emotional consequences of joint music making.

5. 25

   1. A. Research on mechanisms.

   2. B. Research on the relation between music, emotion, and health.

   3. C. A wider assessment palette.

6. 26

   1. A. A better understanding of the emotionalizing of performers. What is happening cognitively when performers use a mood-induction strategy in performance? Do the emotional language and imagery examples so prominent in instruction promote the idea that performers themselves must feel what they hope to convey to audiences?

   2. B. Objective measures of emotion in performing and listening. Perhaps using brain-scan technology to understand if emotion centres are activated in expressive performance conditions. Or using other physiological indicators (heart rate, skin conductance) to study emotional experiences with music, and compare with self-reports.

   3. (p. 952 ) C. How can formal education better address the music-emotion relationship? For example, conventional instruction is largely analytical in nature. Music-listening instruction often involves identifying discrete properties (melody, timbre, form). This goes against the natural desire to listen to be emotionally moved. Do such instructional experiences affect student understanding of the emotional nature of music?

7. 27

   1. A. The first is that self-reports of emotion to music need to become more subtle, systematic and musically ‘intelligent’. I see it as a plea for having researchers trying out new methods, including similarity judgements of emotional music that do not require any verbal labelling and that can be analysed by cluster analysis or multi-dimensional scaling techniques.

   2. B. A second issue is that more attention needs to be paid to mechanisms and moderators of musical emotion induction. Moderators are aspects of performance, of listeners, and contexts (such as room acoustics) in addition to the traditional focus on musical structure as emotion-driving forces.

   3. C. Finally, music therapy should grow out of its impressionistic phase and become a real clinical science, in which any study needs to involve rigorous control groups that make findings interpretable. For example, is there more to the pain-reducing effects of music than simple distraction? Amazingly, we don’t know.

Note: A few responses have been edited for brevity or to preserve the anonymity of the respondent.