Abstract: This essay offers a critical examination from a Peircean perspective of Susanne Langer's account of the role of emotion in musical meaning. Langer maintains in Philosophy in a New Key that the purpose of music is to offer a logical representation of human emotions, not to express emotions directly or to bring about emotive responses in listeners. In arguing thus, Langer, despite her intention to secure a place for emotion in accounting for musical meaning, perpetuates a dualistic view of the relation between reason and emotion that has tended to dominated musical aesthetics for more than a century. But Charles S. Peirce's concept of the emotional interpretant, taken together with his claim that a piece of music is the quintessential example of a sign that is interpreted by feeling, provides a means to refute Langer's position and overcome the dualism by showing how musical meaning can be both derived from and intended for affective experience on the one hand and properly conceptual on the other.
In 1854, the Viennese music critic Eduard Hanslick, champion of Brahms and opponent of Wagner, sparked a controversy about the meaning of music that has shaped the terms of the issue to this day. In his On the Beautiful in Music (Vom Musikalisch-Sch'nen), Hanslick argues that musical meaning consists entirely in "tonally moving forms" (t'nend bewegte Formen). Musical meaning is entirely autonomous; it is constituted by the relationships among tones in the musical work, and it does not signify any content or referent outside or beyond the music itself. Hanslick's purpose in developing this formalist account of musical meaning is to refute the commonly held stance that the aim of music is to represent emotions. Hanslick rejects both the view that the purpose of music is to arouse emotions in the listener and the position that its aim is to express the emotional content of a composer's experience. 
Ever since Hanslick's time, musicologists and aestheticians have struggled with the question of what role emotion plays in the experiences of performing and listening to music, and, more specifically, in the understanding of musical meaning. The terms according to which the question has been debated were identified by Leonard Meyer in his influential 1956 work, Emotion and Meaning in Music, and these terms have changed little in the subsequent fifty years. Meyer identifies two distinct but overlapping debates. The first is between absolutists, who, like Hanslick, "insist that musical meaning lies exclusively within the context of the work itself, in the perception of the relationships set forth within the musical work of art," and referentialists, who maintain that "music also communicates meanings which in some way refer to the extramusical world of concepts, actions, emotional states and character"(Meyer 1956: 1). Meyer correctly notes that this debate reflects a false dichotomy concerning musical meaning: absolute and referential meanings "can and do coexist in one and the same piece of music, just as they do in a poem or painting" (ibid.). The second debate is engaged on the basis of a somewhat more subtle dichotomy, that between formalism and expressionism. Formalists contend that musical meaning "lies in the perception and understanding of the musical relationships set forth in the work of art and that meaning in music is primarily intellectual": expressionists maintain that "these same relationships are in some sense capable of exciting feelings and emotions in the listener" (ibid: 2-3). All musical formalism is, of course, musically absolutist, but a musical expressionist can be either absolutist or referentialist. Absolute expressionists "believe that expressive emotional meanings arise in response to music and that these exist without reference to the extramusical world of concepts, actions, and human emotional states," whereas referential expressionists "would assert that emotional expression is dependent upon an understanding of the referential content of music" (ibid.: 3).
Despite Meyer's recognition that the arguments between absolutists and referentialists "are the result of a tendency toward philosophical monism rather than a product of any logical opposition between types of meaning" (Meyer 1956: 1), I believe that the way in which musical aestheticians and philosophers of music have tended to formulate the issues concerning emotion in musical meaning perpetuates two false dualisms. The first dualism is reflected in the idea that the emotive and intellectual dimensions of experiencing music are separable parts of that experience, or that they even constitute two different experiences of the same phenomenon. The second dualism is reflected in the common assumption that there is an ontological difference between musical and extramusical meaning. In the space allotted, I can only examine the first of these dualisms. I will draw upon Peirce's notion of the emotional interpretant in order to show how musical meaning is intrinsically both emotive and cognitive. However, I will also indicate briefly why a rejection of the first dichotomy points to a need to reexamine the second one as well.
Let us examine one important theory of the role of emotion in musical meaning: that of Susanne Langer in Philosophy in a New Key (1957). I choose Langer because her theory, unlike that of some scholars in the field, makes a subtle attempt to overcome the reason/emotion dichotomy by accounting for how musical meaning can be at once conceptual and emotive in the same experience of listening (as opposed to thinking of the conceptual and emotive dimensions of the experience as somehow constituting two different experiences). For Langer, music is a presentational symbolic form. Presentational forms, the symbolic forms that are expressed through artworks, convey meanings differently than do discursive forms, the means by which analytical reasoning proceeds. Musical meaning has a conceptual dimension, but, like the meaning conveyed through visual art forms, it is not reducible to linguistic description (1957: 94-96). The idea of musical meaning as presentational is promising, for it makes it possible to consider how the act of grasping musical meaning may be emotive as well as intellectual, and even to analyze the possibility that an emotive response to a piece of music might constitute a kind of significant interpretation of it as well. We will see, though, that this promise is inadequately fulfilled in Langer's theory, because of her commitment to the position that emotions cannot be interpretive or significant without being filtered through concepts. Peirce will show us, on the contrary, that they can.
