Multiple-Row Serialism in Three Works by Edison Denisov

by

Zachary A. Cairns

Submitted in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

Supervised by Professor Dave Headlam Department of Music Theory Eastman School of Music

University of Rochester Rochester, New York 2010

ii CURRICULUM VITAE

Zachary Cairns was born in Champaign, Illinois on July 11, 1978. He attended the Pennsylvania State University from 1996 to 2000, graduating with a Bachelor of Science in Music Education with a Performers Recognition Certificate in Percussion Performance. He returned to Penn State in 2001 and graduated with a Master of Arts in Music Theory in 2003. His masters thesis dealt with rhythmic organization of the music of Edgard Varse and was supervised by Dr. Julian Hook. After teaching at Northwestern Lehigh High School from 2003 to 2005, he came to the Eastman School of Music, University of Rochester, and began doctoral studies in Music Theory. He has pursued his research on the music of Edison Denisov under the direction of Dr. Dave Headlam. He will begin work as Assistant Professor of Music Theory at the University of Missouri-St. Louis in August, 2010.

iii ACKNOWLEDGMENTS

The completion of this dissertation would not have been possible without the generous help of many people. First, I must offer deep and sincere thanks to the members of my dissertation committee, Dave Headlam, Jeannie Guerrero, and Peter Schmelz. Working with each of them has helped me to fine-tune my ideas and my writing in very important ways. I would like to thank my advisor, Dave Headlam, for his patient, thoughtful assistance throughout the entire process. He has allowed me the opportunity to explore my own ideas, while continually drawing my attention to connections between those ideas and the work of others. I would also like to thank Jeannie Guerrero for her invaluable comments on various drafts of this dissertation, particularly in helping me to shape some of my thoughts in Chapter 3. I am especially grateful for the willingness of Eastmans theory department to support my request to have Peter Schmelz officially act as a member of the committee. Professor Schmelz has been exceptionally gracious with his time from the earliest stages of this project, sending thoughtful and thought-provoking responses to my many, many email queries on various aspects of Soviet music and music theory, as well as my efforts at translating the Russian language. Having him as a member of the committee has been a true honor, and I thank him for his willingness to serve in this capacity. I also thank Jonathan Dunsby, chair of the Department of Music Theory, who helped to make this collaboration possible. Over the last several years, I have had many fruitful conversations and interactions with a number of people that have impacted this work in both direct and indirect ways. I have been fortunate to be able to communicate with Edison Denisovs widow, Ekaterina

iv Denissova-Bruggeman, and it has been thrilling to read her considerate responses to my emails which come to her in what I can only imagine to be horribly broken French. She has been helpful in providing background information on her late husbands music and compositional approach. Professor Truman Bullard deserves special recognition for his help with my translation of the Kurbatskaya chapter found in Chapter 2. I have fond memories of our meetings during the summer of 2009, going sentence-by-sentence through that translation over lunch(es). The experience of working with him has been not only educational, but deeply rewarding on a personal level as well. I would like to thank Tuukka Ilmaki, Sebastiano Bisciglia, Justin Lundberg and Andy Flowers, the members of Dave Headlams Super Computer Club, for their help with various aspects of computer programming and for a productive exchange of ideas on twelvetone music. Though this dissertation, in its present form, shows little evidence of my endeavors in C++ programming, our discussions have been very profitable, particularly in dealing with some of the mathematical aspects of Chapter 3. My interest in the music of Edison Denisov can be traced back to the two saxophonists with whom I shared an apartment during my masters studies at Penn State. Erik Holmgren and Matt Sisia have always been wonderful friends, and they introduced me to Denisovs Sonata for Alto Saxophone and Piano. I never would have thought that a sentence as simple as, Zac, I think youll like this piece could be a life-changing thing, but in this case, it was. It would be neglectful of me to omit the names of people who have helped me in acquiring scores and recordings of Denisovs music. Much of his music is, unfortunately, not readily available on the shelves of your local public or University library. I would

v especially like to thank Sibley Music Librarys Special Collections Librarian David Peter Coppen for helping me to locate Russian music-theoretical sources, and the Interlibrary Loan staff at Sibley for their tireless efforts in processing my many, many requests to borrow scores, books, and recordings from other libraries. I also offer my thanks to Elizabeth Mulhall in the Promotion department at Schirmer, who helped me to obtain perusal scores and recordings for a large number of Denisovs works which are otherwise only available through rental. Finally, I thank my entire family for their love and encouragement. My parents, Scott and Ginny, and my wifes parents, Barry and Trish, have given me constant support not only through the dissertation writing, but through the entire degree program. And, above all, I thank my wife Whitney for her immeasurable support in this and all in my endeavors. I thank her from the bottom of my heart not only for her tireless efforts in proofreading and correcting problems in the text, notes and bibliography, but more generally for keeping me motivated and on a forward-moving path. She is the reason that I have finally finished this dissertation.

vi ABSTRACT

The music of Russian composer Edison Denisov (1929-1996) has not been widely studied by music theorists in the United States. He was one of the student composers who came of age in the Soviet Union during the period following Stalins death. During this time, young composers began experimenting with some of the modernist approaches to composition found in Europe and America. In this dissertation, I have chosen to focus on one specific aspect of Denisovs approach to serial composition. My goals are twofold. First, my fundamental objective will be to give analytical attention to what I strongly believe is an important, though neglected, repertoire. Second, the dissertation explores technical issues involving multiple-row serialism in general, and demonstrates the utility of my own analytical approach developed in the study of this music. The dissertation is divided into two large parts. The first, comprising Chapters 1-3 serves to set up a music-theoretic backdrop against which Denisovs serial music might be profitably examined. Chapter 1 gives the reader a sense of the cultural, political and societal situation in which Denisov and his student colleagues found themselves. This chapter also provides some general comments about Denisovs highly varied compositional style over the course of his career. Chapter 2 presents an annotated translation of a chapter from Svetlana Kurbatskayas book, Seriynaya muzka: vopros istorii, teorii, estetiki [Serial Music: Questions of History, Theory, Aesthetics]. Written in 1996, this book has the interesting distinction of being the first Russian-language book dedicated entirely to the topic of serial music. This chapter also uses many of Kurbatskayas concepts in an analysis of Andrey Volkonskys piece Musica Stricta (1956-1957). Generally accepted as the first piece of serial music composed in the Soviet Union, this piece occupies an important place in the history of Soviet music, and is

vii one that Denisov knew well. In Chapter 3, I move to a discussion of some of the general issues of multiple-row serialism. Within this chapter, I introduce a new analytical device, the interval-class succession graph, which can illuminate Denisovs music by offering a way of conceptualizing the structure of a twelve-tone row which heretofore has been unexplored. Chapters 4 through 7 make up the second part of the dissertation: the analytical discussion of three specific pieces involving multiple twelve-tone rows. The pieces are the following: the first of Five Etudes for Solo Bassoon (1983), and the finales of Sonata for Alto Saxophone and Piano (1970) and Octet for Winds (1991).

viii NOTE ON TRANSLITERATION AND TRANSLATION

In this dissertation, I use Gerald Abrahams system of transliteration outlined in The New Grove Dictionary of Music and Musicians (1980), as amended by Richard Taruskin and described at the beginning of his book Stravinsky and the Russian Tradition: A Biography of the Works through Mavra (1996). Exceptions to Abrahams system are discussed in some detail in Taruskins book. The transliteration of Russian names has been carried out literally throughout the dissertation, with only a few exceptions. The common rendering of Schnittke is used in place of the more literal Shnitke. I use the Czech spelling of the name Ctirad Kohoutek, rather than the Russian Kogoutek. Finally, for the sake of consistency, I have rendered the name of the famous Russian music theorist as Yuri Kholopov throughout, despite the common Germanic rendering (Iurii Cholopow). Unless otherwise indicated, all translations in this dissertation are my own, with two qualifications: 1) I have been fortunate to have had significant editorial assistance with the Russian translations from Dr. Peter Schmelz and Dr. Truman Bullard; 2) in cases where a source exists both in the original Russian and an English translation (most importantly, Kholopov and Tsenovas monograph on Denisov), I have worked with the two sources sideby-side, departing from the published English translation where necessary. I have also approached Denisovs analysis, published in Russian and Italian in the same journal, of Weberns Piano Variations Op. 27 in a similar manner.

ix TABLE OF CONTENTS

Curriculum Vitae . ii Acknowledgments iii Abstract .. vi Notes on Transliteration and Translation ...viii Table of Contents ix List of Figures and Examples .. xi

Chapter 1 Introduction 1 Introduction . 1 Denisovs Stylistic Evolution, Piano Variations5 Denisovs Individual Style ... 13 Chapter 2 The Russian Understanding of Serialism 23 Annotated Translation: Svetlana Kurbatskaya, Serial Music: Questions of History, Theory, Aesthetics (Ch. 2)23 Analysis: Andrey Volkonskys Musica Stricta (1956-1957).. 43 Movement 1: Andantino44 Movement 2: Allegretto. 48 Movement 3: Lento Rubato54 Movement 4: Allegro Marcato 59 Chapter 3 On Multiple-Row Serialism 65 General Issues of Multiple-Row Serialism65 Issues of Structure .. 68 Issues of Function... 70 Issues of Unity 76 INTs, INT-Cs, and BIPs..78

x The Adjacency Matrix and Interval-Class Succession Graphs.. 82 Comparison and Categorization of ICSGs.. 85 Chapter 4 Analysis: Five Etudes for Solo Bassoon (No. 1). 93 General Observations 93 Issues of Structure .. 99 Issues of Function..105 Issues of Unity . 112 Chapter 5 Analysis: Sonata for Alto Saxophone and Piano (Mvt. 3). 116 Existing Studies of the Sonata 119 Form 125 Techniques of Serial Modulation in Denisovs Sonata 128 The ICSG in Denisovs Sonata . 132 Chapter 6 Analysis: Octet for Winds (Mvt. 2) 143 General Observations ... 143 Forms ... 145 Multiple Rows in the Octet .153 Chapter 7 Conclusions and Extensions 161 Further Research163

Bibliography . 169 Appendix I: Kurbatskaya Translation175 Appendix II: Graph Theory Terminology .187

xi LIST OF FIGURES AND EXAMPLES Example 1.1A, B Webern, Op. 27 and Denisov, Variations Rows and their INTs Example 1.1A Example 1.1B Example 1.2 Example 1.2A Example 1.2B Example 1.2C Example 1.3 Example 1.4 Example 1.5 Example 1.6 Example 1.7 Example 1.7A Example 1.7B Example 1.8 Example 2.1 Example 2.2 Example 2.3 Example 2.4 Example 2.5 Example 2.6 Example 2.7 Example 2.8 Webern, Op. 27 Denisov, Variations Denisov, Piano Variations Excerpts from Variations 1, 3, 6 Variation 1: mm. 10-13 Variation 3: mm. 25-29 Variation 6: mm. 73-78 Webern, Op. 27 mm. 1-7 Octet for Winds (Mvt. 1) mm. 1-5 Sonata for Alto Saxophone and Piano (Mvt. 1) m. 77 The Sun of the Incas (Mvt. 3) mm. 9-13 The Sun of the Incas (Mvt. 5) mm. 19-21 (Pianos and Percussion omitted) mm. 34-38 (Pianos and Percussion omitted) Sonata for Alto Saxophone and Piano (Mvt. 1) mm. 17-19 Webern, String Quartet (1905), mm. 1-13 Excerpt from Webern op. 10, no. 3 (from Kurbatskaya) Peter Schmelzs Soviet Serial Bulls-Eye Diagram Musica Stricta (Mvt. 1) Form Musica Stricta (Mvt. 1) mm. 1-9 Musica Stricta (Mvt. 1) mm. 21-22 Musica Stricta (Mvt. 1) mm. 19-20 Musica Stricta (Mvt. 1) m. 30 9 9 9 10 10 10 10 12 15 17 18 19 19 19 20 31 36 41 45 45 46 47 47

xii Example 2.9 Example 2.10 Example 2.11 Example 2.12 Example 2.13 Example 2.14 Example 2.15 Example 2.16 Example 2.17 Example 2.18 Example 2.19 Example 2.20 Example 2.21 Example 2.22 Example 2.23 Example 2.24 Example 2.25 Example 3.1 Example 3.2 Example 3.3 Example 3.4 Example 3.5 Example 3.6 Example 3.7 Musica Stricta (Mvt. 2) Form Musica Stricta (Mvt. 2) Rows A-D Musica Stricta (Mvt. 2) ic5 segments in Rows B and D Musica Stricta (Mvt. 2) mm. 3-4 Forms of Row A in two voices Forms of Rows B, C and D in three voices Musica Stricta (Mvt. 3) Rows E-F Musica Stricta (Mvt. 3) mm. 2, 6 Musica Stricta (Mvt. 3) mm. 7, 9 Musica Stricta (Mvt. 3) mm. 8, 10 Musica Stricta (Mvt. 3) m. 11 Musica Stricta (Mvt. 4) Rows G-H Musica Stricta (Mvt. 4) Form Forms of Rows G and H in mm. 1-13 Triple counterpoint in mm. 28-54 Musica Stricta (Mvt. 4) mm. 19-23 Musica Stricta (Mvt. 4) mm. 67-74 Concerto for Guitar and Orchestra Form Concerto for Guitar and Orchestra Rows Concerto for Guitar and Orchestra, Guitar, mm. 349-352 Concerto for Guitar and Orchestra, mm. 130-132 Sonata for Saxophone, Row A Sonata for Saxophone, Row B Berg, Lyric Suite (Primary Row) 48 49 50 51 52 53 54 56 57 57 58 59 60 60 61 61 62 73 74 74 75 79 79 81

xiii Example 3.8 Example 3.9 Example 3.10 Example 3.11 Example 3.12 Example 3.13 Example 3.14 Example 4.1 Example 4.2 Figure 4.1 Example 4.3 Example 4.4 Example 4.5 Example 4.6 Example 4.7 Example 4.8 Figure 4.2 Example 4.9 Example 4.10 Example 4.11 Example 4.12 Example 4.13A Example 4.13B Example 4.14 Adjacency matrix for Sonata for Saxophone, Row A Sonata for Saxophone, Row A, successions of ic2 Sonata for Alto Saxophone, Row A, ICSG Three Webern rows and their ICSGs Webern, Kinderstck (1924) Row and ICSG Different representations of Kinderstck ICSG Webern, Drei Volkstexte (Op. 17) Rows and ICSGs Five Etudes for Bassoon (No. 1) row chart Five Etudes for Bassoon (No. 1) Rows A-J and their ICSGs Angle measurements for row pairs in Five Etudes for Bassoon (No. 1) INTs of Rows C and J (prime forms) Abstract inclusion relationship between ICSGs of Rows H and Rows A, C, E, and J Five Etudes for Bassoon (No. 1) mm. 1-3 (Row A) Five Etudes for Bassoon (No. 1) mm. 5-6 (Row A) Five Etudes for Bassoon (No. 1) mm. 3-5 (Row B) Five Etudes for Bassoon (No. 1) mm. 6-9 (Row C) 82 83 85 86 87 88 89 95 96 101 102 103 103 104 104 105

