Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives


SRB Archives

This review appeared in Volume 3 (2) of The Semiotic Review of Books.

Is Music Too Definite For Words?

by Robert S. Hatten

Playing with Signs: A Semiotic Interpretation of Classic Music. By Kofi Agawu. Princeton: Princeton University Press. 1991 Pp. 154 ISBN 0-691-09138-2

Gesture, Sign and Song: An Interdisciplinary Approach to Schumann's Liederkreis Opus 39. By David L. Mosley. New York: Peter Lang (1990) Pp. 225 ISBN 0-8204-1102-7

Felix Mendelssohn wrote in 1842. "Words seem to me so ambiguous, so vague...in comparison to genuine music... The thought which are expressed to me by music that I love are not too indefinite to be put into words, but on the contrary, too definite." (Mendelssohn, 313.) Although music has long been analyzed in structural terms, attempts to interpret expressive meaning in music have been hampered by the supposedly indefinite nature of that meaning. In the early Romantic criticism of E. T. A. Hoffmann and A. B. Marx, structural and expressive features often shared equal status in a musical analysis. But the appearance of Edward Hanslick's critical essay, Vom Musicalisch-Schonen (1854), and the increasingly positivistic tendencies of the new science of musicology that came into being in the late nineteenth century, put a damper on efforts to capture elusive expressive meanings. In the twentieth century, changing compositional aesthetics, influenced in part by a reaction against the hyper-emotionalism of late-nineteenth century music, led on the one hand to a more objective/ironic neoclassicism (Stravinsky, with his claim that music was incapable of expressing anything), and on the other to a more abstract structuralist aesthetic leading from serialism to electronic music. Such aesthetics, increasingly focused on the individual piece and its structural coherence, also affected analytical systems being developed for the study of earlier classical music-obscuring the expressive motivations for formal structure.

The need to reclaim lost territory has become increasingly apparent in musicology, where historical research is beginning to provide clear evidence for more carefully considered musical symbolism or expressive meanings. But the theoretical problem of musical meaning, with all its attendant philosophical issues, remains a thorny one. Aestheticians such as Peter Kivy have begun documenting clear cases of expression (1980) and representation (1984) in more programmatic music. But as Kivy's recent book, Music Alone (1990), makes clear, we have a long way to go to understand the expressive capacities of purely instrumental music in genres such as sonata, string quartet, or symphony, for which the composer has provided no verbal clues or indications of expressive purport:

Music alone is a quasi-syntactic structure of musical properties, some of which are describable in phenomenological terms. That among the phenomenological properties are expressive ones makes this quasi-syntactic structure more interesting to human beings but not semantic in any but an adulatory sense of meaningful that is unhappily misleading. (196)

Although Kivy sanctions the analysis of expressive- meanings where they are "just too prominent to disregard" (192), he argues that expressive meanings must not lead to an interpretation that makes extramusical claims of "aboutness" that would suggest an emotive plot.

Does semiotic theory have anything to offer the music theorist hampered by an inability to reconstruct an expressive "language" for music in its own terms? These two recent books, both purporting to be semiotic, tackle this problem. Kofi Agawu relies on a Sausurean semiotic for his analysis of textural and generic types, or "topics", as they interact with the deeper structural forces of harmony and counterpoint. David Mosley proposes a Peircean approach, enhanced by a gestural theory influenced by George Herbert Mead's (1934) focus on gesture in language.

Agawu's book is directed toward the community of music scholars (see endnote), and accordingly begins with a summary of Charles Rosen's (1972) and Leonard Ratner's (1980) significant studies of the Classical style. This is followed by a succinct introduction to semiotics (drawn primarily from Benveniste and Saussure), and an equally succinct summary of work in semiotics of music (Jean-Jacques Nattiez, Wilson Coker, and David Lidov). Though emphasizing the symbiotic relationship of theory and interpretation in his introduction, Agawu tends towards Kivy's caution in keeping his "play of signs" more formalist than expressive.

The next chapter moves to the heart of Agawu's contribution, a semiotic consideration of "topics" as defined by Leonard Ratner (1980) for Classical music. Topics are types of music that would be readily identified by a listener in the Classical period in terms of their various stylistic or generic derivations. Topics include dance types (minuet, gavotte), marches, hymns, or general styles, such as the "learned style" characterized by an imitative, or at least contrapuntal, texture. Ratner's topics are not pieces in themselves; rather, they are passages in larger pieces. As such, they are often thematized as part of the musical discourse, providing an efficient contextual orientation for the expressive purport of a piece. Topics carry expressive associations drawn from their sources in well known secular or sacred genres. They also suggest a hierarchy of social classes, which can be used to great effect in opera characterization, as Allanbrook (1983) has demonstrated for Mozart's Don Giovanni and Figaro.