To discuss Langer's concept of musical meaning, it is necessary to review some basic features of her theory of symbolic forms. The title of her book, Philosophy in a New Key, is significant in two ways. First, it indicates Langer's intent to recover the philosophic importance of the arts from the then-dominant positivistic ideology that dominated Anglo-American philosophers, which identified meaning exclusively with discursivity and tended to reject art as irrational and hence meaningless. Second, the metaphor that constitutes her title suggests that music plays a paradigmatic role in her theory of artistic symbolization.
What, according to Langer, is a symbol? She distinguishes between signs, which both animals and humans use, and symbols, the use of which is an exclusively human process. Indeed, Langer maintains that symbolization is "the fundamental process of [the human] mind" (1957: 28), as well as "a primary human need" (ibid.: 40-41). Her definition of the term "sign" is quite circumscribed, and is roughly equivalent to a Peircean index when it is considered primarily in terms of its generating an energetic interpretant (i.e., an action that interprets a sign). Her examples attest to this rough equivalence (ibid.: 29-30):
As soon as sensations function as signs of conditions in the surrounding world, the animal receiving them is moved to exploit or avoid those conditions. The sound of a gong or a whistle, itself entirely unrelated to the process of eating, causes a dog to expect food, if in past experience this sound has always preceded dinner; it is a sign, not a part, of his food. Or, the smell of a cigarette, in itself not necessarily displeasing, tells a wild animal that there is danger, and drives it into hiding.
The difference between a sign and a symbol is that a sign "indicates the existence"past, present, or future"of a thing, event, or condition" (ibid.: 57), whereas a symbol represents an object without implying its existence. Signs announce their objects; symbols lead us to conceive their objects. In at least some cases, the same thing or phenomenon can function either as a sign or a symbol depending upon how it is interpreted. "If you say "James" to a dog whose master bears that name, the dog will interpret the sound as a sign, and look for James. Say it to a person who knows someone called thus, and he will ask: "What about James"?" (ibid.: 62). There is, nonetheless, an essential difference between sign and symbol: the function of a sign is "to evoke action appropriate to the presence of its object" (ibid.: 60), but symbols are "vehicles for the conception of objects" (ibid.: 60-61; italics deleted). In a sign relation, there are three terms: subject, sign, and object, but a symbol is a four-term relation among subject, symbol, conception, and object. According to Langer's theory, a symbol is always interpreted by a conception, so Langer's symbol could be defined in Peircean terms as a sign that is necessarily interpreted by a particular kind of logical interpretant (a habit, a habit-change, or a thought that interprets a sign).
A conception is the meaning of a symbol as the individual interpreter grasps it. Each conception is associated with a concept, the abstract form that is embodied in conceptions. The concept is "that which all adequate conceptions of an object must have in common" (Langer 1957: 71; italics deleted). Langer illustrates the relation between conception and concept by inviting the reader to consider "a photograph, a painting, a pencil sketch, an architect's elevation drawing, and a builder's diagram," all of which show the front view of the same house. We are able to recognize each depiction as representing the same house because each image "expresses the same relation of parts." All correct visual conceptions of the house reveal this same fundamental pattern; hence "we can talk together about the "same" house despite our private differences of sense-experience, feeling, and purely personal associations. . . . The same concept is embodied in a multitude of conceptions" (ibid.: 71).
Langer's rather unusual treatment of the term "connotation" in relation to symbols is central to understanding her account of the nature of musical meaning. To Langer, the denotation of a symbol is its relation to its object. The symbol's connotation is the conception it conveys. "Because the connotation remains with the symbol when the object of its denotation is neither present nor looked for" (Langer 1957: 64), connotation is the means by which we are able to conceive of objects in abstraction from any possible immediate existential response to them, and thus to treat them symbolically rather than merely signally. To say that symbolization is abstractive is to say that is a process of comprehending a significant form.
The significant forms conveyed through symbols may be either discursive or presentational. Discursive forms are those which unfold logically and temporally by means of analytical reasoning. Discursive thought must always be expressed linguistically precisely because it is temporal. On the other hand, "any idea which does not lend itself to this "projection" is ineffable, incommunicable by means of words" (Langer 1957: 81). Such concepts are embodied in presentational forms. In the face of Carnap's claim that artistic and poetic expressions are nonsensical because they cannot be reduced to propositions with a truth value, Langer argues that these presentational symbols do in fact convey meanings. The primary function of presentational symbolism is to render the "flux of sensations" into concepts in such a way as to give us "concrete things in place of kaleidoscopic colors or noises..." (ibid.: 93). Discursive symbolism has a primarily general reference, but there is no intrinsic generality to presentational forms. This is why she calls them presentational; a presentational form represents a singular, unique object.