Invariances in row-form successions in Five Etudes for Bassoon (No. 1) 107 Symmetry of Invariant Tetrachords Location of Tetrachords {056e}{349t}{1278} in mm. 6-18 Five Etudes for Bassoon (No. 1) mm. 18-19 Five Etudes for Bassoon (No. 1) mm. 17-20 (CP6, X, GP7) Five Etudes for Bassoon (No. 1) Small Sets of GP7 Five Etudes for Bassoon (No. 1) mm. 20-24 Five Etudes for Bassoon (No. 1) mm. 36-50 108 109 110 111 112 112 113

xiv Example 5.1 Example 5.2 Example 5.3 Example 5.4 Example 5.5 Example 5.6 Example 5.7 Figure 5.1 Example 5.8 Example 5.9 Example 5.10 Example 5.11 Example 5.12 Example 5.13 Example 5.14 Figure 5.2 Example 6.1 Example 6.2 Example 6.3 Example 6.4 Example 6.5 Example 6.6 Example 6.7 Sonata for Alto Saxophone and Piano (mvt. 3), form Sonata for Alto Saxophone and Piano (mvt. 3), Row A RICH and TR-Chain of Row A Sonata for Alto Saxophone and Piano (mvt. 3), saxophone, mm. 3-7 Sonata for Alto Saxophone and Piano (mvt. 3), piano right hand, mm. 7-13 Sonata for Alto Saxophone and Piano (mvt. 3), saxophone, mm. 35-37 Sonata for Alto Saxophone and Piano (mvt. 3), piano right hand, mm. 25-26 Four Techniques of Serial Modulation Sonata for Alto Saxophone and Piano (mvt. 3), Rows A and B Sonata for Alto Saxophone and Piano (mvt. 3), m. 1 Sonata for Alto Saxophone and Piano (mvt. 3), piano part, mm. 17-19 Sonata for Alto Saxophone and Piano (mvt. 3), mm. 34-42 Sonata for Alto Saxophone and Piano (mvt. 3), mm. 53-68 (triplet figures only) Sonata for Alto Saxophone and Piano (mvt. 3), mm. 76-78 Row C created by left hand in mm. 76-78 Derivation of Row C Octet for Winds (Mvt. 2), mm. 1-3, horn 2 shooting Octet for Winds (Mvt. 2), m. 6, oboe 1 pointillistic bursts Octet for Winds (Mvt. 2), mm. 20-21, bassoon 1 smooth threads Octet for Winds (Mvt. 2) Form Octet for Winds (Mvt. 2) Rows A and B, and ICSGs Octet for Winds (Mvt. 2) mm. 17-18, Canon #5 Octet for Winds (Mvt. 2) mm. 68-70 125 128 129 130 130 131 131 132 133 134 135 136 139 140 141 141 144 144 145 146 149 150 152

xv Example 6.8 Example 6.9 Example 6.10 Example 6.11 Example 6.12 Example 6.13 Example 7.1 Example 7.2 Example 7.3 Example 7.4 Octet for Winds (Mvt. 2) Row A and B, ICSGs re-drawn Octet for Winds, Row A & Saxophone Sonata, Row A ICSGs as non-directed graphs Octet for Winds, Row B & Saxophone Sonata, Row B ICSGs Concerto for Guitar Rows A and B, and ICSGs Octet for Winds (Mvt. 2) mm. 19-22 Octet for Winds (Mvt. 2) mm. 61-64 Octet for Winds (Mvt. 2) mm. 1-5, horn 2 Octet for Winds (Mvt. 2) mm. 1-5, horn 2, Beat division succession graphs Graph of functional chord successions in major mode pieces Campions Regle de loctave (1716) presented as a succession graph 153 154 155 156 158 159 163 164 165 166

1 CHAPTER 1 INTRODUCTION During the rule of Stalin (1929-1953), composers in the Soviet Union lived and worked within a world of severe restrictions. 1 The dogma of socialist realism prohibited music, literature and the arts from revealing any sort of formalist tendencies. 2 Music was expected to play a societal role, one of glorifying the state and the working classes and promoting the spirit of socialism. 3 Without offering much in the way of specific criteria, a composition was only deemed acceptable by the Soviet authorities if it was accessible to a general audience, tuneful, optimistic, folk-influenced and fell into a traditional style. 4 Composers during this time had virtually no legal access to scores, recordings, or books related to avant-garde music of Europe and the Americas; influences of the modern West had no place in the world of Soviet music. Ideologically, the composition of serial music in Stalins Soviet Union, then, was doubly problematic: first, it was derided as formalist by the state, and second, the method itself demonstrated an unacceptable Western influence. Dmitri Shostakovich is probably the best known of the composers who suffered under the communist regime; the various facets of his own personal struggles with the

The situation faced by Soviet composers during this time period has been addressed in a number of sources. See Hakobian 1998, Schmelz 2002, 2005 and 2009, Schwarz 1972, as well as the various essays from Taruskin 1997. 2 A precise definition of formalist was never offered to those composers whose works were publicly denounced as such. In general, though, the word seemed to imply that a compositions musical language was abstract and complex in some way. See Fay 2000, 87-8 and Taruskin 1997, 517-28. 3 Tertz 1960, 24-26. Tertz takes the phrase spirit of socialism from the First All-Union Congress of Soviet Writers. He goes on to describe that socialist realism demands that an artists work have a singular Purpose with a capital P. On p. 26 he writes, A poet not only writes poems but helps, in his own way, to build Communism; so, too, do sculptors, musicians, agronomists, engineers, laborers, policemen, and lawyers, as well as theaters, machines, newspapers, and guns. 4 Fay 2000, 89.
1

2 Soviet system have been widely discussed. 5 We can use the following 1959 statement by Shostakovich to summarize the Partys stance on serialism, as, in the late part of his career, circumstances increasingly forced him to become a mouthpiece of official Soviet aesthetic policy. 6 I am firmly convinced that in music, as in every other human endeavor, it is always necessary to seek new paths. But it seems to me that those who see these new paths in dodecaphony are seriously deluding themselves. The narrow dogmatism of this artificially invented system rigidly fetters the creative imagination of composers and deprives them of individuality. It is no accident that in the entire legacy of Schoenbergs dodecaphonic system there is not a single work that has gained wide acceptance. Dodecaphony not only has no future, it doesnt even have a present. It is just a fad that is already passing. 7 In official Soviet publications such as Sovetskaya muska, the journal in which this statement was printed, such polemical statements were common, but it is difficult to know whether or not they were representative of the authors true feelings. 8 However, for our present purposes, it is not particularly relevant whether this statement is truly how Shostakovich, in particular, felt about such music, but rather that this was the general opinion of the Soviet state on serialism. Following Stalins death in 1953, the situation became somewhat less severe. During the so-called Khrushchev Thaw, access to Western musical sources was allowed to varying extents. While completely open access was still an impossibility, the composers who were studying at the Moscow Conservatory during this time were aware of the trends that had been developed thirty years prior by Arnold Schoenberg and his
Shostakovichs situation is well documented from a variety of different perspectives. Among the most important are Fay 2000, Ho and Feofanov 1998, the three essays dealing with the composer in Taruskin 1997 and the much-debated publication of Shostakovichs so-called memoirs, Volkov 1979. 6 Fay 2000, 214. 7 Quoted in Fay 2000, 214. While a discussion of the context of this quote and the circumstances regarding Shostakovichs turbulent relationship with the Communist Party is beyond the purview of the present dissertation, some of the relevant details can be found in Fay 2000, 173-175, 214 and 216-219. 8 Shostakovich himself experimented with aspects of the twelve-tone organization of pitch material. See Child 1993 and Schmelz 2004.
5