Little of this potential explication is plumbed in Agawu's treatment. Instead, we are presented in the next chapter with one kind of formal function that topics might be assumed to support, depending on their character as potential opening, developmental (continuational), or closing material. Such functions are then shown to coordinate with a hierarchical pitch-structural analysis (Schenkerian), thereby leading the analysis to a rhetorical or dramatic account of formal pacing.

With these theoretical tools, Agawu proceeds to analyze one major chamber movement (string quartet or quintet) by each of the principal Classical composers (Haydn, Mozart, and Beethoven) in the middle three chapters of the book. The final two chapters offer a concise outline of his theoretical positions, and a brief application of his topical analysis to early Romantic music.

In terms of semiotic theory, Agawu speaks of signifier/signified relationships, but the signified he claims for his topics are often merely identificational labels, not even the formal functions Agawu skilfully demonstrates. Furthermore, the "universe of topics" is merely listed, without attending to its interesting semiotic problems: (1) what features are sufficient to identify a given topic, and (2) since topics are drawn rather heterogeneously from familiar musical types or styles, how are they oppositional or hierarchically constituted as a coherent system?

Agawu does attribute to his topics one other signified, one all too familiar from a less creditable brand of style analysis:

In the Classic style, topics are not extramusical in the sense of having nothing to offer a syntactic reading of the piece. On the contrary, they are able to account in part for a work s historic specificity, which ultimately locates its syntax in the same historical continuum. (41)

While historical location may be part of the meaning of a topic for contemporary listeners (the stylistic fingerprints that enable us to place an unknown work historically), it is not usually signified to a contemporaneous listener except in those cases where the referenced style is already historical, as when the learned style, derived from the Baroque, is alluded to in a later Classic composition.

After acknowledging that the succession of topics clearly "conveys part of the drama" in the first movement of Mozart's Piano Sonata, K. 322, Agawu is reluctant to go as far as Allanbrook (1983) in exploring the "human dimension of this drama" (1991: 48). Thus, the reader is left with rather pale topical signified, "designated by conventional labels drawn mostly from eighteenth-century historiography"(49), and the caveat that "assigning fixed signification to topics . . . is very much a contextual matter" (50).David Mosley's book presents an even sketchier theoretical argument of some thirty pages before dealing in successive chapters with each of the twelve Eichendorff poems set by Schumann in his second collection titled Liederkreis (Op.39). A three-page conclusion considers poetic and tonal relationships among the songs, and stresses the "intersemiotic transmutation" that occurs between the music and the words. Two short appendices provide information on historical figures leading to the particular semiotic and gestural theories Mosley espouses.

Peirce's system of categories provided the outer framework for Mosley's theory of the intertextual relations that constitute song, but not all the inner workings of that theory. Music contributes to the interpretation of a poem in leading from the Firstness of gesture (music itself,) through the Secondness of song, where music and poem interact, to the Thirdness of poem as sign. In filling out the details in this framework, which starts with the premise that music itself is simply gestural, Mosley, like Agawu, intends his theory to complement the currently popular analytical practices derived from Heinrich Schenker (d.1935).Schenker was a German music theorist whose reductive hierarchies purported to reveal the whole "content" of music as voice-leading (as harmonic and contrapuntal structure alone). Adopting this perspective, Mosley weds it to a parallel reductive scheme for the poetic texts, a method he models on some suggestive remarks by Northrop Frye (1956) about poetry. Mosley's metric and semantic reductions extract significant words or "kernels". The verbal reductions are less systematic than the tonal ones. Words deleted at one stage of his reductions may reappear at a later stage. There is a further difference: kernel words retain their thematic significance and referential specificity, whereas the kernels of his tonal analysis are merely the expressively undifferentiated cadence patterns that define a key.

Mosley's method of interpreting a song is to force the two reductions--harmonic and poetic-to points of convergence in order to establish a kernel expressive meaning (and its Peircean ground) for each song for example "Mondnacht" expresses "longing", and the ground of the relationship between poem and musical setting is "compulsion" (89-90). Longing is expressed, according to this analysis, by delaying a return to the home chord (tonic) suggesting a metaphor of "harmonic desk, the longing of the dominant for the tonic" (89).