In her initial discussion of presentational forms Langer draws largely upon examples from the visual arts, especially painting. All the constituents of visual art forms appear to us simultaneously, so it is easy to understand how their meaning is "presentational." The expressiveness of discursive forms is limited by the linear nature of logical argumentation, but the presentational symbols do not have this restriction. They can convey ideas of greater complexity than can be expressed discursively, especially ideas which convey insight into feelings (ibid.: 93, 101). In painting and similar arts all this complexity is offered to the viewer in a single vision at a single moment. But despite the presentational advantage she finds in visual, nontemporal symbols, Langer considers music, an essentially temporal medium, to be the exemplar of presentational meaning.
Music serves to symbolize human emotions, and it is far better equipped than discourse to reveal meanings pertaining to them. Emotions, Langer maintains, can be conceived as "possible ingredients of rationality" insofar as they have they have "definite forms" which can be articulated symbolically (ibid.: 100). The purpose of music is not to stimulate emotions in the listener or to discharge feelings that a composer has as he or she writes, but rather to offer a logical representation of "emotions, moods, mental tensions and resolutions" (ibid.: 222). In Langer's view, the emotional content conveyed through music "reflects the composer's knowledge of human feeling, not his emotional constitution at the time" (Alexander 1992: 260; cf. Langer 1957: 221). Musical meaning, then, is purely connotative; music does not denote objects that we can "point to," "announce," or "look for." Furthermore, the connotations of musical tones, like those of the elements of the visual arts, are not fixed; there is no "dictionary meaning" of a musical tone. Music turns out to be superior to language for the purpose of representing feeling symbolically precisely because of this connotative flexibility. Music, says Langer, "is not usually derived from affects nor intended for them; but we may say, with certain reservations, that it is about them" (1957: 218).
As I have already indicated, Langer's account of musical meaning is an attempt to articulate a position balanced between musical formalism and musical expressionism. She allies herself with Hanslick in her critique of the common notion that the proper aim of music is to bring about emotional catharsis, and that the meaning of music is the feeling, or the complex of feelings, it is intended to evoke. "Most people connect feelings with music, and ... believe they have the feelings while they are under the influence of the music, especially if you ask them which of several feelings the music is giving them (1957: 213). Inherent in the catharsis view is the idea that "music is essentially a form of self-expression" (ibid.: 216); i.e., musical meaning is reducible to the mere discharge of emotion. The history of music, she argues, gives the lie to this na"ve assumption, for it "has been a history of more and more integrated, disciplined, and articulated forms, much like the history of language, which waxes important only as it is weaned from its ancient source in expressive cries, and becomes denotative and connotative rather than emotional" (ibid.). Yet Langer perceives a kind of extremism in Hanslick's response to the catharsis view. "Vehemently he declared that music conveys no meanings whatever, that the content of music is nothing but dynamic sound-patterns ("t"nend bewegte Formen"), and that "the theme of a musical composition is its proper content." But especially the true Wagnerian aim"the semantic use of music, the representation of emotive life"aroused his opposition" (Langer 1957: 225). Hanslick's main error, Langer maintains, was to assume that symbolic representation can only be denotative and discursive. Hanslick thus is led to exclude music from the realm of symbolic meaning"which, given Langer's position, is tantamount to excluding it from the realm of meaning altogether.
Langer is certainly correct, as is Hanslick, to resist the reduction of musical meaning to emotional discharge (a more accurate term, I think, than her "self-expression"). But her means of rescuing musical meaning from positivist criticism on the one hand and sheer emotivism on the other leads her to the first kind of dualism I outlined earlier in this essay, namely, the view that the affective and intellectual aspects of experiencing music are separable parts of the experience. As we have seen, although she characterizes musical meaning as a "language of feeling" (Langer 1957: 221), this meaning consists not in the direct experiencing of emotional content, but in the conceptualization of emotions. In other words, music signifies emotional concepts, but not emotions per se. We draw upon material from our affective lives in our experiences of listening to and performing music, but the way in which we draw upon it is through reflective acts of the intellect; "[m]usic is not the cause or the cure of feelings, but their logical expression" (ibid.: 218). This intellective activity renders possible the "psychical distance" from immediate emotional investment in an artwork that is required for aesthetic appreciation and judgment to take place (see Bullough 1912).
