Proceedings paper


Musical Expression and Musical Meaning in Context
Justin London




1. Some preliminaries.

There is a growing body of work in the philosophy of music and musical aesthetics that has considered the various ways that music can be meaningful: music as representational (that is, musical depictions of persons, places, processes, or events); musical as quasi-linguistic reference (as when a musical figure underscores the presence of a character in a film or opera), and most especially, music as emotionally expressive. Here I will focus on the last topic, for I believe it will be useful for researchers in music perception and cognition to avail themselves of the distinctions that aestheticians have worked out regarding the musical expression of emotion.

Now we often say that music is "expressive," or that a performer plays with great expression, but what exactly do we mean? There are at least things one may be saying. First, one may be praising a performer for their musical sensitivity, that he or she has a keen sense of just how a passage is supposed to be played. Such praise is often couched in terms of the performer's "musicality" (in statements that border on the oxymoronic, as when one says that a performer plays the music very musically). Such praise may also be couched in terms of expression--i.e., that a performer plays "expressively." I have little to say about these attributions, save that they are often linked to the second thing one often means when speaking of the music or a performance being expressive: an expressive piece or performance is one that recognizably embodies a particular emotion, and indeed may cause a sympathetic emotional response in the listener. Thus if one plays "expressively," this means that the music's particular emotional qualities--its sadness, gaiety, exuberance, and so forth, are amply conveyed by the performer.

Before going further, a number of other preliminary remarks are in order. When we speak of the expressive properties of music, these are distinct from the expressive properties of sound. Sounds may be loud, shrill, acoustically rough or smooth, and so forth. These acoustic qualities have expressive correlates and may trigger emotional responses, and of course one cannot have music without sound. But musical expression is more than this: it requires the attention to the music qua music, rather than as mere sounds. The opening "O Fortuna" of Carmina Burana may shock (and indeed scare) the listener due to its sudden loudness (especially when the bass drum starts whacking away), but this shock isn't a musical effect--we get the same reaction when we here a sudden "bang" at a fireworks display or when a car backfires. By contrast, in hearing the opening of Mozart's 40th symphony as having a quality of restless melancholy, one attends to both the musical syntax and its sonic embodiment.

Another caveat: as Hanslick has noted, at times a musical work may arouse feelings in the listener through ad-hoc associations. In other words, one must be on guard for the "they're playing our song" phenomenon. These associative properties may be quite strong, and can operate in marked contrast to the innate expressive qualities of a given piece, as in the paradigmatic case of a happy piece that arouses sadness because it reminds the listener of a lost love or deceased friend. As will be noted in some detail below, context plays a pivotal role, and here context can include not only genre, but extra-musical information such as lyrics, the image track of a film score, and literary programs. I take it, however, that a primary interest for researchers in the perception and cognition of musical expression will be in the intrinsic expressive properties of the music itself.

Finally, in philosophical discussions of meaning and expression, there is usually what might be called "the inter-subjective agreement requirement." Here is an example from visual art. If I show you a picture of a man on a horse, and you and everyone else says "that's a man on a horse," this confirms that the picture is a successful representation of a man and a horse. Moreover, I don't have to give you any cues or hints regarding its representational subject. By the same token, in order for a piece of music to be "an expression of emotion X" (or "expressive of X") there must be broad consensus among listeners that the music expresses X, a consensus arrived at without any extra-musical prompting. One problem for accounts of musical expression is that such inter-subjective agreement often does not happen: one listener says a given piece is an expression of anger, while another says it expresses hate, another jealousy, and yet another of sinister passion. What emotion does this piece express? While anger, hate, jealousy, and sinister passion are related emotions, the piece nonetheless fails to individuate any one of them in particular. Musical expression is plastic enough so that the same passage might be expressive of a wide variety of emotional states.