Such an analysis engenders a host of problems. It is certainly the case, as he indicates, that longing can be musically expressed by delaying a return to the tonic chord, but surely there are many other ways in which this song could be said to express the archetypal Romantic concept of longing. It does not follow in any case that longing constitutes the whole meaning of the song or that the music realizes this interpretation only because of an intersemiotic interaction with its text. There is no shortage of untexted music in the same stylistic milieu imbued with Kindred feelings.

Schumann's music cannot be educed to mere gesture without losing Its other cognitive levels of expressive meaning (and neither, for that matter can Schenker's analyses Furthermore, a creditable theory must establish the conditions under which Romantic music can express longing, free of an accompanying text For this end, the method of Schenkerian analysis adopted here is inappropriate;it leads our attention away from the details of the music to more standard aspects of structure which guarantee coherence conclusions from his reductions, picking out nuances as crucial gestures because they fit key words. Working at cross purposes to his Schenkerian structuralism, Mosley's instincts seem correct, for much of the expressive interest is found in such nuances as he singles out, where normative progressions are disrupted in various ways, but finding these moments is not enough. There is more meaning in the configuration of the musical surface than can be accounted by its disruption of underlying structures, and this is just what Agawu's topical analysis is at pains to demonstrate. Mosley has not provided any account that rich.

It is ironic that Mosley adopts such a formalist perspective. One might assume that given his subject, he would offer more expressive and poetic interpretation than Agawu, but his model is hampered by its a priori devaluation of music's own semiotic, its own claims to Thirdness as an artistic system capable of producing meaning independently of its interaction with a verbal text. Mosley fails to consider whether the music merely illustrates the words or whether the music might also further interpret the text or, indeed, might at times counter it. The interpretation of Romantic song demands recognition of an interaction between semiotic equals.

As contributions to a new discourse which is still underdeveloped and undersubscribed both these essays in musical interpretation are constructive. Agawu's concentrated study offers a rich analytical discourse sprinkled with a wealth of theoretical insights, each of which could be developed fruitfully in future work. The book is the work of a mature and intelligent musician and music theorist, but it presents a rather narrow view of semiotics, which may well explain its overly cautious explication of expressive meaning. Despite its somewhat limited aims, it is well-written and addresses a current concern of the music theoretical community.

Mosley's invocation of Peirce suggests a step in the right direction for music semiotics, in that Peirce's triadic approach to interpretation leads one to take responsibility for meaning of all kinds and at all stages of inquiry. But with the particular methods he adopts here Mosley has narrowed the interpretive enterprise. Neither Agawu nor Mosley offers a semiotic of music powerful enough to rebut Peter Kivy's rather dour assessment of musical meaning.

Note: For a more technical view of music theoretical issues in Agawu's book, see Hatten (1992).


Allanbrook, Wye (1983) Rhythmic Gesture in Mozart: Le Nozze di Figaro and Don Giovanni. Chicago: Univ. of Chicago Press.

Frye, Northrop (1956) "Melos and Lexis." In Sound and Poetry, Northrop Frye (ed.). New York: Columbia University Press.

Hatten, Robert (1992). Semiotics, Semiology and the Problem of Meaning in Music. Double review of Agawul Playing with Signs, and Nattiez, Music and Discourse. Music Theory Spectrum (to appear).

Kivy, Peter (1980). The Corded Shell; Reflections on Musical Expression. Princeton: Princeton University Press.

---. (1984). Sound and Semblance; Reflections on Musical Representation . Princeton: Princeton University Press.

--- (1990). Music Alone; Philosophical Reflections on the Purely Muscial Experience. Ithaca: Cornell University Press.

Mead, George Herbert (1934) Mind, Self and Society. Chicago: University of Chicago Press.

Mendelssohn-Bartholoy, Felix (1945). Lettersl G. Selden-Goth, ed. (New York: Pantheon)

Ratner, Leonard (1980) Classic Music; Expression, Form and Style. New York: Schirmer.

Rosen, Charles (1972). The Classical Style. New York: Norton.

Robert Hatten, who teaches music at Pennsylvania State University in University Park, writes extensively on musical semiotics and has recently completed the book, Musical Meaning in Beethoven: markedness, correlation and interpration.

Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives