A Theory of Musical Narrative
168 pages

Vous pourrez modifier la taille du texte de cet ouvrage

A Theory of Musical Narrative


Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus
168 pages

Vous pourrez modifier la taille du texte de cet ouvrage

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus


Byron Almén proposes an original synthesis of approaches to musical narrative from literary criticism, semiotics, historiography, musicology, and music theory, resulting in a significant critical reorientation of the field. This volume includes an extensive survey of traditional approaches to musical narrative illustrated by a wide variety of musical examples that highlight the range and applicability of the theoretical apparatus. Almén provides a careful delineation of the essential elements and preconditions of musical narrative organization, an eclectic analytical model applicable to a wide range of musical styles and repertoires, a classification scheme of narrative types and subtypes reflecting conceptually distinct narrative strategies, a wide array of interpretive categories, and a sensitivity to the dependence of narrative interpretation on the cultural milieu of the work, its various audiences, and the analyst. A Theory of Musical Narrative provides both an excellent introduction to an increasingly important conceptual domain and a complex reassessment of its possibilities and characteristics.


Part 1. A Theory of Musical Narrative
1. An Introduction to Narrative Analysis: Chopin's Prelude in G Major, Op. 28, No. 3
2. Perspectives and Critiques
3. A Theory of Musical Narrative: Conceptual Considerations
4. A Theory of Musical Narrative: Analytical Considerations
5. Narrative and Topic
Part 2. Archetypal Narratives and Phases
6. Romance Narratives and Micznik's Degrees of Narrativity
7. Tragic Narratives: An Extended Analysis of Schubert, Piano Sonata in B<FLAT> Major, D. 960, First Movement
8. Ironic Narratives: Subtypes and Phases
9. Comic Narratives and Discursive Strategies
10. Summary and Conclusion




Publié par
Date de parution 27 octobre 2017
Nombre de lectures 0
EAN13 9780253030283
Langue English
Poids de l'ouvrage 1 Mo

Informations légales : prix de location à la page 0,0087€. Cette information est donnée uniquement à titre indicatif conformément à la législation en vigueur.