Langer's position might be convincing were it not for the fact that most people do experience music as giving rise to emotional content. Her recognition that the catharsis theory of musical meaning was "widely accepted by musicians and philosophers alike" in her time is significant (ibid.: 218), as is her admission that "anyone who has a voice or an instrument can verify the relief of musical outpourings, from his own experience" (ibid.: 219). These admissions attest to my suspicion that just about anybody other than a philosopher, an academic music theorist, or a composer ideologically committed to formalism would find Langer's idea that musical meaning only indirectly signifies emotional content to be at least partly alien to their own experiences of music. I know it is alien to my experiences, both as a listener and as a performer. Langer fails to recognize that musical meaning is, indeed, "derived from affects" and, at least in part, "intended for them." Musical meaning does involve the conceptualization of emotional content, and conceptualization is necessary for the conscious recognition of musical meanings. But much, if not most, music is at the same time directly emotive, both in its origin and in its signification. If it were not, human beings would not be driven to create it.
Charles S. Peirce's theory of the interpretant"that which connects a sign with its object--provides a means to articulate how musical meaning can be, on the one hand, derived from affective experience and intended to some extent for it, and, on the other hand, properly conceptual. To see how this is possible, let us remind ourselves briefly of the nature of the Peircean sign.
A Sign, or Representamen, is a First which stands in such a genuine triadic relation to a Second, called its Object, as to be capable of determining a Third, called its Interpretant, to assume the same triadic relation to its Object in which it stands itself to the same Object. The triadic relation is genuine, that is [,] its three members are bound together by it in a way that does not consist in any complexus of dyadic relations (c. 1902; CP 2.274).
This definition, which is one of Peirce's most general and abstract, is instructive insofar as it reveals the universal structure of the sign wherever it may occur. For Peirce, a sign is not a "thing," but a nexus of relations. The most fundamental characteristic of the sign relationship is that it is irreducibly triadic. The three terms that constitute the relationship are sign (or representamen), object, and interpretant. Each component stands in a relationship to each of the other components, but it is only the triadic relation, the relation among all three, that properly constitutes the sign relationship. Peirce's concept of sign is metaphysically and logically rather than psychologically based; everything whatsoever has a semiotic dimension, anything can be examined in semiotic terms, and signs do not necessarily require human interpreters in order to signify. Indeed, the great project, left incomplete, of the last third of Peirce's life was to show how the logic of the universe itself is constituted by ever-evolving sign relations.
The sign relation manifests Peirce's ontological/phenomenological system of categories. Unlike Aristotle's categories, which are articulated mainly in terms of natural beings, and Kant's categories, which are intended to articulate the structure of the mind, Peirce's categories are designed to articulate what is common to all phenomena regardless of whether they can lay claim to or are interpreted as having existence independent of the human mind. These are Firstness, the category of pure qualitative possibility; Secondness, the category of resistance, efficient causality, or dyadic action; and Thirdness, the category of generality, intelligibility, or mediation. The way in which signs function constitutes an example of Thirdness, and may even be the paramount example (1903: CP 1.537). In the sign relation, as we see from the definition at CP 2.274, the representamen is a First, the object is a Second, and the interpretant is a Third. In order for the sign relation to obtain, all three elements must be present.
Our present concern is with the musical sign, and in particular with understanding how musical signs are interpreted"in other words, with their interpretants. Another of Peirce's definitions of sign is helpful for this purpose.
A Sign, then, is anything whatsoever"whether an Actual or a May-be or a Would-be"which affects a mind, its Interpreter, and draws that interpreter's attention to some Object (whether Actual or May-be or Would-be) which has already come within the sphere of his experience; and beside this purely selective action of a sign, it has a power of exciting the mind (whether directly by the image or the sound or indirectly) to some kind of feeling, or to effort of some kind or to thought. (1911; MS 670; quoted in Cumming 1999: 457).
This second definition suggests all the ways in which the object of a sign may manifest itself ontologically--as a May-be (a First), as Actual (a Second) or as a Would-be (a Third). We will return to this issue shortly. For now, though, let us focus on how Peirce's definition reveals the nature of the interpretations to which a sign can give rise. He identifies three possible ways in which a sign can be interpreted: by feeling, by effort, or by thought. In a slightly earlier manuscript, he refers to these as the emotional, energetic, and logical interpretants of a sign (c. 1907: MS 318, 00035; CP 5.475). Here he explains that the effort that constitutes an energetic interpretant can be either physical or mental and that the logical interpretant should not be considered to include only human thought, but is a self-controlled habit or a habit-change (MS 318: 00035; CP 5.475). The "logic" of the logical interpretant may be embodied in the nature of the habit itself, regardless of whether the interpreter is a reasoning human being.
As is the case for the sign relation itself, the operative principle in terms of which interpretants are distinguished is Peirce's categorial scheme. In the order of interpretation, the emotional interpretant is a First, the energetic interpretant is a Second, and the logical interpretant is a Third. Each latter kind of interpretant includes the former but not vice versa, as is true of the categories themselves. Every instance of an energetic and a logical interpretant, then, will include the embodiment of a feeling, since "in all cases . . . there must, at least, be a sense of comprehending the meaning of a sign" (MS 318: 00079). Every instance of a logical interpretant will also include some sort of physical or cognitive effort (MS 318: 00079-80 and 00156-159; CP 5.475-76).
It is, of course, the emotional interpretant that is of most immediate interest to us here. Note what Peirce is saying: all acts of signification include, at their root, a feeling that interprets the sign, and this feeling is always present, whether or not the sign is further interpreted by an action, a habit, or a concept. It is, I think, remarkable that Peirce offers the example of the act of listening to a piece of music as the paradigmatic case of a sign interpreted emotionally.
Every sign whatever that functions as such must have an emotional interpretant; for under that head comes the feeling of recognizing the sign as such; and it is plain that a sign not recognized is not a sign at all. The performance of a piece of music may excite musical emotions without being a sign. But if the hearer discerns in the notes the musical ideas or emotions of a composer; then the music conveys to him a message from the composer, and it becomes a sign. In such cases the emotional interpretant is highly developed (MS 318: 00035-36).
Later in the same manuscript (MS 318: 00201), Peirce maintains:
A piece of concerted music is evidently a sign. For it mediates between the series of musical feelings and emotions of the composer and those of the auditor. It conveys; namely it conveys feelings. A feeling is a cross-slice, or lamina, out of the current of consciousness, taken in itself, without any analysis and tearing apart any comparison (since comparisons consist in community of elements, and feeling is not cut up into elements). Only "feeling" is to be understood in the sense of a quality, not in that of an event, which would be existential.
Peirce's account would appear to be at least partly compatible with Langer's insofar as he seems to be saying that the feelings conveyed through music are mediated through conceptual signs. If "the performance of a piece of music may excite musical emotions without being a sign," it functions cathartically in Langer's sense and does not convey meaning. But for a piece of music to convey meaning (i.e., to signify), which Peirce says requires the listener to "discern in the notes the musical ideas or emotions of a composer," it seems that the listener must also form concepts"logical interpretants"that make acquaintance with such musical ideas or musical emotions possible. If this is so, does this mean that Peirce intellectualizes emotion in a way similar to Langer? Does Peirce's brief account of musical meaning necessarily lead to the same reason/emotion dichotomy that Langer's does?
The answer to this question is "no," and it is to be found in Peirce's understanding of the relation between feeling and consciousness. Recall that Firstness is the category of pure qualitative possibility. "The typical ideas of firstness are qualities of feeling, or mere appearances," Peirce maintains (1904: PW 24), and one of his favorite examples to illustrate a First is that of "a feeling of red." (e.g., MS 318: 00203) But feelings never arise in experience in a pure state, although each feeling in itself is sui generis. Because it is impossible to "inhibit the processes of mental elaboration in reproducing that instantaneous state . . . we have to put up with generalized feelings in place of the very feelings themselves; and in these substitutes we only find remnants of the sui generis character" (ibid.: 202). These "generalized feelings" are feelings as they are experienced. The fact that experience is irreducibly temporal guarantees that it is a semiotic process. And every experience involves the experience of feeling because, according to Peirce, consciousness is feeling, as he indicates in the passage cited above (see also1907: CP 5.492)"not feeling as pure First, but that "congeries of feelings" (CP 5.494), "varying greatly in quality and in intensity, which are symptomatic of the interaction of the outer world [i.e., the world of Secondness, or efficient causality] ... and of the inner world [the world of the psyche] ... " (CP 5.493). Consciousness emerges from this temporal interaction, and because it mediates the inner and outer worlds, it is a component of a triadic relation. Insofar as feelings are Firsts, they are sui generis and nonrelative (except for the relation of identity that a First "has" to itself). But insofar as they become components of consciousness, they enter into a triadic relation and hence an interpretive framework. When these feelings are conveyed to an interpreter from a representamen in accordance with its dynamical object, they function as emotional interpretants.
According to Peirce, a piece of music "mediates between the series of musical feelings and emotions of the composer and those of the auditor." Langer, as we have seen, would modify this claim to state that music mediates between the musical concepts of feelings conveyed by the composer and the conception of those feelings the listener forms when he or she hears the music. The introduction of the emotional interpretant allows Peirce to account for why the experience of being moved emotionally by, say, Bach's Cello Suites or Vaughan Williams' The Lark Ascending is a properly musical response--and, I wish to suggest, indispensable for the further experience of conceptualizing the feelings expressed in the performance of the piece. Furthermore, once we understand how a feeling can be interpretive"how it can mediate between the musical sign and its dynamical object--we can see that not all emotive responses to music should be dismissed as emotional discharges (Langer's "self-expression"). If a loud trombone blast makes me jump in my seat and experience panic, that is emotional discharge. But if I am moved to tears by a performance of Beethoven's Ninth Symphony, that is an instance of interpretation by means of emotional interpretants.
This is not to say that emotional interpretants are adequate by themselves for giving rise to or for mediating musical meanings. When I am moved by The Lark Ascending, my emotional response occurs in connection with a complex array of energetic and logical interpretants as well. When I pay attention to the music, retain the memory of what I have just heard, and anticipate what I am about to hear next, I engage in efforts that constitute energetic interpretants. The habits of listening that I have formed on the basis of my lengthy acquaintance with Western art music and study of its history, my education in music theory, and my experience as a (former) violinist constitute logical interpretants. Many, but not all, of these habits are predominantly conceptual (for, as we have seen, an intellectual concept is only one kind of logical interpretant). My conceptual knowledge of violin technique enhances my appreciation of the effortless character of Christopher Warren-Green's recorded performance of the Lark. My habit of attentive (and emotive) response to certain harmonic progressions in the piece is also conceptual to the extent that I recognize those progressions when I hear them, but the formation of this habit (and therefore its "logic") depends in part upon conditions over which I do not exercise conscious control. The formation of concepts is indispensable to experiencing musical meaning, but musical concepts depend upon actually existing interpretive feelings. These feelings, moreover, are not abstracted from in the experience of listening. Even the most rigorously analytical listening experiences, such as a technically educated listener might have in hearing Milton Babbitt's Piano Concerto, involve at least "a sense of comprehending the meaning" of the signs that make up the piece, and therefore cannot entirely escape the realm of feeling. Most composers do not demand such severe intellectual puritanism of their listeners, and the interpretive feelings generated in most listening experience are considerably richer than the ones Babbitt calls for.
Peirce's account of the irreducibly triadic character of meaning and of the interdependence of emotional, energetic, and logical interpretants holds considerable promise for investigating this richness. The interpretants, or "proper significate effects," of music include habits, concepts, actions and emotions, and these are intertwined in an extremely complex manner in almost all experiences of listening to and performing music, as Cummings has demonstrated (1999, 2000).
A fully developed Peircean musical aesthetics might also be a means of countering the second dualism mentioned at the beginning of this essay, the dichotomy between musical and extramusical meaning. Peirce's semiotic theory is compatible with musical referentialism insofar as Peirce agrees that meaningful musical emotions emerge from the context of the so-called "extramusical" world of concepts, actions, and feelings. However, a Peircean aesthetics of music would have to reject the idea that this world stands "outside" the world of musicmaking and listening in an absolute or substantive way. The functional relativity of the elements of the sign relation and of Peirce's categories themselves"the fact that, for example, what serves as an interpretant (a Third) in one sign relation can become a representamen (a First) in a subsequent act of semiosis"suggests that in the performance and hearing of a musical work, the boundary between "musical" and "extramusical" meanings is a relative and shifting one. The interdependence of the three kinds of interpretant, too, suggests that even those musical meanings conveyed in "absolute"works (those without a programmatic narrative: Brahms' Symphony No. 4 in E minor as opposed to, say, Richard Strauss' Don Juan) may not be formally self-contained, and programmatic meanings may in certain respects or in certain cases be more "musical" than commonly thought. These are tantalizing possibilities, for they might well serve to guide us toward a musical aesthetics that would remind us continually of how and why music is a unique and wondrous manifestation of the human drive to create and express meaning.
Felicia E. Kruse
Alexander, Thomas M. 1992. Langer, Susanne. In A companion to aesthetics. Ed. David E.Cooper. Oxford: Blackwell.
Bullough, Edward. 1912. Psychical distance as a factor in art and as an aesthetic principle. British Journal of Psychology (5) 2, 87-118.
Cumming, Naomi. 1999. Musical signs and subjectivity: Peircean reflections. Transactions of the Charles S. Peirce Society (35) 3, 437-474.
______________. 2000. The sonic self: musical subjectivity and signification. Bloomington and Indianapolis: Indiana University Press.
Dewey, John. 1934. Art as experience. The Later Works, 1925-1953. Vol. 10. Ed. Jo Ann Boydston. Carbondale and Edwardsville: Southern Illinois University Press, 1987. All references to this work are according to the standard form for this scholarly edition (LW 10 followed by page number).
Frogley, Alain. 2001. Vaughan Williams's ninth symphony. Oxford: Oxford University Press.
Goehr, Lydia. 1992. The imaginary museum of musical works: an essay in the philosophy of music. Oxford: Clarendon Press.
Hanslick, Eduard. 1891. On the musically beautiful [Vom musikalish-sch'nen]. 8th German edition trans.Geoffrey Payzant. Indianapolis: Hackett, 1986.
Hausman, Carl R. 1993. Charles S. Peirce's evolutionary philosophy. Cambridge: Cambridge University Press.