3 pupils, Anton Webern and Alban Berg. 9 Though these techniques certainly were not officially taught within Conservatory walls, composition student Edison Denisov (19291996) and others reportedly sought the aid of their open-minded composition instructor Vissarion Shebalin to help in their self-study of as much of this music as they could acquire. Since access to this music was now legal but still not widespread, the scores and books studied by the young composers were not openly available through the Conservatory library, and had to be found elsewhere. Unfortunately, it is virtually impossible to know with any certainty what specific pieces these composers studied. It was during this time that Denisovs fellow student Andrey Volkonsky (19332008) composed Musica Stricta (1956-57). This four-movement work for solo piano is generally considered to be the first piece of Soviet serial music. 10 As I will discuss in Chapter 2, the first movement of this piece is organized around manipulations of an unordered pitch-class set (pc set), and therefore does not fit into standard Western definitions of serialism. 11 It may be described more comfortably by the general label of freely atonal music. On the other hand, the second, third and fourth movements are serial, as they are based on the transformation of ordered twelve-tone rows. 12 But, in non-Schoenbergian fashion, all three of those movements derive their pitch material from multiple, simultaneous twelve-tone rows: the second movement features four distinct rows used in counterpoint, while the third and fourth movement each use two, and all three movements make use of non-row-based, freely atonal material.
Schmelz 2009 provides a great deal of evidence regarding the acquisition of this awareness. See his Chapter 2, especially 28-33, 39-49 and 54-66. 10 Schmelz 2009, 81. 11 To cite just one of these standard Western definitions, Dave Headlam defines twelve-tone music as music in which all pitch events in a piece can be traced, without an unreasonable amount of difficulty, to the common source of the ordered intervals of a twelve-note row, taking pieces like Schoenbergs Variations for Orchestra op. 31 and Weberns Symphony op. 21 as examples. See Headlam 1996, 195. 12 Throughout this dissertation, I will use the word row for what should more accurately be called a row class the set of up to 48 row forms related by the basic twelve-tone operations.
9

4 Peter Schmelz, one of the main researchers of the unofficial music of the Thaw, has speculated that the young Russian composers interest in serialism may have grown from the desire to experiment with a forbidden fruit. 13 Because they and the composers of the previous generation had been unable to explore this method in the past, they may have felt a strong urge to catch up with their European and American contemporaries. 14 Schmelz goes on to offer a view of serialism as a symbol of resistance in the Soviet Union. Having been forced to ignore this compositional opportunity, the young composers may have felt that Soviet music had been artistically compromised. In Schmelzs view, the adoption of serial techniques became a symbol of intellectual and moral integrity. 15 A number of young Soviet composers experimented with serial techniques in the 1950s and 1960s. Edison Denisov, Andrey Volkonsky, Alfred Schnittke, Arvo Prt, Sofia Gubaydulina, Valentin Silvestrov, Nikolai Karetnikov and others were among those who tried their hands at dodecaphonic writing. However, among these composers, Denisov appears to be one of the few for whom the twelve-tone approach found a permanent home in his compositional style. For Denisov and the other young composers, dodecaphony was used as a compositional technique rather than a compositional method. Schmelz convincingly argues that the young composers did not view [serialism] as a monolithic methodthat implied a single style. They were more

The phrase young composers was used in official Soviet publications to describe these students who were experimenting with new approaches to composition. Though the words translate literally from the Russian molodye kompozitor, Schmelz demonstrates that, during the Thaw, this phrase acquired condescending connotations. See Schmelz 2009, 5-6, n. 11. 14 Schmelz 2005, 140-41. 15 Ibid., 143.
13

5 intent on applying the techniques and integrating them into their own, already developing styles. 16 It is one particular aspect of Denisovs serial music that will be the focus of this dissertation: his use of multiple twelve-tone rows in counterpoint or juxtaposition within a single composition (or movement). I have two primary goals. First, my fundamental objective will be to give analytical attention to what I strongly believe is an important, though neglected, repertoire. To this end, I will examine three pieces in which Denisov uses multiple, simultaneous twelve-tone rows: Five Etudes for Solo Bassoon (1st etude, 1983), Sonata for Alto Saxophone and Piano (3rd movement, 1970), and Octet for Winds (2nd movement, 1991). Furthermore, owing to its importance in Soviet music history, I will also provide an analysis of the multiple-row serialism of Volkonskys Musica Stricta. Second, the dissertation will explore technical issues involving multiple-row serialism in general, and demonstrate the utility of my own analytical approach developed in the study of this music.

DENISOVS STYLISTIC EVOLUTION, PIANO VARIATIONS Apart from a few scattered articles, the translation of an important monograph by Russian musicologists Yuri Kholopov and Valeria Tsenova, and Peter Schmelzs work, Denisovs music has received little attention in the English-speaking musictheoretic community. 17 As such, I shall provide a summary of some aspects of his musical career and compositional output here. 18 As early as June 1948, Denisov began

Schmelz 2002, 179. Throughout this dissertation, references to Kholopov and Tsenovas combined work will be given to show the location of the cited information in the Russian text and in the most recent English translation: 1993 and 2002, respectively. 18 Denisovs biography is treated more fully in Kholopov and Tsenova 1993, 1-43; 2002, 1-49.
16 17

6 corresponding with Shostakovich, in an effort to get the elder composers advice and evaluation of his compositions. Shostakovichs appraisal was generally positive, but he indicated in several different letters during the year 1950 that Denisovs melodic writing was the weakest aspect of his nascent compositional skill. 19 But it is clear that Shostakovich saw a budding talent in Denisovs work, and he continued to encourage him to apply for entrance into the Moscow Conservatory, even after he failed the required entrance exams in 1950. But this proved to be only a minor setback, as Denisov was successful in those exams in his second attempt a year later. From 1951-1956, Denisov studied at the Moscow Conservatory, and he continued as a post-graduate student from 1956-1959. During the early part of this time period, Denisovs music was largely tonal, likely due to the restrictions of the curriculum and the demands of the Soviet government. 20 Soon after graduating from the Conservatory, however, he began exploring non-tonal and dodecaphonic music. Denisov himself described the time period from 1959-1969 as his second conservatory. 21 During this time, he engaged himself in a personal study of music of many composers who had been left out of the Conservatorys official curriculum: Stravinsky, Bartk, Hindemith, Debussy, Schoenberg, and Webern. 22 It will be immediately apparent that Bergs name is conspicuously absent from this list, which includes his Second Viennese School partners. According to Kholopov and Tsenova,

19 Kholopov and Tsenova provide a number of Shostakovichs letters to Denisov. See Kholopov and Tsenova 1993, 172-183; 2002, 269-282. These letters are also found in French in Denisov and Armengaud 1993, 272-277. 20 Of course, it would be foolish to deny that Denisovs own personal aesthetic preferences might have played a role in the nature of his compositions. It is impossible to know the extent to which those personal preferences might have been influenced by the music to which he was exposed as a student at the Moscow Conservatory. 21 Denisov and Shulgin 1998, 22. 22 Kholopov and Tsenova 1993, 20-22; 2002, 19-21.

7 Denisov failed to have time for [Bergs music]. 23 It is surprising, therefore, that Denisovs later music should share several traits with that of Berg: in particular, the use of multiple rows (both derived from one another, and with no apparent connections) and material based on a twelve-tone row or rows presented side-by-side with non-rowbased material. 24 The year 1961 marked Denisovs first attempts at twelve-tone composition: Music for Eleven Wind Instruments and Timpani and Piano Variations. 25 The serial writing in these two pieces contrasts strongly with the first serial efforts of Volkonsky. While Volkonskys Musica Stricta will be discussed in detail in Chapter 2, as mentioned above, the work is fairly non-Schoenbergian in the approach to twelve-tone composition used in the piece: in particular, multiple rows, neither of which, separately or together, accounts for the pitch organization of the entire musical texture. But while Volkonsky uses his own idiosyncratic style of dodecaphony in his earliest work, Denisovs earliest twelve-tone works actually represent the most conventional of any of his works using the technique. 26 Volkonskys unorthodox twelve-tone traits may thus be understood as the result of a young composer experimenting with an incompletely understood technique, while Denisovs later unorthodox trespasses may be interpreted as willful transgressions of a system with which he had already developed a good deal of familiarity and comprehension. To demonstrate the awareness that Denisov acquired during the early part of the 1960s, it will be informative to examine one of his early twelve-tone works in
Ibid., 21; 20. Headlam 1996, 196-197. 25 Also composed in 1961, the second movement of Denisovs String Quartet No. 2 is a theme and variations in which the theme is a twelve-tone melody. See Kholopov and Tsenova 1993, 55-56; 2002, 61. 26 For reasons that will be discussed later, I consider Denisovs earliest twelve-tone works to be those written before his 1964 chamber cantata The Sun of the Incas.
23 24

8 conjunction with a staple of the Viennese repertoire. That Viennese piece is Weberns Op. 27 Piano Variations, which I believe may have provided a model, however informal, for Denisovs own Piano Variations. In 1957, Glenn Gould made a series of concert appearances in Moscow and Leningrad. Among his performances in Moscow was a lecture recital in the Small Hall of the Conservatory, where he briefly discussed Schoenbergs twelve-tone method and performed, among other works, Weberns Op. 27. 27 Denisov was in attendance at this recital, and later recalled the impact the performance had by saying, Ce fut une vritable rvlation. 28 The Webern piece must have made an enormous impression on Denisov, either at the recital or at some future time, as he published a theoretical analysis of the work in 1970. Weberns overriding concern for symmetry and palindromes in his Op. 27 is not at all present in Denisovs Variations. However, the canonic writing in Denisovs first and third variations, combined with the fragmented, pointillistic musical surface of the sixth variation all bear at least a superficial resemblance to the music of Webern. Within the structure of the row for Denisovs Variations lies a deeper structural similarity, though. Example 1.1 shows the two rows, with their respective ordered-interval sequences (INTs). It is easily seen that the first six entries in the two INTs are identical, but presented in a slightly different order. There is some evidence that Denisov found the interval content of row subsets to be an important part of the structure of a twelvetone row. In his analysis of Weberns Op. 27, he discusses the discrete trichords of the row in the following manner:

27 Schmelz discusses the details of Goulds visit to Moscow are discussed at some length. See Schmelz 2009, 55-59. 28 It was a true revelation. See Denisov and Armengaud 1993, 66.

9 EXAMPLE 1.1: WEBERN, OP. 27 AND DENISOV, VARIATIONS ROWS AND INTS A. WEBERN, OP. 27 29

B. DENISOV, VARIATIONS

The series is broken up into four three-note segments that constitute intervals of seconds and thirds. The first group is a combination of a minor second and a major third; the second group is a minor third and a major second; the third group is two minor seconds; and finally, the fourth group is, again, a minor second and a major third (but, because of the directional orientation of the thirds, [the first and last groups] are not isomorphic). 30 From this description, it is clear that Denisov is thinking of Weberns row in terms of unordered interval-class (ic) content. The parenthetical note at the end of the quotation shows an intervallic intuition similar to modern set theory: even though the first and last trichords are both made of a minor second and a major third, Denisov says they are not isomorphic, because the major third in the first trichord is descending, while the one in the final trichord is ascending. Perhaps not coincidentally, three of the discrete trichords from Denisovs row can be described in a similar manner: the first and third trichords are made of a minor second and a major third (enharmonically spelled), with the
This is what Denisov considered to be the prime form of Weberns row, though other authors consider this to be a retrograde-inverted form. See Denisov 1970, 48 and Bailey 1991, 24, 341 and 350. 30 Denisov 1970, 47.
29

10 direction of both intervals flipped in the third trichord. Additionally, the second trichord is made of two seconds, like the third trichord of the Webern row, but two major seconds. Two notable characteristics of Denisovs later style are already present in this early work. As previously mentioned, the first is the close canonic writing of the first, third, and sixth variations. Excerpts from each of these three variations are shown in Example 1.2.

EXAMPLE 1.2: DENISOV, PIANO VARIATIONS EXCERPTS FROM VARIATIONS 1, 3, 6 A. VARIATION 1: MM. 10-13

B. VARIATION 3: MM. 25-29

C. VARIATION 6: MM. 73-78

11 This kind of canonic writing bears a striking resemblance to that found in a number of Weberns works: the Five Canons Op. 16, the first variation of the second movement of his Symphonie Op. 21, the second movement of his Variations Op. 27, and the second movement of his Cantata Op. 29. 31 The second aspect of Denisovs later style found in the Piano Variations lies in the row segmentation of the third and sixth variations. As seen in Examples 1.2B and 1.2C, Denisov is fond of dividing a row into segments, and then manipulating the order of the pitches within each segment. In the first five measures of Variation 3 (Example 1.2B), the pianists right hand presents the first pentachord of row form I1, forward and backward, followed by the final pentachord of I1, backward and forward. 32 Finally, the left hand completes the row statement by providing the missing pitches, at order positions 5 and 6. 33 Denisov frequently treats row segments in this manner, blurring the distinction between P- and RP-forms of the row (as well as I- and RI-forms). 34 This technique is not unlike Denisovs own interpretation of the row manipulations in the first movement of Weberns Op. 27. In his article, Denisov writes, all the row forms in the first variation are wrapped [zavorachiayutsya] (i.e. if one hand begins a row form from its beginning, the other hand simultaneously begins the same row form from its ending). 35 Essentially, he is saying that one hand presents the early order positions of a single row form while the other hand presents the later order positions, as shown in Example 1.3. This has the net effect of rendering any row forms
On Weberns use of canonic techniques in these and other works, see Bailey 1991, 94-146. Throughout this dissertation, I will label row forms using the letters P, I, RP and RI to indicate which of the standard twelve-tone operators is used. The row form label will be followed by a subscript indicating the first pitch class of the row form. In the case of RP and RI, the subscript will represent the last pitch class of the row form (which is, of course, the first pitch class of the related P or I form). 33 Following the convention in Mead 1985 and Headlam 1996, I will underline order position numbers to visually distinguish them from pitch-class integers. 34 This work is also discussed in Schmelz 2009, 139-145. 35 Denisov 1970, 48.
31 32

12 labeled as retrograde or retrograde-inversion virtually indistinguishable from their nonretrograde counterparts.

EXAMPLE 1.3: WEBERN, OP. 27 MM. 1-7

Denisov labels all row forms in the first movement with curved arrows like those shown in Example 1.3. The direction of the arrow is intended to show the path of the row between the two hands. With this notation, it is clear that Denisov interprets the first four measures as one statement of P4 which wraps itself from right hand to left hand: the rows first hexachord is presented in order in the right hand, and simultaneously the rows second hexachord is presented in retrograde in the left hand. Variation 3 from Denisovs own Variations (Example 1.2B) seems to involve a similar kind of wrapping, but on an entirely horizontal plane. Instead of presenting a row segment forward and in retrograde simultaneously, Denisovs Variation 3 presents those versions of a given row segment successively, as indicated by the brackets in Example 1.2B. An article from 1973 indicates that, for Denisov, the manipulation of subsets of a twelve-tone row was a standard part of the composers serial technique. The article The Compositional Process consists of Denisovs personal musings on what are more or less self-evident statements about music composition. But among those musings, Denisov has the following to say about dodecaphonic composition:

13 In 12-tone and serial technique there are other guiding principles for organizing sound-materials[,] and over the years these have become increasingly sophisticated. Here we work with sets in the mathematical sense of the word set, and the process of composition lies in (1) the selection of segments from these sets; (2) organizing from these segments a succession of sub-sets (for instance, of harmony); (3) the establishment of definite logical correlations, both between the segments of the chosen sets (the establishment, for instance, of a pattern of consonant relationships) and between those sounds which are included in the complex of sub-sets (the organization of definite patterns of harmonic sequence, or the creation, for instance, of thematic and tonal relationships by intervals); and also (4) the ordering of this complex of sounds, precisely and coherently. 36 This statement suggests not only the importance of subsets of a twelve-tone row, but also touches upon a topic of crucial importance in the present study: the organization of definite patterns of harmonic sequence between the elements of those subsets. A way of understanding patterns created by organized successions of interval classes will be presented in Chapter 3.

DENISOVS INDIVIDUAL STYLE Kholopov and Tsenova single out the year 1964 as the beginning of Denisovs individual style. 37 Composed in this year, his chamber cantata The Sun of the Incas is cited not only by Kholopov and Tsenova, but also by Susan Bradshaw as the first work in which the composers individual voice can be heard. 38 This piece even brought Denisov his first taste of international recognition. 39 Following the premiere performance in Leningrad, the piece was performed in Darmstadt and Paris, under the baton of Bruno Maderna. There is also evidence that Igor Stravinsky and Roger
Denisov 1973, 9. Kholopov and Tsenova 1993, 51; 2002, 57. 38 Kholopov and Tsenova 1993, 22; 2002, 21; Bradshaw 1984, 3. 39 While an analysis of The Sun of the Incas is not within the scope of this present study, some of its relevant technical aspects have been discussed in Schmelz 2002 and 2009, Kholopov and Tsenova 1993 and 2002, and Bradshaw 1984.
36 37

14 Sessions knew and were quite impressed with the work. Predictably, this positive impression of the piece was not shared by the Soviet authorities. There was a significant amount of controversy about the pieces Leningrad premiere, and it was uncertain whether the performance would be allowed. Additionally, a performance in Moscow was scheduled and abruptly cancelled. 40 From this point until his death in 1996, his music maintained a remarkable degree of stylistic continuity. As a way of describing aspects of Denisovs style, Kholopov and Tsenova catalog a number of characteristic genres [zhanr ]. These genres are better understood as rhythmic and/or textural gestures, typical of Denisovs individual style. 41 A general understanding of several of these gestures will prove to be useful in the analytical portions of this dissertation, particularly in Chapter 6, an analysis of the Octet for Winds. The most relevant of Kholopov and Tsenovas genres for the pieces discussed in the present study include the following: 1) High lyricism [Vsokaya lirika] 2) Lyrical bands [Liricheskaya vyaz ] 3) Shooting, pricking, and sharply rhythmicized points [Strelba, ukol, ostroritmizovannye tochki ] 4) Pointillistic bursts [Puantilisticheskiye vspleski ] 5) Rustles, smooth threads [Shorokhi, Gladkiye niti ] These are five of the nine genres given by Kholopov and Tsenova. Two of the remaining four (sonorous masses [sonor-mass ] and sonorous duplications, mixtures and clusters [sonor-dublirovki, mikst, klaster ]) deal primarily with ways of thickening a melodic line by adding voices to create a series of dissonant, moving clusters. 42 The final two genres not included here are aleatory [aleatorika] and traditional artistic genres

Kholopov and Tsenova 1993, 22-24, 29-30; 2002, 22-23, 28-29. Ibid., 61-72; 67-84. 42 Ibid., 66-68; 76-79.
40 41

15 [traditsionnye khudozhestvennye zhanr ], which, as their names imply, are not so much rhythmic/textural gestures as they are indications of broader stylistic attributes. EXAMPLE 1.4: OCTET FOR WINDS (MVT. 1) MM. 1-5

Kholopov and Tsenova consider the lyrical band gesture-type to be the most distinctive calling card [vizitnaya kartochka] of Denisovs style. 43 Example 1.4 shows a lyrical band which occurs at the beginning of the first movement of Denisovs Octet for Winds. Kholopov and Tsenova note that the term lyrical band refers to strings of several voices in a quasi-arhythmic and ametrical presentationIt sounds as a thick melody, for the constituent voices merge together without [any single voice] as the

43

Ibid., 64; 70.

16 leading one 44 In short, Kholopov and Tsenova refer to the entire texture shown in Example 1.4 as a lyrical band. The individual lines of a lyrical band are defined by Kholopov and Tsenova as rustles or smooth threads. The distinction between rustles and smooth threads is primarily the tempo and dynamic level of the gesture: rustles are faster, softer gestures with an illusory effect [effekt prizrachnosti ], and are often notated as unmeasured grace notes. When the gesture is slowed down enough that the figure acquires an identifiable rhythmic profile, it is termed a smooth thread. 45 It is through the combination of these smooth threads that the texture of a lyrical band results. These smooth threads tend to consist primarily of small intervals, which combine to create a weaving chromatic line, which typically results in a chromatically saturated (or nearly saturated) pitch range. 46 Any of the individual instrumental lines in Example 1.4 can be seen to represent an extremely slow smooth thread: the first bassoon part in mm. 2-5, for example, winds around to chromatically fill in the space between E3 and B3. Often, though, smooth threads occur in faster rhythms, as shown in Example 1.5. Though not noted as such by Kholopov and Tsenova, the shooting gesture is found perhaps as frequently in Denisovs music and is every bit as much a calling card of his style. Their definition refers to quasi-unordered pointillistic simultaneous statement[s] of accented staccato sounds or disjointed chords in all registers in turn 47

Ibid. Ibid., 66; 75. 46 Kholopov and Tsenova suggest that these weaving chromatic lines might be related to the EDS intonation, which is the pitch realization of several of the letters of the composers name.
44 45

We might imagine these smooth threads as fortspinnung-like continuations of the intervallic succession given by either of these two motives. See Kholopov and Tsenova 1993, 83; 2002, 101. 47 Ibid., 65; 72.

17 EXAMPLE 1.5: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 1) M. 77

The examples mentioned in the text reveal, to some extent, the importance of repeated pitches in this gesture-type. This appears to be the primary difference between shooting and pointillistic bursts, though the authors do not explicitly state this. 48 Shooting is most often found in closely imitative textures, as Examples 1.2A and 1.2B show. The third and fifth movements of Denisovs The Sun of the Incas demonstrate two different realizations of the gesture. In the third movement, the individual rhythms are syncopated, but they are all based on a uniform subdivision of the beat, as shown in Example 1.6. The entirety of this twenty-six-measure movement utilizes thirty-second note beat divisions, in close imitation: a clear example of shooting. 49

In support of my understanding of shooting as being primarily centered around repetitions of a single note or chord, Kholopov and Tsenova refer to the broken rhythms of the finale from Stravinskys Rite of Spring. See Kholopov and Tsenova 1993, 65; 2002, 72. 49 This movement also demonstrates Denisovs penchant for using segments of a single twelve-tone row. While The Sun of the Incas is based on a single twelve-tone row (combined with non-row-based material, particularly in the second, fourth and fifth movements), the third movement derives all of its pitch material from transformations of the first hexachord of this row: the second hexachord is never used.
48

18 EXAMPLE 1.6: THE SUN OF THE INCAS (MVT. 3) MM. 9-13

Examples 1.7A and 1.7B show two brief shooting statements in the fifth movement of The Sun of the Incas. In this movement, as in many of Denisovs later works (including his Octet for Winds, as we shall see in Chapter 6), the shooting gesture first appears as part of an imitative texture, yet is presented later in a homorhythmic texture at the climax of the movement. Denisovs pointillistic bursts often have similar rhythmic shapes as the shooting gestures shown in Example 1.7: highly syncopated, with frequent shifts of beat division. According to Kholopov and Tsenova, the distinction between the two gestures is that shooting implies an extremely tense rhythm while [the pointillistic burst] is a multi-element splash [pyatno]. 50 It seems that, for the most part, the intensity of shooting is replaced by the lighter splashing of pointillistic bursts by means of replacing repeated-note figures with single statements of individual pitches, often at wide melodic intervals. Example 1.8 shows three measures from the first movement of Denisovs Sonata for Alto Saxophone and Piano which demonstrate how these pointillistic bursts (m.

50

Kholopov and Tsenova 1993, 65; 2002, 74.

19 EXAMPLE 1.7: THE SUN OF THE INCAS (MVT. 5) A.

MM. 19-21 (PIANOS AND PERCUSSION OMITTED)

B.

MM. 34-38 (PIANOS AND PERCUSSION OMITTED)

19) can be used as a tension-releasing eruption to contrast the building intensity of shooting. In this case, a uniform beat division is used throughout. Kholopov and Tsenova are not clear on the specific manner in which pointillistic bursts offer a release of tension, and, certainly, tension in a musical passage can be relieved in a number of ways. In the particular case of the Sonatas first movement, the repeated-note gesture is

20 gradually introduced from the beginning of the movement, starting in the piano part in m. 5, with the repetitions increasing in number from mm. 7-12. In mm. 17-18 (shown in Example 1.8), the repeated-note figure is divided between the three voices of the musical texture. With each repetition of a particular single note, there is an ever-greater yearning for change of pitch. As such, when the repeated notes disappear in m. 19 and are replaced with large intervallic leaps, this desire is fulfilled, thus suggesting a subtle release of the tension of the preceding measures. EXAMPLE 1.8: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 1) MM. 17-19

Kholopov and Tsenovas taxonomy of characteristic rhythmic/textural gestures leaves each individual genre broadly defined; they do not analyze thoroughly any single gesture in great detail. The authors primary goal is to provide the reader with a set of audibly identifiable traits that are particular to Denisov. In Chapter 6, I will take some preliminary steps toward the goal of attaining a more thorough understanding of the nature and usage of these gestures, but even my study will focus more on the interaction between specific gestures.

***

21 A broad description of Denisovs music is a characteristic that runs throughout Kholopov and Tsenovas monograph. The book deals with biographical information, issues of general compositional style, and some brief analytical observations about a number of specific pieces. Their research has provided an important foundation for Denisov scholarship; without it the present study would scarcely be possible. It is from their work and the work of Peter Schmelz (which will be discussed further in Chapter 2) that the present dissertation arose. This dissertation divides into two large parts between Chapters 3 and 4. The first part serves to situate Denisovs music within its proper historical context and to discuss the analytical methodology which will be used in the second part of the study. In the present chapter, I have been mostly concerned with introducing the reader to broad stylistic issues. Chapter 2 will provide an annotated translation of a chapter from Svetlana Kurbatskayas book Serial Music: Questions of History, Theory, Aesthetics [Seriynaya muzka: vopros istorii, teorii, estetiki]. This book was the first Russian-language monograph devoted entirely to the study of serial music, published in 1996, the year of Denisovs death. Though this books circulation appears to have been fairly limited, it was written under the advice of Yuri Kholopov (1932-2003), the most prominent Russian theorist of his generation. This book provides interesting insight into a uniquely Russian understanding of what serial music is. In one of the early chapters of the book, Kurbatskaya defines no fewer than twelve different categories of what she calls twelve-toneness (dvenadtsatitonovost ). Between this source and the Russian Musical Encyclopedic Dictionary (Muzkalny Entsiklopedicheskiy Slovar, henceforth MES ), which has different definitions for terms that Western readers would generally consider to be essentially synonymous (serial music, serial technique, dodecaphony, twelve-tone music), one can begin to accept the

22 possibility that Russian theorists might be more restrictive in their definitions than their counterparts in the West. The purpose of Chapter 2 is, thus, to discuss these twelve brands of twelve-toneness and to demonstrate how they can inform an understanding of some of the apparently non-serial aspects of Volkonskys Musica Stricta. In Chapter 3, I will move to a discussion of some of the general issues of multiple-row serialism. As I see it, there are three broad categories of issues that arise when considering multiple-row serialism: issues of structure, function and unity. Within this chapter, I shall introduce a new analytical device, the interval-class succession graph, which can illuminate Denisovs music by offering a way of conceptualizing the structure of a twelve-tone row which heretofore has been unexplored. Chapters 4 through 7 make up the second large part of the dissertation: the analytical discussion of three specific pieces involving multiple twelve-tone rows. The pieces are the following: the first of Five Etudes for Solo Bassoon (1983), and the finales of Sonata for Alto Saxophone and Piano (1970) and Octet for Winds (1991). I have selected these particular pieces because their dates of composition span a full twenty-one years of the mature period of Denisovs career. It is important to remember that the use of multiple rows in a single piece is not the composers typical modus operandi. His compositional output is characterized by the use of a wide variety of methods of which serialism is but one subset. Furthermore, not all the serial pieces in Denisovs oeuvre are based on multiple twelve-tone rows. I must make it clear at the outset of this study that it is not my goal to unlock the secrets of Denisovs compositional style. My goal, as previously stated, is to examine the details of one specific aspect of this composers wide-ranging style.

23 CHAPTER 2 THE RUSSIAN UNDERSTANDING OF SERIALISM: AN ANNOTATED TRANSLATION AND ANALYSIS Before beginning to examine Denisovs own idiosyncratic approach to serialism, we must pause to take account of some Russian theoretical thought about serialism, and to see how this theorizing might inform analysis. For this purpose, I shall present an annotated translation of the second chapter from Svetlana Kurbatskayas book Serial Music: Questions of History, Theory, Aesthetics [Seriynaya muzka: vopros istorii, teorii, estetiki ]. Following this, I shall demonstrate the relevance of Kurbatskayas writing for an informed discussion of the earliest piece of Soviet serial music, Andrey Volkonskys four-movement work for solo piano, Musica Stricta.

ANNOTATED TRANSLATION: SVETLANA KURBATSKAYA, SERIAL MUSIC: QUESTIONS

OF HISTORY, THEORY, AESTHETICS (CH. 2)

As mentioned in Chapter 1, Svetlana Kurbatskayas book Serial Music: Questions of History, Theory, Aesthetics (1996) carries the interesting distinction of being the first Russian-language monograph dedicated entirely to the topic of serial music. 1 Peter Schmelz book Such Freedom, If Only Musical addresses the history of Russian writings on serialism in some detail, and, while it is not my intent to duplicate his efforts here, a few notes on the topic are in order. 2 In the foreword of her book, Kurbatskaya notes that there is a large body of significant writings on serialism from outside Russia and identifies some of the reasons for the apparent delay in the creation of a Russianlanguage monograph on the topic.

1 2

Kurbatskaya 1996, 2 and Schmelz 2004, 353 n. 49. See, especially, the section of his Chapter 2 titled The Second Conservatory, Schmelz 2009, 39-66.

24 The doubtless numerical superiority of foreign studies speaks of the well known, sad circumstances of the Soviet epoch, when ideological dictates in musical science for a long time seriously prevented the purposeful studying of serial (as well as any other formalistic, incomprehensible) music. Prior to the beginning of the 1980s, publications on problems of [serialism] were extremely rare and isolated (it was not always even possible to call things by their proper names, for example, Descriptions of Modern Harmony, by Yuri Kholopov, 1974). Sometimes between the writing of a work and its appearance in the press, several years passed, and it is not possible to explain this only by the expensive costs of the publishing process. So Edison Denisovs article Dodecaphony and the Problems of Modern Compositional Techniques is extremely valuable for its time, with its thoroughly developed historical and theoretical base, and the abundance of musical examples. It was published in 1969, only six years after it was written. 3 Kurbatskaya goes on to mention three Russian books dealing with the topic to some degree: Kholopovs Lessons in Harmony [Zadaniya po garmonii ] (1983), Natalya Sergeyevna Gulyanitskayas Introduction to Contemporary Harmony [Vvdenie v sovremennuyu garmoniyu] (1984), and Anton Webern [Anton Vebern] (1984) 4 co-authored by Yuri Kholopov and his sister, musicologist Valentina Kholopova. While none of these books are exactly about serial music, the first two do contain special sections 5 on serial techniques and the third, at least, deals with a single composers approach to the serial method. Kurbatskaya notes that, with the appearance of these works, the position changed, as if serial dodecaphony was terra incognita (or musica non grata) for domestic musicologists. 6 The Russian sources on serial music which pre-date Kurbatskayas book fall into three categories: propagandistic articles in official publications, unofficial essays, and chapters in books on the broad category of twentieth-century harmony. 7 The first