A Theory of Musical Narrative
Robert S. Hatten, editor
Approaches to Meaning in Music
Byron Alm n and Edward Pearsall
Voicing Gender: Castrati, Travesti, and the Second Woman in Early Nineteenth-Century Italian Opera
Naomi Andr
Neil Young and the Poetics of Energy
William Echard
Interpreting Musical Gestures, Topics, and Tropes: Mozart, Beethoven, Schubert
Robert S. Hatten
Musical Meaning in Beethoven: Markedness, Correlation, and Interpretation
Robert S. Hatten
Intertextuality in Western Art Music
Michael L. Klein
Is Language a Music? Writings on Musical Form and Signification
David Lidov
Pleasure and Meaning in the Classical Symphony
Melanie Lowe
The Musical Topic: Hunt, Military and Pastoral
Raymond Monelle
Musical Representations, Subjects, and Objects: The Construction of Musical Thought in Zarlino, Descartes, Rameau, and Weber
Jairo Moreno
Deepening Musical Performance through Movement: The Theory and Practice of Embodied Interpretation
Alexandra Pierce
Expressive Forms in Brahms s Instrumental Music: Structure and Meaning in His Werther Quartet
Peter H. Smith
Music as Philosophy: Adorno and Beethoven s Late Style
Michael Spitzer
Music and Wonder at the Medici Court: The 1589 Interludes for La pellegrina
Nina Treadwell
Debussy s Late Style: The Compositions of the Great War
Marianne Wheeldon
A Theory of Musical Narrative
Bloomington Indianapolis
This book is a publication of
Indiana University Press
Office of Scholarly Publishing
Herman B Wells Library 350
1320 East 10th Street
Bloomington, Indiana 47405 USA
First paperback edition 2017
2008 by Byron Alm n
All rights reserved
No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher.
The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences-Permanence of Paper for Printed Library Materials, ANSI Z39.48-1992.
Manufactured in the United States of America
The Library of Congress has cataloged the original edition as follows:
Alm n, Byron, [date].
A theory of musical narrative / Byron Alm n.
p. cm. - (Musical meaning and interpretation)
Includes bibliographical references and index.
ISBN 978-0-253-35238-5 (cloth : alk. paper) 1. Music-Philosophy and aesthetics. 2. Musical analysis 3. Music-Semiotics. 4. Music and literature. 5. Music theory. I. Title.
ML3800.A46 2008
ISBN 978-0-253-03009-2 (pbk.)
ISBN 978-0-253-03028-3 (eb)
1 2 3 4 5 22 21 20 19 18 17
Phil Abalan
A. DeWayne Wee
J. Peter Burkholder
Robert S. Hatten
and, as always,
For Sarah
1. An Introduction to Narrative Analysis: Chopin s Prelude in G Major, Op. 28, No. 3
2. Perspectives and Critiques
3. A Theory of Musical Narrative: Conceptual Considerations
4. A Theory of Musical Narrative: Analytical Considerations
5. Narrative and Topic
6. Romance Narratives and Micznik s Degrees of Narrativity
7. Tragic Narratives: An Extended Analysis of Schubert, Piano Sonata in B Major, D. 960, First Movement
8. Ironic Narratives: Subtypes and Phases
9. Comic Narratives and Discursive Strategies
10. Summary and Conclusion
The inspiration for this project dates back to 1992, to the preliminary research period of my dissertation Narrative Archetypes in Music: A Semiotic Approach (1996), and to my near-simultaneous discovery of three books from disparate fields: Northrop Frye s Anatomy of Criticism (1957), Eero Tarasti s A Theory of Musical Semiotics (1994), and James Jak b Liszka s The Semiotic of Myth (1989).
Frye s book, an acknowledged masterpiece, is a remarkable taxonomic rewriting of the principles of literary criticism; its most influential constituent essay, Archetypal Criticism, introduces his four mythoi -romance, tragedy, irony, and comedy-that represent fundamental, pregeneric patterns of narrative motion. This formulation has influenced countless scholars in many fields, most notably Hayden White, who has observed (1973) the tendency of historians to consciously and unconsciously emplot historical events according to temporal narrative schema. I had been acquainted with these mythoi since high school, but my first reading of the essay in 1992 convinced me that they are eminently applicable to music.
Tarasti s book was very nearly my first introduction to the semiotic discipline. Although I was not then familiar with Charles S. Peirce, Ferdinand de Saussure, or Algirdas Julien Greimas, several aspects of Tarasti s writing immediately appealed to me. First, it is systematic and thorough (although his writing style is quite expansive), but these qualities never unseat his sensitive musical insight. Second, his application of the notion of modality to music to account for the encoding of human values into musical discourse seemed to offer a way out of the arbitrary assignment of expressive characteristics to music. Third, his willingness to tackle a large conceptual terrain and a broad representation of musical literature was refreshingly ambitious and welcome. With respect to my own development, Tarasti was an important model for bringing together methodological rigor, solid musical intuition, and an eclectic breadth of interests. My choice of title for this book thus represents an acknowledgment of the debt I owe to his example.
Frye s deductive taxonomic system and Tarasti s inductive analytical methodology embody balancing impulses that might work effectively together. The means to achieve this balance in the current volume is accomplished by The Semiotic of Myth . It does for the field of mythology what I am attempting to do for music: locate an analytically rigorous approach to narrative within a socially and psychologically methodological frame-and it specifically invokes Frye s mythoi as its upper-level taxonomic principle.
Music, like mythology, is a temporal phenomenon, and both are amenable to narrative organization. Liszka s concept of narrative as transvaluation -the change in markedness and rank within a cultural hierarchy over time-is crucial for the understanding of musical narrative, not only because it sidesteps lengthy detours into literary narrative theory, but because it accounts for the social and psychological function of narrative: revealing the implication of the necessary conflict between the violence imposed by hierarchy and the violence required to counter it (Liszka 1989: 133). This factor informs critiques of musical discourse that reinforce the status quo (Adorno, McClary) and transformative approaches that implicate music as a vehicle of change and challenge (Schoenberg). It also allows the analyst to see music as a mirror of psychic processes of development and integration.
Understanding narrative as transvaluation also bridges the rhetorical gap between context-centered and structure-centered approaches to music. I suggest in this volume that it is both possible and desirable to balance careful attention to musical details with an attunement to music s location within a social and significatory network. Further, the recognition-from Frye-that there are multiple and functionally equivalent realizations of narrative conflict prevents narrative analysis from becoming a force either for reactionary repression or for the relativistic erosion of all stable value systems.
In the years since I encountered these works and began developing my approach, I have encountered other studies that have effectively investigated issues of expression in combination with analytical rigor and cultural insight. In particular, the research undertaken by Robert S. Hatten and Vera Micznik has proved very influential to me. To Hatten s work I owe my integration of topic into the multi-leveled signifying network of musical narrative. In Micznik s narrative writings I found an effective methodology, entirely different in flavor from that of Tarasti, that supported the analytical eclecticism of this volume.
I have taken care to make the following theory of musical narrative as widely applicable as possible without sacrificing its necessary conceptual weight. One should not come away from this volume with a method for analyzing musical narrative. Instead, a theoretical basis for understanding the implications of narrative analysis is given, along with a range of analytical approaches that are methodologically suggestive rather than prescriptive. In that I am both engaging with the primary texts of musical narrative theory and presenting a flexible theoretical apparatus for investigating narrative, this book can serve as a graduatelevel text on musical narrative that does not unduly limit the student s latitude to move beyond the provided examples.
With respect to the field of music narrative theory, I have also been influenced by Carolyn Abbate, Edward T. Cone, Nicholas Cook, M rta Grab cz, Marion Guck, Peter Kivy, Michael Klein, Lawrence Kramer, Fred E. Maus, Susan McClary, Patrick McCreless, Raymond Monelle, Jean-Jacques Nattiez, Anthony Newcomb, Jann Pasler, and many others. I hope that I have done justice to their ideas and to the spirit of their work in this volume.
A Theory of Musical Narrative casts a wide analytical net: the music analyzed in this volume features works from the early eighteenth century to the 1960s. My primary intent is to illustrate the breadth of focus of musical narrative as an analytical enterprise, and I hope that I will be forgiven for any overreaching that may result in the interests of promoting an eclecticism of approach.
Chapter 1 provides a brief narrative analysis of Chopin s G-major Prelude, op. 28, no. 3 that introduces some of the key concepts and theoretical assumptions of my approach. In chapter 2 , I consider various explicit and implicit approaches to narrative analysis by J r me-Joseph de Momigny, Eero Tarasti, and Susan McClary, looking for a disciplinary consensus about musical narrative. I also examine the important critiques of musical narrative in light of that consensus and my reconfiguration of it. Chapters 3 and 4 are concerned with the explication of the theory itself. Here I introduce important concepts from semiotics, narrative theory, and the works of Frye, Liszka, Tarasti, and Micznik, among others, that significantly contribute to that theory.
The focus shifts in chapter 5 to the balance of theoretical and analytical elements that will characterize the remainder of the book. In this chapter, I consider the relationship between musical topic and musical narrative theory, with particular attention to their differences in scope and complexity and to topic s various intersections with the three levels of narrative analysis discussed in chapter 4 .
Chapters 6 - 9 each illustrate one of the four narrative archetypes, corresponding to Frye s four mythoi . Both extended and abbreviated analytical examples are given to flesh out the definitions of the archetypes. In addition, each chapter considers a different set of theoretical issues. Chapter 6 illustrates the romance archetype and features an exploration of Vera Micznik s analytical methodology and her application of Gerald Prince s notion of degrees of narrativity, applied to several early symphonic movements of Gustav Mahler not covered by Micznik in her 2001 article. Chapter 7 , on the tragic archetype, contains an extended analysis of the first movement of Franz Schubert s B -major Piano Sonata, using a parametric approach roughly derived from Tarasti (1994). The various methodological orientations taken in these two chapters are meant to suggest an eclectic approach to application of the theoretical concepts previously outlined.
Chapters 8 and 9 are more taxonomic in nature, serving to refine the four archetypal categories with respect to a number of additional criteria. In chapter 8 , concerned with the ironic archetype, the notion of archetypal phases is introduced, subdividing the four archetypes according to various oppositions. Other subcategories relating to the rhetorical situating of interpretation, the role of topic, and the musical elements used to formulate narrative conflict are considered here. Chapter 9 examines the comic archetype and proposes a partial list of comic discursive strategies: typical patterns of discursive musical organization organized according to similar features.
Finally, chapter 10 presents a summary of the theory and its associated analytical components. A glossary of new or appropriated specialized terms is also given to help the reader navigate the concepts in this volume.
I hope that readers will find much to interest them in the following pages. I have been profoundly enriched by this intellectual journey, and I am delighted to share my thoughts with the wider community of students and professionals.
Many individuals contributed in significant ways to the development of this project. Since my research into this topic began more than fifteen years ago, I am afraid that there will be some omissions from the following list, for which I offer my apologies and belated thanks.
First, I am profoundly grateful for the advice, encouragement, and input of J. Peter Burkholder of Indiana University. Dr. Burkholder was the primary advisor for my dissertation, which represented my first stab at this subject, but he was also involved at a much earlier stage, and provided great insight and many suggestions without which this book would not be possible. His profound knowledge of music literature, his perceptive editing abilities, his interdisciplinary acumen, and his valued friendship continue to inform my contributions to the field of music scholarship.
I would also like to thank Drs. Marianne Kielian-Gilbert, Lewis Rowell, Mary Wennerstrom, A. Peter Brown, and David Neumeyer for their expertise and assistance on countless occasions over the years in matters of semiotic terminology, analytical insight, and cultural implications.
Next, I am fortunate to have an excellent group of colleagues in the theory department at the University of Texas at Austin who have contributed many hours of conversation, honest feedback, insightful dialogue, moral support, and social camaraderie. To these colleagues-James Buhler, Eric Drott, David Neumeyer (again), Edward Pearsall, Winton Reynolds, and Marianne Wheeldon-I offer my humble thanks. I would like also to recognize the many graduate and undergraduate students with whom I have had the pleasure to work.
The field of musical narrative and the study of music expression are disproportionately populated with remarkable scholars who also happen to be delightful individuals. Although I do not know everyone in these small but important fields, I have nevertheless drawn inspiration from all of them. For those I have had the privilege to work with-Patrick McCreless, Jann Pasler, Nicholas Cook, Vera Micznik, Michael Klein, Marion Guck, and Fred Maus-I have nothing but the greatest respect and admiration. Thanks also to the growing community of Jungian music scholars, especially Jeffrey Kurtzman and Robin Wallace, for their contributions and for opening up a productive new dialogue between music and Jungian studies.
A special word of thanks is due to Robert S. Hatten, whose work has been an inspiration to me for over a decade, and whose support and interest in my own contribution to musical narrative have meant a great deal. I thank him for his many editorial contributions to this volume and for the opportunity to make it available to the public. I am also deeply indebted to Eero Tarasti and James Jak b Liszka for their comments on this volume and for their obvious and lasting impact on my work.
Thanks to everyone at Indiana University Press-proofreaders, editors, indexers, and support staff-for helping me to get our second collaborative project from proposal to publication. Special thanks to Miki Bird, Jane Behnken, Donna Wilson, and Katherine Baber for answering my many e-mails and phone calls with good humor and prompt attention. I am also particularly indebted to Ryan Beavers, whose excellent engraving skills saved me many an hour searching through the manuals of my music notation software, and to Carol Kennedy for her excellent copyediting.
I am grateful to the University of Texas School of Music and College of Fine Arts-and in particular to Glenn Chandler, Michael Tusa, and Douglas Dempster-for the administrative and financial assistance they provided to enable this book to be written.
I am especially indebted to my parents, Anthony and Marilyn, for their love, support, and encouragement, and also to many friends and relatives-especially Sue Alm n-Whittaker, Alice Bartusiak, Leslie Bush, Doug Blake, Patrick Boley, Pat Budelier, Victoria Redfearn Cave, Rick and Catherine Colvin, Leonie Esselbach, Philip Ford, Joanna Gentry, Brice Gerlach, Jennifer Kropp, Joan and Noel Kropp, Linda and Gary Lawrence, Andrea Marttila, Liz Matassa, Peter Mowrey, Barbara Myrom, Olga Myrom, Ludim Pedroza, Erika Pierson, Kurt Runestad, Matt Whittaker, and everyone at Westminster Presbyterian Church.
Finally, as always, I am eternally grateful to my wife, Sarah, for being there every day and going through it all with me. I couldn t ask for a better friend and companion.
Part One:
A Theory of Musical Narrative
1 An Introduction to Narrative Analysis: Chopin s Prelude in G Major, Op. 28, No. 3
When we think of narrative music, certain assumptions come quickly to mind-assumptions that have strongly colored our responses to the topic. First, narrative music is often thought to be in some way problematic or idiosyncratic; that is, we tend to resort to narrative interpretations when traditional formal, harmonic, and generic paradigms do not apply. Anthony Newcomb (1984, 1987), for example, has located a certain type of narrative music within a particularly nineteenth-century mode of expression that attaches plot archetypes to nonstandard or unusual compositional designs. Second, narrative music tends to be associated with programmaticism, dramatic or epic texts, evocative titles, or any of a variety of attachments that ready the listener to hear the music in a special way. Carolyn Abbate (1991) and Jean-Jacques Nattiez (1990a) have both discussed the problematic semantics of musical discourses that require supplementation. Third-and most tellingly-narrative music is typically understood as a derivative phenomenon. Its formal strategies, subject matter, and critical metalanguage are all apparently imported from literature or drama.
This book endeavors to question some of these common assumptions about narrative music and to suggest that narrative organization is far more normative and common than is generally conceded. This is not a new endeavor: Fred E. Maus, Vera Micznik, Michael L. Klein, Eero Tarasti, and others have suggested alternative ways of conceiving of narrative music and have developed powerful analytical tools to examine it. In the following chapters, I will argue that a new consensus is developing about musical narrative that is aware both of the limitations of musical expression and of the rich potential of music as a narrative medium.
Before addressing the theoretical basis and possible pitfalls of musical narrative design, let us look ahead at the ground to be covered by means of a short example. The piece to be analyzed here is Chopin s Prelude in G major, op. 28, no. 3, chosen for its nonconformance with the traditional notions of narrative music. It does not feature a particularly unconventional formal design, and it is purely instrumental, containing no textual or programmatic cues that would suggest a narrative trajectory. Yet our analysis, derived firmly from the musical discourse, articulates just such a trajectory. This short example suggests that if we reimagine the conceptual basis of narrative theory and practice, we will find in it a rich field of study and insight.