2. Simple Emotions, Higher Emotions, and Moods

In the late 19th century Eduard Hanslick famously denied that music had any ability to express emotions, and many 20th century aestheticians (and composers, most notably Stravinsky) held this to be true. Why would one take up such a counter-intuitive view? Well, philosophers often take up counter-intuitive views, and if you are a philosopher, there are two problems to be surmounted if one wants to claim that a piece of music expresses a particular emotion. The first is the "who" problem: whose emotion is being expressed? Emotions are felt by living, sentient creatures, and as Malcolm Budd has noted, "It cannot be literally true that [a piece of] music embodies emotion, for it is not a living body" (Budd, Music and the Emotions, p. 37). One is thus tempted to claim that a piece of music is an expression of its composer's emotion. But when one examines the compositional history of most works this claim also falls apart, for composers often write sad music, for example, even when they feel no particular sadness (as in the case of Funeral March from of Beethoven's "Eroica" symphony). Nor are they in the throes of sadness during the entire course of composing a piece of music, since the compositional process may last weeks, months, or even years (see, for example, the Adagio Mesto movement of Brahms' trio for horn, violin, and piano, which is purported to be an elegy to his mother). Thus if pieces of music are expressions of emotion, they are disembodied, and usually disconnected from any particular "emotional cause" in the life of their composer.

The second problem is the "why" problem: emotions typically require what are referred to in philosophical parlance as intentional objects, that is, particular people or events that play a causal role in trigger an emotional state. Thus we are jealous of a particular person, frustrated at a particular state of affairs, feel grief at the death of a particular friend or relative, and so forth. One does not, for example, feel "jealous" in general (though one may have a disposition toward jealousy).

Not all emotions are like jealousy and frustration, as some do not always require intentional objects. While one can be sad due to particular event, one also can be generally sad, for example, and such sadness is not dependent upon any particular person, state of affairs, and so forth. As Colin Radford has pointed out, "not all emotions, or occasions of emotion are rational, i.e., they are not informed by, explained and justified by appropriate beliefs [that is, intentional objects]" ("Muddy Waters," p. 249). Radford also explicitly acknowledges that "we naturally call such feelings 'moods.'" ("Muddy Waters," p. 250). Thus there is a distinction between higher emotions (which require an intentional object) and simple emotions and moods which may/do not.

There is now general consensus that music can express moods and simple emotions, contra Hanslick. Some aestheticians, most notably Jerrold Levinson, have claimed that in some music contexts music can do more, in that it is capable of mimicking the characteristic "look and feel" of at least some of the higher emotions (see Music, Art, and Metaphysics, chapter 14).

3. How music expresses emotions I: Cognitivism

But just how does music express simple emotions? There are two main points of view on this question. The first, developed (and much defended) by Peter Kivy, is known in philosophical circles as "cognitivism" or "cognitivist" theories of musical expression. The second, one with a long historical pedigree, can be termed "emotivist" or "arousal" theories of musical expression. Taking up the cognitivist charge, Kivy has repeatedly denied that music really arouses what he has termed the "garden varieties" or real-world instances of sadness, happiness, anger, and other simple emotions in the listener (though music may move the listener through its sheer beauty). For even simple emotions, when fully aroused, usually relate to an intentional object. Thus if we say that a piece of music makes us sad or angry, what exactly are we sad or angry about--the music? ("that damn Symphonie Pathetique!") or its composer? ("that damn Tchaikovsky!"). And has already been noted, a piece that seems expressive of happiness may actually trigger sadness due to extra-musical associations.

For the cognitivist, the expressive properties of music are properties intrinsic to the music, and not, to quote Kivy, "dispositions to arouse emotions in [the] listener" ("Feeling the Musical Emotions," p. 1). Kivy takes this position from O. K. Bouwsma, but he also acknowledges psychological antecedents for this view, in particular Charles Hartshorne's The Philosophy and Psychology of Sensation (1934), and Kivy cites Hartshorne observation that 'Thus the "gaiety" of yellow (the peculiar highly specific gaiety) is in the yellowness of the yellow' (see ibid., note 2, p. 1). In making this move, one allows that music that is expressive of sadness need not make the listener sad.