Langer, Susanne K. 1957. Philosophy in a new key: a study in the symbolism of reason, rite, and art. 3rd ed. Cambridge and London: Harvard University Press.
Meyer, Leonard. 1956. Emotion and meaning in music. Chicago and London: The University of Chicago Press.
Peirce, Charles S. All references to Peirce's works are incorporated into the text of this essay and follow the standard scholarly conventions for citing them.
CP references are by volume and paragraph to The collected papers of Charles Sanders Peirce. Vols. 1-6 ed. Charles Hartshorne and Paul Weiss. Cambridge: Harvard University Press, 1931-35. Vols. 7-8 ed. Arthur W. Burks. Cambridge: Harvard University Press, 1958.
MS references are to Peirce's unpublished manuscripts as listed in Robin 1967.
W references are to Writings of Charles S. Peirce: a chronological edition. Ed. by the Peirce Edition Project. Bloomington and Indianapolis: Indiana University Press: 1982- .
PW references are to Semiotic and significs: the correspondence between Charles S. Peirce and Victoria Lady Welby. Ed. Charles S. Hardwick. Bloomington: Indiana University Press, 1977.
Robin, Richard S. 1967. Annotated catalogue of the papers of Charles S. Peirce. Amherst: The University of Massachusetts Press, 1967.
Schelling, Friedrich Wilhelm Josef von. 1803. Philosophie der Kunst. Lectures delivered at Jena, 1802-03. Schellings Werke, Erg"nzungsband vol. 3. Ed. Manfred Schr"tzer. Munich: Beck, 1959.
 The musical paradigm Hanslick has in mind (as does Langer, as we will see shortly) is Western art music, or "classical music." When Hanslick wrote On the Beautiful in Music, the Romantic idea that there is a species of music that can be labeled a "fine art," essentially separate from and elevated above the artifacts of ordinary experience, was still fairly new. Lydia Goehr (1992) argues that Hanslick's formalism is a logical consequence of the Romantic impetus to view music as naturally autonomous and to desire to "emancipate" it from the extramusical social and religious ends that it had previously served. According to the Romantics, the significance of music "lies not in its service to particularized goals of a moral or religious sort, or in its ability to inspire particular feelings or to imitate worldly phenomena. It lies, rather, in its ability to probe and reveal the higher world of universal, eternal truth" (Goehr 1992: 153). Instrumental music "is the most plausible candidate for being the "universal language of art"" because it "provides a direct path to the experience of a kind of truth that transcends particular natural contingencies and transitory human feelings" (ibid.: 154). It has no intermediary content (unlike, say, visual representational art or poetry, both of which use media that have a particular cognitive content), so it can "actually embody and become a higher truth" (ibid.). This "transcendent move," Goehr maintains, allowed music to become liberated from particularized extramusical ends and from the cognitive specificity of language. It was not sufficient to render music entirely autonomous, however, since music was still not free of "its obligation to be meaningful in extra-musical, spiritual, and metaphysical ways" (ibid.: 155). Formalist aestheticians such as Hanslick attempted to solve this problem by arguing that music "is intelligible not because it refers to something outside itself, but because it has an internal, structural coherence. It consists in an internal and dynamic stream of purely musical elements, in Hegel's terms, in an "abstract interiority of pure sound"" (ibid.). Goehr points to Schelling as an example of a Romantic aesthetician who reconciles formalism with transcendence through his argument that music is the Platonic form of the motions of physical bodies""it represents pure motion as such, abstracted from any other object and borne on invisible, almost spiritual wings" (Schelling 1803: 152; quoted in Goehr 1992: 156).
 The title hints as well that there may be something about the nature of music that is also crucial for the activity of philosophizing.
 In the preface to the second edition of Philosophy in a New Key (1951), Langer states that, had she been able to write the book over again, she would have replaced the term "sign" with "signal," following Morris 1938. Furthermore, it should be noted that Langer does not overtly acknowledge a role for anything like a Peircean interpretant in sign relations, but only in relations involving symbols. Following Morris" misreading of Peirce, she equates the interpretant with the "subject" (i.e., sign user) in a sign relation (Langer 1958: 58). For Peirce's definition of interpretant and his classification of interpretants, see below, pp. 12 ff.
 Hence her use of the term "symbol," like her use of the term "sign," differs considerably from Peirce's. For Peirce, a symbol is a particular species of sign relation. See note 8 below.
 Unfortunately, Langer presents this distinction as a (false) dichotomy: a symbolic form is either discursive or presentational but cannot be both. Her position is undoubtedly a consequence of the fact that she developed her theory of artistic meaning polemically in response to the positivist claims that language is the only means of articulating thought and that feeling does not play any role in intellectual activity. This leads Langer seriously to undervalue the presentational dimensions of linguistic meaning, and it also leads her to fail to reconcile the fact that music is the paradigmatic presentational form with the fact that music is as inescapably temporal as discursive language.