Example 1.1. Chopin, Prelude in G major, op. 28, no. 3, motives a (measures 3-4) and b (measures 4-6)
The Prelude begins with the establishment of a sixteenth-note ostinato figure that becomes the accompaniment to a lilting, dance-like melodic line. (This figure can be seen from measure 3 onward in example 1.2 below; the opening measures appear as example 5.1 in chapter 5 .) The undulating, tonally stable, and repetitive character of Chopin s ostinato figure evokes the hypnotic stasis-through-motion of the Romantic Spinnerlied , suggesting an atmosphere of rustic simplicity. It further recalls certain ubiquitous seventeenth- and eighteenth-century Figuren that appear in nature-pictorial movements and represent running water or gentle breezes. As a musical topic, we might describe the overall expressive effect as harmony-with-nature, nature being understood in its gentler aspect. If we trace this figure through the entire Prelude, we see that it is employed continuously, undergoing occasional changes of harmony, until the final measures. There it is doubled by the right hand and subjected to fragmentation and a registral ascent prior to the two final tonic chords that signal the work s conclusion. Overall, one function of the topic harmony-with-nature in this piece is to provide a specific background environment within which the thematic material can move.
Measures 3-6 feature the first melodic phrase of the piece, which divides into two uneven subphrases distinguished by contrasting registral spaces and motivic directional contours, as shown in example 1.1 .
These motives are obviously related: both are harmonically stable, and, like the accompaniment figure underneath, they both arpeggiate the tonic triad. Further, both contain short, dotted rhythmic cells that give them a lilting yet dignified quality and suggest a dance-like derivation. Despite these similarities, several musical aspects contribute to motivic contrast. The first motive, a , with its predominately half-note pulse and upward octave thrust in the piano s middle range, is more buoyant than the second. The second motive, b , reverses the upward directionality of a , is located almost an octave higher, and contains slower note values and a decreasing dynamic level. Thus, an element of striving upward in a is answered by the yielding descent of b , but in a higher register, suggestive of distance and a lack of energy. Like the accompaniment, the a motive s upward striving is self-contained and uncomplicated: largely arpeggiating the tonic triad, it results in an impression of assuredness rather than restlessness. Motive a s dynamic character is reinforced by its beginning on the fifth, rather than the root, of the tonic key. The persistence of the tonic key in motive b along with the termination of that motive on the tonic pitch seems to suggest an endorsement of a s tonal and registral motion.
The relationship between a and b , however, lies somewhere between oppositional and complementary. On the one hand, their registral separation implies a need for its removal to overcome a sense of motivic opposition. The oppositional aspect is further outlined by the contrast in directional contour, with the profile of b seeming to give way in the face of a s ascent. On the other hand, there are rhythmic similarities between a and b that suggest an echo effect between the two motives. The anacrusis-downbeat rhythm that concludes motive a in measures 3-4 is also used to initiate motive b in measures 4-5, giving the latter motive the character of a retrograde, continuation, or reprise of the first movement. In a certain sense, a and b can be considered two halves of a larger figure-note that b inverts the intervallic content of a . Their operation suggests a kind of inner narrative, akin to what Robin Wallace (1999) has termed an introverted reading, in which two aspects of a single personality enter into dialogue.
Three further, more hidden aspects seem to support this interpretation. The first is the continuity of the melodic line outlined by the combined motives, which (in measures 3-5) describes a stepwise segment E-D-C-B that is hidden by the registral shift. The second aspect is the continuation of the registral ascent of motive a (D4 to B4 to D5) up to the B5 in motive b (measure 5, beat 1). The third aspect is the resemblance of the entire melodic profile a+b to the notes of the accompaniment figure. The pitches in a+b (D-B-E-D-C-B-G) are found in exactly the same order in the accompaniment figure (notes 2, 5, 8, 9, 10, 11, and 13; see also measures 3-6 in example 1.2 , below), and both contain the climactic neighbor-tone figure E-D. (I consider the upper line to be the melodic line, although the right-hand material is actually in parallel thirds, lending a euphoric quality to the melody. In the subsequent discussion of the melodic line, the parallel voice will be considered with respect to issues of register.)
The narrative implication, borne out in later material, is that motives a and b constitute the Prelude s primary oppositional elements. These motives possess some form of kinship, but are prevented from realizing this kinship by two factors: their registral separation and the directionality of their melodic contours. The narrative program in this case consists of various attempts to bring the two motives into a more harmonious relationship based on the removal of these two obstacles. (Note in particular the registral overlap of a fourth, B4-E5, between the upper voice of a and the lower voice of b that will play a role in the eventual resolution of this conflict. The overlapping space thus becomes-both registrally and semantically-the area of common ground between the two motives.)
Motives a and b , thus understood, function as anthropomorphized syntactic units, or musical agents , within the meaningfully unfolding temporal process of a narrative trajectory . (See the glossary for the definition of this and other terms.) The character of this trajectory is generated by the fluctuating nature of their relationship, which sets up a semiotic opposition between the potential for relatedness and the potential for separateness. The former is musically realized by the various similarities between motives a and b : their dotted rhythmic cells, their high degree of harmonic stability, their shared initial arpeggiation of the tonic triad, and their overlapping registral compasses (the span B4-E5 is common to both). In addition, they combine to form a linear voice E-D-C-B at the middleground level, suggesting an organic connection through this hidden contiguity. This connection is reinforced by the continuation of the registral ascent of a into the beginning of motive b , preventing too great a disjunction between them. Finally, the two motives both partake of the character and shape of the accompanimental figure: apart from sharing its lilting, dignified, and dance-like quality, the composite contour of the right-hand phrase, with its staggered arch shape, is a reflection of the same shape found in the left-hand part.

Example 1.2. Chopin, Prelude in G major, op. 28, no. 3, measures 3-33
By contrast, the potential for separation is also available to be exploited as the piece progresses. As mentioned above, this potential is primarily expressed through registral separation, contrasting directional contours, and differences in the respective sequences of rhythmic events.
This conflict between two possible paths that the music might take-toward an apparently restored unity or toward a greater degree of distinctness-can be expressed in terms of a specifically narrative opposition between an order-imposing hierarchy and a transgression of that hierarchy (discussed more fully in chapter 4 ). The listener must then determine the standpoint from which this opposition will be interpreted. (Is the reestablishment of an order-imposing hierarchy to be understood as desirable or undesirable? To the contrary, would a definitive departure from this hierarchy be understood as desirable or undesirable?) In this case, the topical environment plays a significant role in suggesting narrative context. Given that the peaceful, pastoral accompanimental frame, the major tonality, and the leisurely tempo all suggest calm and avoidance of conflict, the listener might be inclined to prefer a synthesis or mediation (a new hierarchical order ) rather than a separation or transgression. Thus, the narrative trajectory will involve the question of whether the centrifugal elements of the two motives will lead to fragmentation or to synthesis.
What the Prelude in fact unfolds is a narrative romance, a reestablishment-through a registral and directional synthesis-of order, of the kinship between motives a and b .
As shown in example 1.2 , measures 7-12 contain a modified version of the primary melodic material and serve to move away from relative stability and balance: the narrative action commences in an attempt to mediate between the contrasting elements from measures 3-6. The motive now in play (measures 7-8 and 9-10) is a variant of both a and b and represents a first, unsuccessful attempt at mediation between them. Its rhythmic profile most resembles a , absent the first pitch, but the lengthened initial note and the high registral location (above the aforementioned overlapping region B4-E5) suggests the less energetic b motive. The melodic profile still emphasizes an ascent, but one made less emphatic by the hesitancy of the initial appoggiatura note F , and the motive ends on a melodically unresolved high A. The harmonic motion away from the tonic to a tonicization of the dominant further contributes to this motive s insufficiency as an effective mediation between a and b ; its repetition, reintroducing the tonic key through the addition of a seventh to the final chord, has the effect of a halfhearted but unfruitful insistence. Leading into the second section, the repeated dotted anacruses of measure 11 present in succession the two as yet unconnected registral spans.
An exact repetition of measures 3-6 occurs in measures 12-15, including a return to the tonic-the obvious tonal location or goal for an attempt at mediation. Also recurring, however, is the original separation of melodic contour and register-the transgression. The difference between this passage and its earlier appearance is that the listener has now experienced the intervening material, such that measures 12-15 are no longer an initial condition, but a retreat from activity and a return to the status quo.
Measures 16-19 contain the first in a pair of phrases representing the expressive climax, phrases that together outline an octave descent from G5 to G4 (ending in measure 26). Instead of resolving the narrative conflict, the phrase intensifies it through a move to the subdominant-from the sharp side to the flat side, as it were-and an emphasis on the chromatic F in measure 16. Recalling Tarasti s (1994) notion of key regions as related to the home key as distant locales to a spatial here, this harmonic motion has the effect of overshooting the goal. Further, the qualities of each motive are present, but do not coexist peacefully. On the one hand, the rhythms of the melodic material derive from an almost obsessive repetition of the rhythms of the a motive, producing a sense of restless agitation. The directional profile of a , an assured arpeggiated ascent, has been replaced by a weaker reiterative figure that cannot rise higher than the F but descends inevitably to E in measure 18. On the other hand, the melody appears in the register of b and displays its characteristic descent, but without the intervallic and rhythmic contour of b .
The second phrase of the pair (measures 20-27) is the crucial section with respect to the narrative, in that it enacts a resolution of the initial oppositions. This process is highlighted by an active return to tonic (IVVI) that passes through the previously tonicized subdominant and dominant regions (the harmonic locations of the earlier mediation attempts). In this section-the longest unbroken melodic span of the piece-the melodic descent combines the rhythmic profiles of a and b into a single, extended line. This passage also evokes a previous attempt at a synthesis: the rhythm in measures 21-26 is identical to the attack pattern of measures 7-10. The rhythm of a appears in measures 20-21 and, in a partially augmented form, in measures 24-26, while the rhythm of b appears in measures 22-23 and 24-25. These relationships are summarized in example 1.3 .

Example 1.3. Motivic variants of a and b in the resolution passage (measures 20-26)
The registral placement also supports reconciliation, since the phrase opens within the register shared by a and b (in measures 3-6). Furthermore, the two motives are no longer separated by rests, and the melodic contour is singly directed toward the tonic pitch G, with no directional opposition as before. A synthesis of both motives, supported by rhythmic, harmonic, melodic, and registral elements, has taken place. The rhythmically assured, downbeat-oriented character of a has combined with the yielding/accepting melodic contour of b . Most importantly, since the harmonic placement of each section has functioned as a commentary on the appropriateness of the mediation, the arrival of tonic harmony in measure 26 indicates the resolution of the narrative conflict. Recalling the piece s initial measures, the accompanimental passage in measures 26-27 clears the stage of musical agents.
The final passage of the piece (measures 28-33) features a repositioning of this accompaniment figure-the topical environment-into the foreground. As this figure ascends through the entire registral space of the piece in an effect akin to gap fill (Meyer 1956: 128-50), it effaces the narrative action, subordinating that action to the pastoral background, and restoring a sense of wholeness. The motion to the extreme upper keyboard register lends a transcendent quality, affirming the piece s synthesizing teleology. The final two chords abstractly present the combined registral space of the accompaniment figure and the a-b synthesis. The overall effect of the postlude is to universalize the narrative action by filling the registral space and to complete the pastoral frame.
This interpretation is clearly not the only narrative analysis the piece could support. I have chosen to put forward an introverted narrative of similarity and difference with respect to variations of a single, underlying motive, rather than an extraverted narrative of motive against motive or theme against theme. The above interpretation is effective insofar as it acknowledges the pivotal role of the overall topical environment in both setting up expectations (a restful non-tragic or non-ironic narrative) and reinforcing the resultant narrative trajectory. The analysis has also suggested that narrative involves the coordination of multiple elements: the articulation of conflicting elements or possibilities, their temporal engagement resulting in shifts of hierarchical emphasis, and an interpretive frame that establishes a meaningful perspective of the whole. In subsequent chapters, the implications of these suggestions will be developed, with attention being given to points of consensus among scholars of musical narrative, a consideration of critical questions, the articulation of a self-consistent theory, and the relationship of that theory to analytical and interpretive applications.
Portions of this chapter appeared in altered form as part of Alm n 2004 .
2 Perspectives and Critiques
The problematic status of musical narrative as a disciplinary entity today is reflected in a general disagreement about its nature, properties, and range of application. Some scholars have ascribed it primarily to programmatic music (Kivy), while others expand its reach to music that is in some manner formally problematic (Newcomb, Abbate) or to a broader spectrum of works including absolute music and instrumental genres (Maus). A further group (Nattiez, Abbate, Kramer) has questioned its applicability to music. Musical narrative has variously denoted a loose analogy between literary and musical patterning, a historically bounded and primarily Romantic compositional impulse (Newcomb), a meta-analytical rhetorical strategy (White), and a background structure that enables a grammar of meaning (Greimas, Tarasti). It has been employed as a hermeneutic alternative to formalist approaches to music (Hatten), as a way of recovering a Romantic compositional aesthetic (Newcomb), and as the basis for analytical projects. Approaches to musical narrative have emphasized its performative aspects (Maus, Tarasti, Abbate) and investigated its potential for reflecting or addressing social dynamics (McClary).
Common to virtually all approaches to musical narrative is the recognition of a degree of similarity between musical and literary discourse. Having taken this cue, music scholars have tended to appropriate the methodology and vocabulary of literary approaches to narrative, using these approaches as measuring sticks to determine music s propensity for narrative organization. Debates about musical narrative thus tend to concern themselves with the central issues of literary narrative (with respect to which the canonic figure is Genette): plot, agency, temporal manipulation, the role of the narrator, and genre. Alternative narrative theories-notably those of Roland Barthes (by McCreless), A. J. Greimas (by Tarasti and Grab cz), and Paul Ricoeur (by Grab cz, Klein)-have also been investigated. The results of these inquiries have been mixed: while it is possible to identify approximate musical correlates to literary elements (theme or motive agents; formal schema plot archetypes, etc.), such correlations create as many problems as solutions. In fact, a significant proportion of music-narrative literature has concerned itself with articulating (Kivy, Abbate, Nattiez, Kramer) or rebutting (Micznik, Alm n, Klein, Maus) arguments against too close an affinity between music and literature: the absence in music of referentiality, a subject-predicate relationship, a narrator, and a past tense.
I will suggest in the last section of this chapter that such arguments can be reasonably addressed even if their premises are granted. However, we might first question whether the premises are correct, and consider what unfortunate side effects follow from accepting them. By allowing literary narrative theory to establish the first principles of narrative, we virtually ensure that musical narrative will generate its narrative meaning parasitically, via a continuous referring back to a hypothetical and imaginary verbal equivalent. To understand narrative meaning in this manner, we would have to consider what a similarly constituted literary analog might employ, and we would constantly be frustrated by all those elements missing in music that literature displays more effectively. If this is how we derive musical narrative meaning, wouldn t writing a story be much more effective?
Perhaps, then, musical narrative really is a conceptual phantom, a phenomenon to be enlisted in the presence of texts, descriptive titles, and programs that lend their literary qualities to musical signification. Or perhaps, with Maus (2005), it would be better to scale back our theoretical claims, proceeding from a more modest analogical relationship between music and narrative, a relationship buttressed by its obvious appeal for many listeners and the analytical insight that it allows.
Although I fully concur that musical narrative, construed as an analogy, is sufficiently motivated to render it useful and productive for interpretive attention, I do not believe it necessary to throw aside all hope of establishing more substantive theoretical foundations. Instead, I submit that the definition of narrative itself is the source of confusion: because narrative was first conceptualized in relation to literature, we have largely failed to recognize the distinction between narrative proper and narrative as manifested in literature. Lacking such a distinction, a clear understanding of any specifically musical manifestations of narrative, should they exist, would be impossible.
To use a genealogical metaphor, I prefer a sibling model rather than a descendant model for articulating the relationship between musical and literary narrative. The descendant model presupposes a conceptual priority for literary narrative, while the sibling model distinguishes between a set of foundational principles common to all narrative media and principles unique to each medium.
The traditional descendant model presents musical narrative as a derivative phenomenon: it is effective only to the degree that the musical work is able to mimic or approximate the effects of literary narrative. Using this model, we are bound to view music as insufficiently and ineffectively narrative: it apparently lacks semantic specificity, a recognizable narrator, and coherent characters. Music s native significatory processes are thereby deemphasized, while those arising from the mixture of media are given disproportionate priority. Admittedly, our understanding of the signifying properties of program music, music with text, or leitmotivic music have thereby been greatly enriched-and we learn more about music signification by examining the way it incorporates elements from other media. Research has also uncovered important musical analogs to literary devices: theme-actors (Tarasti 1994), framing elements (Abbate 1991, Hatten 1997), cues delineating a narrated past tense (Hatten 1991), and so on. However, it is also critical that the boundary lines between music, literature, and other temporal signifying media be clearly drawn. In the absence of such clear delineation, we risk mistaking breakdowns in translation between media for failures of signification. Furthermore, there is a danger of overinterpreting musical works to render them more like literary events, by, for example, assigning them too great a degree of referential specificity. If musical denotation is seen as a lack in relation to literary denotation, it will be easier to fill in the gaps by projecting dramatic scenarios that strain the credulity of those already skeptical of hermeneutic approaches.
I therefore suggest that the sibling model, which posits an indirect relationship between musical and literary narrative as distinct media sharing a common conceptual foundation, is the more productive one. Such a model separates narrative universals from those arising from specific media, obviating many of the difficulties attached to the descendant model. With respect to the former, I will understand narrative as articulating the dynamics and possible outcomes of conflict or interaction between elements, rendering meaningful the temporal succession of events, and coordinating these events into an interpretive whole. With respect to the latter, I will consider music s own syntactic potentialities, its own devices for negotiating conflict and interaction, in ways that reframe problematic issues in productive ways. Music s lack of semantic specificity might, for example, be viewed as a positive characteristic, in that music can display narrative activity without being limited to specific characters and settings.
A theory of narrative that recognizes the different languages and organizing principles of literature and music would not be focused on the question, how is music really like literature in disguise? Instead, it would highlight far less intractable issues: identifying common essential elements of narrative and the ways in which music uniquely employs these elements, understanding the differences between narrative and non-narrative music, and devising useful strategies for integrating narrative theory with analysis and historical studies.
As an additional example, consider the differences in the representation of temporal events in drama, literature, and music. In a dramatic medium, the audience sees before them a specific realization of those events as portrayed by actors; the strength of a dramatic presentation thus depends in part on this phenomenal immediacy and impact. In a literary medium, the reader must construct a mental image of events; the strength of a literary presentation thus depends in part on the reader s greater flexibility in imagining character and setting. But if descriptions of character and setting in literature allow for a degree of immediacy and verisimilitude, music provides even greater flexibility for the listener to track the interplay of narrative relationships. The relative freedom from descriptive specificity in music allows the dynamic interactions between events to be foregrounded, interactions that are fruitfully homologous with psychological and social dynamics and emerge all the clearer and with greater force in the absence of a descriptive milieu.
An effective medium-specific definition of narrative would also clarify the relationship between musical narrative and what Fred E. Maus refers to as drama. Maus suggests that music is like drama without characters, a drama that presents a series of actions performed at the same time as the audience s perception of the action (thus removing the necessary mediation of a narrator) and that forms a plot that holds the actions together in a unified structure (1988: 71). Using our sibling model and the example in the last paragraph, we can conclude that literature, drama, and music share a potential for meaningfully ordering events in time, but differ with respect to their degree of referential specificity. In the case of literature, characters, settings, and actions are described , but the reader fills in many imaginative details. In a dramatic medium, however, characters, settings, and actions are enacted , and thus are more precisely determined. On the other hand, literature often (but not always) reveals the internal motives and desires of the characters, while these properties must be inferred through the words and actions of the characters. In a sense, then, motive in literature is potentially more precisely determined than in drama. In music, actions are displayed, but character, setting, and motive either are indeterminate or must be supplemented in some manner. These distinctions, as I will observe in the last section of this chapter, are not as rigid as might be supposed; nevertheless, they suggest that artistic media can inflect a more general notion of narrative in different ways. To equate musical narrative with drama, however, is as misleading as equating musical narrative with literary narrative, since it covers over the very real distinctions between them.
I will define the concepts of narrative and musical narrative more thoroughly in chapters 3 and 4 . First, however, it would be useful to see what can be observed about musical narrative, or put forward in its favor, in light of a series of historical and analytical examples.
One factor in support of musical narrative is the prevalence and historical appeal of describing music via an analogy with literature, drama, or language understood more broadly. Although these analogies often do not often invoke what we might think of as narrative elements, the degree of kinship between music, literature, and drama itself lends support to the possibility of narrative organization in the former. Numerous examples can be given:
1. The pervasive human practice of combining music and language through liturgy, ritual, song, and dramatic display. Music so employed can serve to heighten the expressive effect of the text, cue important moments, add parallel or contrasting layers of signification, reveal hidden connections, suggest specific cultural valences (ceremonials, dances, intimate expression), elicit the support or communal solidarity of the listener, evoke emotion through conventional codes, enhance the prestige of a particular social group, and/or effect a meditative, reflective, ritualistic, or social-cohesive frame of mind.
2. The appropriation of rhetoric from the sixteenth century onward as a musical meta-language to render it more amenable to pedagogical systematization. Significant parallels emerging from this tradition include: (a) the correlation of musical formal organization with the stages of an oration, later expanded to include analogies with dramatic sequences; (b) the awareness of music as a persuasive art, leading to the notion that music is capable of expressing objective emotional states, either globally (as exemplified in the Affektenlehre ) or locally (as in the interplay of Classical topoi ); (c) the correlation of the respective compositional processes for music and oratory (including Heinichen s application of Aristotle s loci topici to the process of selecting musical settings appropriate to texts and Koch s plan-execution-elaboration schema ); (d) the use of Figuren as a means of heightening the expressive effect of a passage (this practice highlights the conventional derivation of music-expressive categories, which at various times invoked polyphonic effects, harmonic innovations of the seconda prattica , and Classical topoi ); (e) the relationship between the acts of narration and performance.
3. The eighteenth-century coordination of phrase-structural organization in music with grammatical features: punctuation cadence, the subject-predicate relationship the antecedent-consequent relationship, syntactic hierarchy formal hierarchy, complex sentences phrases with extensions and expansions. It is likely that such equivalences led nineteenth-century theorists to consider semantic as well as syntactic connections between music and literature.
4. Momigny s expressive analyses in the Cours complet d harmonie et de composition (1805), which elaborate the character of the work through the superimposition of isomorphic verbal texts onto principal melodic material and the subsequent tracking of the development of these melodic/dramatic entities. (See below for a discussion of the narrative implications of a Momigny fugue analysis.)
5. Wagner s explicit harnessing of the principles of foreshadowing and recall, supported in part by a web of leitmotivic associations.
6. The development of various theories of thematic and motivic transformation that invoked a semantic analogy with the development of character in literature and drama. This analogy is closely connected with the emergence of an organicist metaphor as a meta-analytical principle. Here we might include the diverse contributions of Reicha, A. B. Marx, Schenker, and Schoenberg, among others.
7. Essays written for the nonprofessional concertgoer (by Berlioz, Kretzchmar, Tovey, etc.) that extrapolate pictorial scenes and dramatic, often heroic, plots from the thematic, harmonic, and formal activity of a piece.
Even confining ourselves to relatively distant historical examples, one is struck by the startling ubiquity of literary and dramatic elements in descriptions of musical phenomena. Indeed, as long as music analysis has existed, it has called upon its sister discipline to provide a map for exploring uncharted terrain. The appeal of the literary domain for theorists, critics, and listeners of music does not, of course, constitute a proof of the existence of musical narrative. It does, however, suggest that models of musical meaning have typically been enriched by invoking languages for which semantic principles, even if equally provisional, are at least closer to hand. The popularity of this approach, however, calls for a cautionary recognition of the ways in which music is not like literature or drama. Narrative, if it exists, will occupy a conceptual space between, on the one hand, an untenable equivalence between the three media and, on the other, an equally improbable disjunction between them.
As indicated above, the limits of musical narrative with respect to Western art music are disputed, particularly with respect to non-programmatic instrumental music. One way to begin an investigation of this disputed territory is to consider existing analyses concerned with the temporal unfolding of semantic content in instrumental music. In the process, we can look for common or interesting features that might contribute to a general definition of our subject matter, to avoid having to reinvent the wheel. Three analyses, chosen for their methodological variety and degree of analytical interest, comprise one relatively early (Momigny) and two relatively recent (Tarasti, McClary) examples that will serve as the basis for our meta-analytical interrogation of narrative properties in music.
J r me-Joseph de Momigny s Analysis of Handel s Fugue from Harpsichord Suite No. 6 in F Minor
This analysis (Bent 1994a: 29-36), first published in 1805, represents an important historical attempt to supplement a descriptive structural analysis of an instrumental work with an interpretive account of that work s expressive features. As such, it provides an appropriate starting point for our investigation of musical narrative properties. Momigny, a Belgian-born music publisher and organist working in Paris, included this analysis in a more comprehensive pedagogical text, the Cours complet d harmonie et de composition (1805), unsuccessfully submitted to the French National Institute for its official sanction.
Along with the approach itself, the analysis is unusual in several important respects. Its description of a fugue in dramatic terms is striking, given its status as embodying an academic or cerebral idiom. Also idiosyncratic is its rather sharp division into two unconnected parts. The first is an account, in the tradition of Marpurg, of the fugue s structural features-its subjects, their arrangement in the various expositions, and the various alterations and contexts in which they appear. The second part is a short dramatic scene intended as an analogical reflection of the fugue-the subjects are each identified with a line of text, resulting in their (apparently) acquiring the status of theme-actors, which then interact in a manner keyed to the musical unfolding. Since this passage is relatively short, it will be given in full; the themes to which Momigny attaches text are shown in example 2.1 . 1
A father, respectable but at the same time severe, commands his daughter, who has fallen into an infatuation, to sacrifice her love. Unable to banish from her mind and heart the object of her affections, the latter says to her father: Father dear, I beg you, soften your heart (C , B A, G F , E F , D C ) [mm. 1-2]. The father, unbending, replies: No, obey you must (A, G , F , E F[ ]) [mm. 2-4]. While the father delivers his uncompromising rebuttal, the daughter turns and says to her mother: Oh plead for me, dear mother (C D, F B, E C , A) [mm. 2-3, 3-4].