How exactly does music then express emotions if not by arousing them in the listener? Here Kivy, Levinson, and many others would agree with this explanation given by Malcom Budd (who takes this view in large part from the music psychologist Caroll Pratt): "music can be agitated, restless, triumphant, or calm since it can possess the character of the bodily movements which are involved in the moods and emotions that are given these names" (Music and the Emotions, p. 47). Likewise Kivy develops a "physiognomy of musical expression" and thus claims that music is expressive of these basic emotions by its resemblance to human utterance and behavior. Music thus distills certain aspects of human expressive behavior, especially that of the voice, and renders those aspects into dynamic musical shapes.. Levinson's claim that music can express some higher emotions (such as hope) is based on the claim that some higher emotions have characteristic physiognomies that can be musically portrayed (see ibid.).

Note, however, on this view that in order for the contours of a musical phrase to express an emotion, one must recognize that this "musical utterance and behavior" is akin to other, non-musical utterances and behaviors. Thus musical expression is mediated through our understanding of social behavior in general, and what might be termed a knowledge of "social musical behavior" in particular. It is for this reason that one may mistake musical expressions in an alien musical culture, not because we do not know the musical language, but perhaps primarily because we do not know the normative social behaviors onto which the musical gestures may be mapped.

To sum up so far: the "cognitivist" theory of emotional expression in music says that a piece of music expresses a particular emotion if a suitably grounded listener is able to recognize that emotion in the musical structure by analogy to human social behavior, but she need not assume that this emotion was felt by the composer (or is felt by the performer), nor does the listener have to experience that emotion while listening.

4. How music expresses emotions II: Emotivism

For many other aestheticians, cognitivism is a necessary but insufficient account of emotional expression in music. As Jenefer Robinson has noted, not only does music frequently express emotional qualities, it also frequently affects us emotionally by evoking or arousing emotions in the listener ("The Expression and Arousal of Emotion in Music," p. 13). But what kinds of feelings does music arouse? Are they the same as our "ordinary" emotions, or are they special "musical versions" of emotions? And what is their relationship to our understanding of musical expression?

A common tack taken by a number of philosophers has been to claim that music arouses our emotions, but in a special way. For Kendall Walton, who approaches all kinds of aesthetic experience, and not just music, as a special kind of imaginative activity, expressive music "evokes the imaginative experience of the emotion expressed: more precisely, music expressive of sadness, say, induces the listener to imagine herself experiencing sad feelings" (this cogent summary of Walton is from Jenefer Robinson, op. cit., p. 18). In other words, for Walton our emotions aren't really aroused, but we imagine they are. For Stephen Davies and Jerrold Levinson, expressive music really does arouse the listener's emotions, but emotions of a greatly attenuated kind--"sadness lite", for example. As Kivy has noted with respect to their theories, such emotional arousals "must be weakened, . . . because they do not have the power to make use behave the way those emotions would do in ordinary circumstances" ("Feeling the Musical Emotions," p. 11). For Kivy, champion of cognitivism, this is inadequate. We do not have imaginary or stunted emotional responses when we listen to expressive music, but real, full blown feelings--albeit feelings of a special kind. For Kivy, what moves us is sheer musical beauty, and this beauty may be emotionally individuated: "Sad music emotionally moves me, qua sad music, by its musically beautiful sadness, happy music moves me, qua happy music, by its musically beautiful happiness, [and so on]" ("Feeling the Musical Emotions," p. 13). For all of these philosophers, however, what we do not experience when we listener are ordinary feelings of sadness, happiness, serenity, or so on. Musical emotions are always of a different order.

Jenefer Robinson takes a different approach, one that tries to avoid making musical expression a special case. She considers most carefully what we really do feel when we hear expressive music, and then what we make of those feelings: "As I listen to a piece which expresses serenity tinged with doubt, [for example], I myself do not have to feel serenity tinged with doubt, but the feelings I do experience, such as relaxation or reassurance, interspersed with uneasiness, alert me to the nature of the overall emotional expressiveness in the piece of music as a whole" ("The Expression and Arousal of Emotion in Music," p. 20). Robinson takes care to note that "the emotions aroused in me are not the emotions expressed by the music" (p. 20), and so for her it is not simply that sad music arouses sadness. Rather, our basic feelings--or perhaps "reactions" is a better term--of tension, relaxation, surprise, and so forth, are combined with our awareness of the musical gesture and syntax, and through this combination we gain a sense of what emotion(s) a piece may express.

5. Conclusion and implications for research

As is now clear, while philosophers of music now generally agree that music can express at least some emotions, there is much disagreement as to which particular emotions can be expressed, whether or not such expression depends upon arousing an emotional response in the listener, and if so, what kind of feelings exactly music does arouse. Nonetheless, philosophical discussions of musical expression have a number of implications for research in music cognition and perception.

1. Musical expression always involves sonic properties, and to things like loudness and roughness I would add the rhythmic properties of sounds (as indicative of coordinate movement, spatial location, and so forth). Moreover, alterations to the "sonic" properties of a musical passage may be made without changing its basic melodic or harmonic structure--the same melody and accompaniment played high, fast, and loud may convey a vastly different expressive character from its low, slow, soft version (the locus classicus of such variations is the various presentations of the idée fixe in Berlioz's Symhonie Fantastique).

2. If one uses "real world" musical stimuli, especially well-known repertoire, one will often be faced with "associative interference," as one cannot control the contexts in which subjects have first heard and come to know such repertoire. Therefore in many cases newly composed or otherwise unfamiliar musical stimuli may be preferable, as they circumvent such interference.

3. While music alone may only express a garden-variety emotion, such as anger, that same music in a richer semantic context may be properly heard as an expression of jealousy or hate. Different visual and/or linguistic cues will give different expressive results. Moreover, a level of musical activity that is most apt for one particular emotion may be inapt for another. For example, a passage that expresses "anxious anticipation" very well will not be made more expressive by making it louder, faster, and so forth. There isn't a simple linear relationship between musical parameters and the robustness of an emotional expression.

4. Some perfectly good musical expressions of emotion may not arouse those emotions (or much of anything, for that matter) in the listener. Yet it would be incorrect to call such passages "inexpressive."

5. Any emotions that are aroused by listening to music, while perhaps similar to "real" emotions that occur in non-musical contexts, nonetheless have important differences. Even if context provides an intentional object for an emotion, transforming a yearning, longing passage into an expression of hope (to take an example from Levinson), it is not at all clear that the listener should feel hopeful, what she should be hopeful about, and so forth. Moreover, such hope (and its emotional stimulation) is commingled with other aesthetic properties--balance, beauty, intensity, coherence--and those properties may (and most certainly will) also stimulate affective responses of their own.

Works Cited

Budd, M. (1985, 1992). Music and the Emotions: The Philosophical Theories. New York, Routledge.

Davies, S. (1994). "Kivy on Auditors' Emotions." The Journal of Aesthetics and Art Criticism 52(2): 235-36.

Goldman, A. (1995). "Emotions in Music (A Postscript)." The Journal of Aesthetics and Art Criticism 53(1): 59-69.

Graham, G. (1995). "The Value of Music." The Journal of Aesthetics and Art Criticism 53(2): 139-53.

Kivy, P. (1989). Sound Sentiment. Philadelphia, Temple University Press.

Kivy, P. (1993). "Auditor's Emotions: Contention, Concession and Compromise." The Journal of Aesthetics and Art Criticism 51(1): 1-12.

Kivy, P. (1994). "Armistice, But No Surrender: Davies on Kivy." The Journal of Aesthetics and Art Criticism 52(2): 236-37.

Kivy, P. (1999). "Feeling the Musical Emotions." British Journal of Aesthetics 39(1): 1-13.

Levinson, J. (1990). Music, Art, and Metaphysics. Ithaca, Cornell University Press.

Martin, R. L. (1995). "Musical "Topics" and Expression in Music." The Journal of Aesthetics and Art Criticism 53(4): 417-24.

Meyer, L. B. (1956). Emotion and Meaning in Music. Chicago, University of Chicago Press.

Radford, C. (1989). "Emotions and Music: A Reply to the Cognitivists." The Journal of Aesthetics and Art Criticism 47(1): 69-76.

Radford, C. (1991). "Muddy Waters." The Journal of Aesthetics and Art Criticism 49(3): 247-52.

Robinson, J. (1994). "The Expression and Arousal of Emotion in Music." The Journal of Aesthetics and Art Criticism 52(1): 13-22.



 Back to index