 Langer uses some embarrassingly distorted"not to mention racist"examples to illustrate what she means by "self-expression." "We have more need of, and respect for, so-called "pure music" than ancient cultures seem to have had; yet our counterpoints and harmonic involutions have nothing like the expressive abandon of the Indian "Ki-yi" and "How-how," the wailing primitive dirge, the wild syncopated shouts of African tribesmen. Sheer self-expression requires no artistic form... The laws of emotional catharsis are natural laws, not artistic."
I find Dewey's distinction between "impulsion" and "expression" in Chs. 4 and 5 of Art as Experience (1934) to be more helpful and less misleading than Langer's term "self-expression." What Langer characterizes as "self-expression" is equivalent to the mere discharge of emotion that Dewey says is not expressive, but is a case of merely giving way to an impulsion (LW 10: 67). To express an emotion requires that we be aware of its meaning and consciously weave that meaningfulness into the act of communicating the emotion (ibid.: 68), and "a person overwhelmed by an emotion""rage, for instance""is thereby incapacitated for expressing it" (ibid.: 75; cf. 67).
 It is a frequent source of confusion that Peirce uses both the term "sign" and the term "representamen" indiscriminately to refer both to the triadic relation as a whole and to the element within it that serves as vehicle of signification. In this essay, I will use "sign" to refer to the former and "representamen" to refer to the latter.
 Unlike Langer, Peirce does not draw a dichotomy between sign and symbol. Peirce's definition of sign is far broader than Langer's, and a symbol for Peirce exhibits a particular kind of sign relation. One way in which signs can be classified, according to Peirce, is in terms of the relation between the representamen and its dynamical object, which is the object as conditions the sign by providing resistance in the process of semiosis, thereby constituting the telos of interpretation. According to this principle, signs can be distinguished as icons, indexes, or symbols. An icon is a sign that signifies its dynamical object by resemblance, or "by virtue of its own internal nature" (1904: PW 33), as does a diagram or a passport photograph. An index signifies its dynamical object by virtue of a relation of existence to it (ibid.); a painful red swelling in a wound signifies an infection insofar as it is caused by the infective agent. A symbol signifies its object by means of a convention, a habit, or a natural disposition of whatever effects the interpretation of that sign (i.e., its interpretant) (ibid.). Thus a skunk instinctively stomps its front paws to warn a potential attacker that it is about to spray; in musical notation for bowed stringed instruments, a dot above or below a note conventionally signifies that the note is to be played with a shortened bow stroke.
 For example, suppose a greyhound sees a squirrel five hundred meters away across an open field and takes off after it. The emotional interpretant in this case is the greyhound's attraction to the squirrel as prey and its feeling of satisfaction as it chases and catches it. The energetic interpretant is its act of chasing. The logical interpretant is the law or rule that determines the innate tendency of greyhounds to track moving prey by sight; individual greyhounds express this logical interpretant by means of their habitual behavior. The habit is "self-controlled," not by any autonomous conscious effort on the part of the animal, but insofar as it is shaped through the natural history of dogs and through selective breeding.
 Among Peirce's most evocative accounts of Firstness is W5: 238-39 (1885): "The category of First ought to be the easiest to seize: it is the mode of thought of a child. Yet I believe that to an educated man, it is the most difficult of the three. The word First fails to express it, for this word suggests the question of a Second. Imagine something definite, which is quite new and fresh, yet not aggressively so; something unlike and unrelated to a second, not merely to your thought, but in itself; something without genesis, flourishing in spontaneous and pristine freedom. Think what should have been the first idea in the consciousness of Adam at the moment of his creation, when he found a lively feeling which had not become defined. Remember, however, that every description of the First must be false... The greatest fault of the description ... is that it conveys the suggestion that the conception is to be attained by straining; whereas the whole difficulty of seeing it is that it lies too immediately in the foreground, beyond the margin of any representation."
 What exactly the dynamical object of a musical piece might be has yet to be fully articulated. But given that a dynamical object is both teleological and evolutionary, which is why Peirce calls it "dynamical" or "dynamic" (see Hausman 1993: 72-78 and passim), a musical dynamical object could only be constitued through performance.
 For a superlative account by a violinist-musicologist of how all three Peircean modes of interpretation enter into the experiences of performing and listening to music, see Cumming 1999. Chs. 3 and 4 of Cumming 2000 are also directly pertinent to the issues raised in the present essay.
 See Frogley 2001 for a comprehensive musical analysis of Vaughan Williams' Symphony No. 9 in E minor, the composition of which was strongly influenced by narrative elements of Thomas Hardy's Tess of the D'Urbervilles. Vaughan Williams himself admitted only that in the early stages of its composition the second movement had originally had a program (Tess' night at Stonehenge before her arrest), but "it got lost on the journey."