Example 2.1. Handel, Suite no. 6, Allegro (fugue), measures 1-4
The delineation of the characters is reflected not only in the interaction of the voices, but also in the higher register at which the daughter s entreaties ring out at 4 and 5 [mm. 18-19, 29-30], and in the increasing frequency with which the subjects occur between 5 and x .
Following g and h , the fretful movement of the bass [mm. 32, 33-4] portrays vividly the father s wrath, which, far from abating, grows steadily more furious. This ire bursts forth for a second time at u, v , and x . At this point, the exchanges become so vehement and so rapid that the words of father, mother and daughter seem to be heard in brief snatches.
Soon everybody is so agitated that they pay no heed to each other, pursuing their own protestations while ignoring all rejoinders. The canon that begins at y [m. 57] in all three parts conveys this perfectly.
Enraged, the father thunders: I insist, don t waste your breath, obey me, I insist (D, C B, A G , F[ ], E , D C , B A, G F ) [mm. 62-4].
The two-part canon [mm. 63-6] is a portrayal of mother and daughter lamenting their inability to soften the heart of the wrathful father.
At letter dd , in despair the daughter breaks off her passionate entreaties abruptly, and declares in ringing tones [mm. 74-81] that, tear out her heart though he might, he would not succeed in banishing the image of her loved-one. She forgets herself, mixing protestations of undying love for the object of her desires with reckless reproaches to her father, full of bitterness at his heartlessness.
The latter, shocked by her audacity, stands stupefied-as the sustained bass note portrays. The mother, ever loving, tries to coax her daughter back to the filial duty and respect required of her toward her father.
This, or something like it, is the range of feeling that we believe Handel might have experienced, or the image that he might have had in mind, as he composed this fugue.
A modern analyst, accustomed to carefully correlating expressive statements with musical features, will likely be struck by the problematic nature-and, occasionally, the absence-of such correlations in Momigny s analysis. 2 We are not, for example, given much explanation for why this fugue should be realized as a domestic dispute, nor for why the subjects acquired their particular text pairings. We learn (in a previous passage) only that the opening two-measure subject segment has a noble character, but nobility seems not to be the primary feature of the pleading text (note the pianto D-C figure in measure 2) applied to it: Father dear, I beg you, soften your heart.
The precise connection between dramatic character and musical event is also somewhat confused. The initial presentation of each line of dialogue as a word-for-word setting of motivically distinct subject or counter-subject phrases creates the impression that these phrases play a role akin to theme-actors in the analysis. There are many problems with this assumption, however. The mother is never associated with a specific theme-her presence is first registered as the object of the daughter s line Oh, plead for me, dear mother -and yet she appears as an active character at other points in the analysis where the establishment of a theme-actor is not indicated by musical data. For example, the two-part canon in measures 63-66 is interpreted by Momigny as a portrayal of mother and daughter lamenting their inability to soften the heart of the wrathful father. The relevant theme, here subjected to canonic treatment, is the initial phrase segment, a segment that Momigny originally associates with the daughter, but that now appears to stand for the mother and daughter acting in concert.
Perhaps, then, the theme fragments suggest certain ideas that different characters can pick up. But this is contradicted by Momigny s interpretation of the other canonic passage involving the opening phrase segment-which begins in measure 57-in which all three characters are imagined to be arguing: if the father is participating in this argument, he would surely not appropriate the idea of pleading from the daughter s text line. It appears that Momigny intends this passage merely to suggest the babel of many voices, irrespective of any particular association.
Furthermore, the characters seem occasionally to correspond to particular registers rather than to subject phrases, as when the father s wrath is cued to a rapid bass-note pattern in measures 32-34 and 51-53. This pattern does not clearly derive from one of the text-linked melodies; instead, it suggests that the father is represented by the bass register. A further example begins in measure 74, when the bass pedal is said to stand for the father s stupefied silence. Here it is probable that Momigny wished to provide a persuasive expressive correlate for the pedal point, using the equation musical stasis = silence. Given all these difficulties, there is no reasonable way to globally link specific dramatic characters or ideas to specific musical qualities or events.
Let us assume that Momigny was not simply being careless and that he did not intend to propose a one-to-one relationship between theme and character (or idea). If this assumption is true, then what function does the dramatic scene serve? Although it superficially engages issues of domestic politics and the role of women in society, Momigny, in contrast to McClary in the Bach analysis discussed below, is obviously not offering up a social critique. Despite Momigny s use of terms such as uncompromising, wrathful, and enraged, the father is not rhetorically portrayed as a repressive villain. Indeed, it is unclear what position, if any, Momigny would take on the social issues he portrays. We are not told how the argument resolves itself, so we are not given the opportunity to feel either pity at the plight of an injured woman or anger at the disobedience and willfulness of the young. Only Momigny s choice of conventional or stock characters-the stern father, the pleading child, the placating mother-creates a strong sense of rhetorical impact.
To find an answer to the above problems, we might begin with the few hints provided in the analysis. For Momigny, the primary functions of music are pleasing the listener and engaging their interest (Bent 1994a: 35). Further, the dramatic scene is designed to interpret the expressive content of the subject (35), and represents what Momigny believes to be the range of feeling, the image that [Handel] might have had in mind, as he composed the fugue (36). These introductory and concluding remarks are also a bit misleading: an appeal to the composer s intentions is not necessarily equivalent to an appeal to the listener s interest. However, the constituent elements involved in communicating musical expressive content did not become a significant object of attention until the twentieth century. Early-nineteenth-century theorists tended to assume a transmissional transparency : if expressive content was conveyed, it would have arisen from the intentions of the composers. 3 This is not to imply that Momigny sees no distinction between the dramatic scene and the essential meaning of the fugue. It is clearly an imaginative exercise, intended to display in as vivid a manner as possible those musical events that contribute to engaging the listener s interest. Note the provisional character of Momigny s assertion: This, or something like it , is the range of feeling that we believe Handel might have experienced, or the image that he might have had in mind, as he composed this fugue.
Thus, although this analysis lacks a consistent correlation between musical event and dramatic character (that is, its agential level is ill-defined), and although it is relatively unformed with respect to an explicit social message (although we are free to impose one), its primary narrative quality lies in its attention to the expressive content derived from temporal process, from action, and from actantial features (the acquisition of narrative roles or functions). Such a strategy is congruent with Maus s understanding of narrative, in which listeners can hear actions in music by understanding musical events in relation to imagined intentions and in which musical actions have general qualities in common with other actions, as well as having specifically musical descriptions (Maus 2005: 468). Thus, Momigny s trio of characters is not introduced primarily for their characterological features or for the overall plot line, but for those semantic features that their interactions share in common with the musical unfolding.
Even were the relationships between melodic unit and character stable in this analysis, little of the character of the musical lines themselves is reflected in the characters described in the scene. The initial phrase, associated with the daughter s pleading, does not, for example, sound particularly pleading. It does have a rhythmically regular, stepwise contour, which perhaps suggests charm or artful winsomeness, but one must stretch interpretation a bit to reach this conclusion. Likewise, one could argue that the father s line depicts the process of submission to authority, since the initial syncopations give way to a conventional melodic cadential figure (dominant embellished with a 2-3 suspension). However, I believe that such interpretations say more about the malleable interpretive proclivity of musical events or the imagination of the interpreter than about the actorial coherence of these melodies. The most, I think, that can be said of the melodies is that they are semantic placeholders, employed to embody the dramatic actions in a more familiar form.
The majority of correlations in this piece simply invoke the notion of dialogue ; the alternation and interplay between fugal voices is shown to resemble the back-and-forth of the dramatic characters as they argue amongst themselves. 4 Thus, Subject 2 is a dissenting voice raised against theme 1 (Bent 1994a: 31) and represents the father s uncompromising rebuttal (35). Later, the canonic repetition is said to resemble the agitation of the three individuals, who pay no heed to each other. The changes in the character of the dialogue are cued by changes in the expressive characteristics of the musical events described. So, the canonic passage that begins at measure 57-with its closely spaced entries, full texture, and dramatic octave leaps-is described as agitated, while the relatively more subdued two-voice canon in measures 63-66 is lamenting. The rapid-fire succession of motives in measures 51-53 becomes a vehement exchange, and, as mentioned above, the active-upper-voice/bass-pedal-point combination in measures 74-81 calls forth the image of a father stupefied at his daughter s audacity. All of these dramatic elements derive from their effectiveness at describing variously inflected modes of dialogue.
Dramatic depictions of instances of musical intensification or changes of momentum also figure in the analysis. Momigny prefigures this issue in the structural analysis, when the immediate repetition of Subject 1 at a higher pitch level in measures 35-39 is said to bring an intensification which conveys a sense of gathering momentum in the voice making the repetition (33). Further examples in the expressive analysis itself include the immediately preceding passage (measures 32-34), in which the rapid bass motion is associated with the increasing ire of the father, and the arrival of the pedal point in measure 74, where the texture change is linked with the daughter break[ing] off her passionate entreaties (36).
If Momigny s analysis is a narrative one, it acquires this label through its attention to specifically temporal and actantial features of the piece-to music s dialogical and intensifying properties. Further, in spite of the lack of agential clarity, it would be misleading to suggest that the articulation of dramatic character plays no part at all in the analytical project. Taken by itself, this example might suggest that narrative analyses call for a relatively high degree of semantic specificity-a fleshed-out story, to be precise. This conclusion, among others, cannot be verified, however, until we have looked at other ostensibly narrative analyses.
Eero Tarasti s Analysis of Chopin, Ballade in G Minor (1994: 154-80)
The contributions of Eero Tarasti to our current understanding of musical narrative have, to some degree, gone unrecognized in the United States, yet his 1994 book A Theory of Musical Semiotics contains one of the most insightful and thorough explorations of the subject to date. The book is an attempt to apply the semiotician A. J. Greimas s theory of narrative grammar to the language of music. This theory combines a deductive model of the semiotic structure of meaning-producing systems with an inductive organization of its content into a dynamic narrative pattern (Monelle 1992: 233); as such, it classifies and coordinates the kinds of insights which Momigny described more informally. Tarasti s analyses, through their careful disentangling of different expressive elements, strikingly reinforce my position that musical and literary narrative exhibit important distinctions in their manifestations. A Tarastian narrative analysis is not a literary commentary-the story this composition might tell-but reflects upon the semio-musical structures that make possible its concrete psychological contents (Tarasti 1994: 138). An examination of his approach will reveal certain aspects of musical narrative theory that have not, I believe, been sufficiently emphasized.
Tarasti s analytical method is discussed in some detail in chapter 4 , to which I refer the reader for more details. In general, however, his approach can be characterized as parametric in orientation. He segments the musical work into isotopies : passages rendered distinct by the employment of redundant semantic categories. Within each isotopy, he highlights the dynamic role played by certain discoursive [ sic ] categories , which roughly correspond either to traditional musical parameters of register, rhythm, harmony, and so on, or to other temporal organizational processes like memory, expectation, or the transformation of theme- and motive-actors. Not all of these processes will be significant in any given piece, but they are all potential fields for discursive activity. These dynamic processes-occurring within each isotopy-mutually affect and influence one another by means of modalities , reflecting fluctuations in activity level, directed movement toward goals, the amount of information available to the listener, the discharging of formal or discursive obligations, and the degree of performative display. Having identified these processes, the analyst can then construct a map of musical-semantic relationships that can be variously emphasized to enable narrative interpretations.
Tarasti illustrates this approach most rigorously in his analysis of Chopin s Ballade in G minor (154-80; see chapter 4 below for a summary of this analysis). In this analysis, he elucidates a tragic narrative plot in which certain theme-actors strive for and ultimately fail to achieve their musical goals. In light of my current discussion, Tarasti s approach leads to several important insights about musical narrative.
1. Musical narrative is fundamentally dependent on the listener s, analyst s, and/or performer s interpretation . If diverse and problematic situations can be identified on different musical levels, even when more basic semantic data are agreed upon, then narrative requires the interpretive lens of the observer to render it meaningful. And this step cannot be omitted in a truly narrative analysis: Tarasti s inductive Ballade analysis, as he himself observes, has only established the building-blocks from which a narrative interpretation might be constructed-deductively, as it were. 5 Of course, a successful-that is to say, persuasive-interpretation will be significantly guided by the phenomenal data, but a great deal of leeway remains with respect to the shaping and contextualizing of these data. One essential feature of a narrative analysis, then, is that sense be made of the temporal succession, even if that sense involves the articulation of senselessness. 6
2. The traditional narrative paradigm initial situation-disruption-resolution does not in every case refer to a reestablishment of the initial situation or a synthesis of opposites . In both analyses discussed above, we find narrative interpretations (including Tarasti s short narrative summary) that contradict the common misconception that narrative plots always reestablish a kind of order. Tarasti s interpretation is essentially tragic (and not merely because it contains tragic topoi ), in that the theme-actors whose shifting fortunes we follow both fail to achieve their objectives at the end of the piece. By contrast, Momigny s family quarrel ends ambiguously, with no particularly satisfying resolution. 7 Neither narrative has what we might call a happy ending.
To expect that every narrative will do this, however, is, I think, a misreading of Todorov s equilibrium-disequilibrium-equilibrium scheme. Just as many ABA forms do not feature an exact return of initial material, so too do the first and third phases of Todorov s scheme (potentially) represent different equilibria. Further, a particular equilibrium need not possess a positive valence with respect to an observer-it may, in fact, be a completely undesirable equilibrium. As will be discussed in chapters 3 and 4 , I believe that it is better to understand narrative as both displaying a particular set of hierarchical relations subjected to crisis and tracking the consequences of this crisis . Recognizing that narratives may resolve undesirably or not resolve at all is as crucial to musical narrative theory as it is to literary narrative theory.
3. Narrative involves the coordination of multiple elements involving different mechanisms of meaning and different levels of focus or temporal scope . Tarasti s analyses illustrate very clearly that narrative has a special status among aspects of musical meaning. In particular, narrative draws upon and gives perspective to many other more localized meaning-generating phenomena. A partial list of such phenomena that contribute to narrative meaning might include: affect, theme- and motive-actors, topic, syntactic units (grammatical and formal significations), motivic interrelationships, gestural processes, expectation, relationship to an ideal schema such as sonata form, borrowing, program, text, descriptive titles, musical symbol, cultural and political relationships, generic and stylistic conventions, and strategically emergent meaning. Given that it is capable of assimilating and directing the interpretation of all these elements, narrative may represent the object of study most capable of evaluating musical meaning in its fullest sense. Only large-scale topical configurations such as Hatten s expressive genre (1994: 67-90) or detailed programmatic content have a similar reach. Even here, expressive genres do not apply with equal force when topical features are minimal, and programmatic music has generally been considered a subdivision of narrative theory.
Furthermore, these meaning-generating elements interact on different levels. For example, local stylistic and generic features help to establish the semantic characteristics of theme-actors, while gestural features (intensification, dissolution, etc.) indicate their change of status over time. Meanwhile, global topical features provide a contextual frame in which to understand the relationships described, and isomorphisms between musical and social relationships allow for the interpretation to connect with a large cultural field of values. 8 Through the coordination of levels and processes like these, narrative provides a unique perspective from which to view the ways in which various elements of meaning interact.
4. Actorial elements are not essential to music narrative, but are one of those devices that can be successfully translated from literary narrative . Actoriality, one of Tarasti s discoursive categories (Tarasti 1994: 98-111), invokes the activity of musical actors-themes, motives, cells, or other musical groupings functioning as narrative subjects; 9 this issue will be discussed at more length in chapter 4 . For now, however, consider that Tarasti s analytical method-in identifying actoriality as one of several discoursive categories-allows for the articulation of a narrative design in the absence of actorial elements. For example, an analyst might track the attainment of registral, formal, harmonic, and/or temporal goals in narrative terms, even where these goals are not embodied as musical characters.
5. Along with the coordination of multiple levels and musical features, the central role of conflict in narrative analysis must be emphasized . This topic will be formalized in chapters 3 and 4 , but we can already observe that, in both analyses discussed above, the analyst has centralized the notion of conflict. For Momigny, the argument between a father and daughter over a marriage prospect provides a suitable environment to display both the dialogic character of the fugue and certain gestural features of the musical discourse, such as the gradual intensification of motivic interplay or the buildup of tension at the dominant pedal. For Tarasti, the conflict between themes as played out in the registral, tonal, and rhythmic domains mirrors a deeper conflict between thematic delineation and interconnection. In both cases, it is conflict-that which projects the discourse beyond its initial conditions-that propels the narrative forward, that gives the piece its psychological and aesthetic logic. I will examine the implications of this idea in later chapters.
Susan McClary s Analysis of Bach, Brandenburg Concerto No. 5, First Movement (1987: 13-62)
With respect to narrative method, McClary s discussion of Bach s music as social discourse (19) is a mirror image of Tarasti s musico-semiotic taxonomy of the Ballade. Tarasti would, I think, agree with McClary s position that inasmuch as every piece of music assembles and problematizes very different elements of the shared semiotic code, the interpretive process is by definition both ad hoc (it derives its strategies from the specific demands and features of the individual composition) and dialectical (it strives to account for particularities in terms of the norms they affirm or oppose) (21). However, whereas the rhetorical weight of Tarasti s analysis rests on what we might call bottom-up issues-the constitution and presentation of the structure and dynamics of discourse-with the interpretation left somewhat open-ended, McClary places her primary attention on the top-down component-on the ideological implications of the piece s dynamic musical impulse[s] (20). Both writers justify their analytical weightings on the basis of perceived disciplinary deficiencies. Tarasti attempts to counter an insufficient degree of attention to the variety of ways that music can signify, while McClary wishes to thematize the idiosyncratic features of compositions in place of the deep-structural universals that are for her the preferred objects of interest for mainstream music theory.
McClary s analysis of the Brandenburg Concerto movement was intended to deconstruct and then reconstruct the music of Bach to support an agenda of empowering the ideologically marginalized artist, thus bringing his music into line with a postmodern musical aesthetic (60-62). By identifying correlations between musical and sociocultural valences, McClary proposes a narrative interpretation of the movement that invokes an eighteenth-century struggle of dominant versus marginalized social groups. To this end, she identifies several stylistic norms and deviations applicable to the Concerto movement that show how contemporary conventional norms of social order were disrupted. According to McClary, the social agenda attributed to Bach would not have been out of character, particularly at a time when the Germanic cultural zone occupied a decentered position with respect to Italy and France (19-20).
The following stylistic norms are presented by McClary as contributing to musical signification in the first movement of Brandenburg Concerto No. 5:
1. The tonal dialectic, which creates both large-scale coherence and localized directionality of tension toward significant arrival points. The goal-directedness and regulated quality of tonality suggests a quality of predetermination and rationality that would have appealed to the propertied consumer of Enlightenment culture (21-23).
2. The concerto grosso formal procedure, by which a large ensemble-characterized by tonal and melodic stability and regular phrase groupings and representative of the dominant social hierarchy-is set against a comparatively free solo instrument or ensemble-characterized by tonal and melodic freedom, virtuosic display, dissonance, and dynamic extremes. McClary sees the resulting dialectic between these two groups as isomorphically reflecting the struggle between individual and society that had become a central issue during Bach s lifetime. Typically, in such a scheme, the composer would enact the proper social coding with respect to the solo passages [the individual] by ceding to the ritornello sections [society] the control of appropriate tonal and temporal boundaries. In so doing, individual expression and social harmony will finally be demonstrated to be compatible (24).
3. The harpsichord, in its conventional service role as a continuo instrument providing harmonic and metric support for the ensemble. McClary emphasizes the thankless quality of this task, mirroring the subordinate position of the composer-harpsichordist or conductor-harpsichordist with respect to his audience (25).
As constituted above, the stylistic conventions of the concerto grosso seem well suited for promoting one of two socially acceptable narrative schemes: (1) the appropriate submission of individual aspiration for the good of society or (2) the reconciliation of the apparently contradictory aims of the individual and society. However, McClary reads Bach as manipulating these conventions in the fifth Concerto so that the goals of the individual cannot be reconciled satisfactorily with those of the group, thereby exceeding socially acceptable limits and subjecting them to a narrative critique.
In demonstration of this assertion, McClary presents a short analysis of the first movement, which, she argues, begins like a typical concerto movement for solo flute and violin. She describes the opening ritornello as confident (note the self-assured arpeggiation of the opening), unified, slightly smug (the repetition of each note of the unison arpeggio yields a quality of complacency or selfsatisfaction), and self-contained (26). To these qualities we might add its selfcontained tonal motion, its registral arch that ascends by almost two octaves and returns gradually to its starting point, and its confidently pervasive eighth-note pulse.
McClary describes the solo instruments as somewhat sentimental due to their use of elaborate and light ornamentation and as dynamic enough to accomplish modulations to other keys (28). She then describes how the harpsichord moves from its original supporting role to a more active one, overtaking and surpassing the activity of the soloists and forcing attention upon itself. McClary notes that in the passage just prior to the cadenza-an extensive harpsichord solo-Bach composes the parts of the ensemble, flute, and violin to make it appear that their piece has been violently derailed. They drop out inconclusively, one after another, exactly in the way an orchestra would if one of its members started making up a new piece in the middle of a performance (28).
Supporting the fact that the cadenza is played by the wrong instrument (the harpsichord), the cadenza is unusually long and reaches extremes of rhythmic diminution and harmonic complexity capable of sustaining enough tension to preclude the appearance of the ritornello. When the ritornello finally does reappear to end the movement, it is allowed to do so only by the subsidence of the harpsichord, a sign that the latter had previously reversed its signifying role to become the dominant narrative agent. The subsequent restoration of social conformity thus seems insufficient and arbitrary after the violence of the harpsichord cadenza (32-40).
Most critical for a theory of narrative in this analysis is McClary s attention to the ideological and psychological foundations of narrative analyses (and of analysis in general). With respect to the former, McClary argues that certain modes of analysis are perpetuated by disciplinary groups who wish to control the power to manipulate musical discourse. Narrative principles can thus be used either to promote or to subvert this state of affairs (15-16). They can also be used unconsciously, in ignorance of their social implications.
With respect to the latter, she suggests that particularities of personal temperament are implicated in the kinds of narratives that we construct. In particular, she distinguishes between
(1) those who seek to immerse themselves in what they wish to regard as the pure order of music in order to escape what they perceive as the chaos of real life and (2) those who turn to music in order to enact or experience vicariously the simulacrum of opposition to the restrictiveness of real life (with real life represented by those abstract though socially grounded norms). The ways in which one composes, performs, listens, or interprets are heavily influenced by the need either to establish order or to resist it. (18)
Here she rhetorically contrasts theorists who bracket off social context from analysis with those who consider its contribution significant. Whether because of her significant influence on subsequent music scholarship or because the initial dichotomy overstated the size of the interpretive chasm (or both), I think this sharp distinction has become murkier in recent years. Instead, I would propose that the opposition has been relocated, reflecting instead a divide between different kinds of social contextualizing. We could rewrite McClary s assessment to distinguish between (1) those who construct narratives that attempt to ward off what they perceive as the chaos of real life and (2) those who construct narratives that oppose the restrictiveness of real life. This change, I believe, reflects recent realignments in a field of music theory more comfortable with the notion of music as discourse. This is not to say that the original formulation is no longer valid, but that at least some of its energies have been transferred into another disputed region.
The implication of this assessment is that narrative analysis involves the complicity-conscious or unconscious-of the analyst in shaping that narrative, that it is shaped as much by the analyst s psyche as it is by musico-cultural convention. Further, this complicity need not, must not, be understood as a deficiency to be factored out or corrected for; it is instead a precondition of-a justification for-narrative patterning.
Here we come to the heart of the matter, the why of narrative. Here, I believe, we bump up against the reason for the prevalence of narrative organization across diverse media and across boundaries of temperament and discipline. If narrative can display, through its coordination of abstract and conventional relationships, the patterns that show us how to preserve stability against chaos, how to subvert an oppressive hegemony, how to recognize the limits of our power, or how to find a new path out of an insoluble dilemma, then its individual and political appeal is obvious. This is, for example, the position taken by Hayden White when he recognizes the indissoluble link between historical discourse and narrative emplotment.
There does, in fact, appear to be an irreducible ideological component in every historical account of reality. That is to say, simply because history is not a science, or is at best a protoscience with specifically determinable nonscientific elements in its constitution, the very claim to have discerned some kind of formal coherence in the historical record brings with it theories of the nature of the historical world and of historical knowledge itself which have ideological implications for attempts to understand the present, however this present is defined. To put it another way, the very claim to have determined the formal coherence of that past world, implies a conception of the form that knowledge of the present world also must take, insofar as it is continuous with that past world. Commitment to a particular form of knowledge predetermines the kinds of generalizations one can make about the present world, the kinds of knowledge that one can have of it, and hence the kinds of projects one can legitimately conceive for changing that present or for maintaining it in its present form indefinitely. (White 1973: 21)
Similarly, the semiotician James Jak b Liszka, whose ideas I will later revisit, argues that narrative patterns are the basic strategies used by fantasy, by the narrative imagination, in playing out the tensions between the violence of a hierarchy that imposes order and the violence that results from its transgression (Liszka 1989: 133). For the individual, then, narrative patterns are psychological templates illustrating possible responses to conflict. For society, they represent paths to visualize and confront structures of power in constructive and/or critical ways.
Looking back on the preceding discussion, it is possible to say with more clarity what music narrative is and is not. It is a psychologically and socially meaningful articulation of hierarchical relationships and our responses to them. It involves the coordination of multiple structures of meaning at multiple levels.

  • Accueil Accueil
  • Univers Univers
  • Ebooks Ebooks
  • Livres audio Livres audio
  • Presse Presse
  • BD BD
  • Documents Documents