Kurbatskaya 1996, 3-4. This book is better known to Western readers in its 1989 German translation, Anton Webern: Leben und Werk. 5 Kurbatskaya 1996, 4. 6 Ibid. 7 It should be noted, of course, that the present discussion does not deal with the assorted Western sources, legal or otherwise, that the young Soviet composers were able to obtain through various channels.
3 4

25 category includes writings like Grigoriy Shneyersons On Music Living and Dead [O muzke zhivoy i mertvoy] (1960) and numerous articles from the official journal Sovetskaya muzka. Schmelzs book shows that, while the purposes of these writings largely were to deride the serial method, they often backfired. 8 Shneyersons book, for example, contained a number of vitriolic statements such as the following: [Schoenberg] succeeded in confusing and destroying much in musical art, but he did not succeed in creating anythingSuch manifestations as dodecaphony, abstract painting, and existentialist philosophy are natural and unavoidable results of bourgeois decadence and its reactionary ideology. 9 But alongside such blatantly propagandistic statements were a wide array of musical examples demonstrating a variety of serial techniques. The same is true of articles by Johannes Paul Thilman and Marcel Rubin that appeared in Sovetskaya muzka. Thilmans article, especially, contained such a detailed description of the serial method that it functioned more as a primer for young composers. 10 Countless articles in the official press, among them Dmitri Kabalevskys Music and the Present [Muzka i sovremennost ] (1960) and Noemi Mikhilovskayas Notes about the Works of the Young [Zametki o tvorchestve molodkh] (1960), while not devoted exclusively to serialism, provided negative comments on any type of Western-influenced formalist writing within the Soviet Union, speaking largely in vague generalities rather than displaying any evidence of close study of specific pieces of music. When specific victims were required, these types of articles frequently turned their attention to the music of the young

These materials, as well as non-printed influences (such as the impact of Glenn Goulds 1957 visit to the Soviet Union) are discussed at length in Schmelz 2009, 45-66. 8 Schmelz 2009, 41-45. 9 Quoted in Schwarz 1965, 92. 10 See Schmelz 2009, 42-44.

26 composers of the Moscow Conservatory, offering condescending remarks couched as words of encouragement. 11 The second category is comprised of articles meant as responses to those in the first category. Denisovs writings figure especially prominently in this category. Schmelz notes that he was the most prolific Soviet writer on the topic, the one who had obviously contemplated and studied serialism the most thoroughly and the most sympathetically. 12 A glance at a list of his writings 13 reveals the following provocative titles: For Objectivity and Fairness in the Evaluation of Contemporary Music [Za obektivnost i spravedlivost v otsenke sovremennoy muzki ] (1965), 14 Some Words about A. Webern [Neskolko slov ob A. Veberne] (1966), The New Techniques Are Not a Fad [Novaya tekhnika eto ne moda] (1966), Dodecaphony and the Problems of Contemporary Compositional Techniques [Dodekafoniya i problem sovremennoy kompozitorskoy tekhniki ] (1969), and On Some Melodic Types in Contemporary Music [O nekotorkh tipakh melodizma v sovremennoy muzke] (1986). 15 Although Dodecaphony and the Problems of Contemporary Compositional Techniques is the only one of the essays which includes the description of specific serial techniques, one of the primary purposes of these articles appears to be providing a rebuttal to the official sources that

11 The official criticism of Andrey Volkonskys student works is especially relevant here. While the Soviet press often praised him as one of the more gifted of the young composers, other articles condemned his atonal and serial efforts. Often, these condemnations placed more blame on Volkonskys liberal teachers (Shebalin), or simply his own youth (and, presumably, navety) than on Volkonskys own intentions. See Schmelz 2005, 148-157. 12 Schmelz 2009, 147. 13 See Kholopov and Tsenova 1993, 237-239; 2002, 323-329. 14 This writing is actually a typescript of the remarks that Denisov made at a conference on contemporary music, held by the Union of Composers of the USSR in 1965. The text is found in Tsenova 1999, 22-33. In her commentary on the remarks, Tsenova notes that many of the ideas in this article are similar to those found in the later Dodecaphony and the Problems of Contemporary Compositional Techniques. 15 In addition to these essays, Denisov also wrote analyses of a number of specific Western serial works: Weberns Variations Op. 27 (as discussed in Chapter 1), Schoenbergs opera Von Heute auf Morgen, Luigi Dallapiccolas opera Ulisse, and Luigi Nonos Il canto sospeso.

27 claimed that serialism was an unnatural method of composition. 16 On the contrary, Denisov attempts to build a case for the naturalness of the serial method by outlining a quasi-evolutionary path leading from the extended tonality of late Romanticism through the dissolution of tonality, to atonality, and finally (and inevitably) to serialism. In an extremely valuable prcis of this article, Schmelz points out that, in addition to the question of serialisms unnatural origin, Denisov also counters at least one other official criticism of the compositional technique: One of the other official criticisms that he is explicitly countering in his thorough discussionis the idea that twelve-tone music is modish, something that anyone can do, especially those without talent.Denisov also took pains to discuss the negative situations that confront composers as they employ serial techniques and especially total serialism, especially the resulting problems with perceptibility that frequently arise. 17 Both these explanations bear a remarkable similarity to the way Schoenberg himself dealt with the same issues. Early in Schoenbergs 1941 essay Composition with Twelve Tones (1) we find the following claim: The method of composing with twelve tones grew out of a necessity, 18 which is echoed by Denisovs statement that the appearance of this [dodecaphonic] technology was the logical consequence of the development of some parts of European musical art of the 19th and early 20th centuries, that dodecaphony was not devised by anyone, or imposed on art 19 While Denisov does not use Schoenbergs oft-quoted phrase emancipation of the dissonance, the paths to serialism described by the two composers are very similar. Additionally, Denisovs refutation of the argument that serial composition requires no talent is reminiscent of Schoenbergs stance on the issue:
Schmelz 2009, 148. Ibid., 149. 18 Schoenberg 1975, 216. 19 Denisov 1969, 489-490.
16 17

28 The introduction of my method of composing with twelve tones does not facilitate composing; on the contrary, it makes it more difficult. Modernistically-minded beginners often think they should try it before having acquired the necessary technical equipment. This is a great mistake. The restrictions imposed on a composer by the obligation to use only one [row] in a composition are so severe that they can only be overcome by an imagination which has survived a tremendous number of adventures. Nothing is given by this method; but much is taken away. 20 While Schoenbergs remarks do not mention total serialism or issues of perceptibility, the underlying tone is much the same. Schoenbergs essay is not directly cited in Denisovs article, but this does not necessarily mean that he was unfamiliar with the ideas expressed in it. Certainly, he could have been exposed to Schoenbergian ideas indirectly, from Western publications or other sources. Schmelz notes that the path to serialism traced by Denisov begins in much the same way as did the remarks Glenn Gould made from the stage at the Moscow Conservatory during a 1957 concert which Denisov attended, as discussed in Chapter 1. While Gould never uttered the phrase twelve-tone composition during his lecture, his remarks were clearly intended to show the underlying connections between the works he performed: excerpts from Bachs Art of the Fugue and Goldberg Variations, Bergs Sonata Op. 1, Weberns Variations Op. 27, and the first and fourth movements from Kreneks Piano Sonata Op. 92, no. 4. 21 Furthermore, Denisovs article does cite Schoenbergs Theory of Harmony (although it is unclear what version of this book he had in his possession), Hans Eislers Reden und Aufstzen, and mentions the titles of three books by Ren Leibowitz (Schnberg et son cole, Introduction la musique de douze sons, and Histoire de lopra). While it may be impossible to trace Denisovs familiarity with Schoenbergs ideas to one single source, it seems irrefutable that this familiarity did exist.

20 21

Schoenberg 1975, 223. For a fuller account of Goulds performance in Moscow, see Schmelz 2009, 55-59.

29 The category of apparently reactionary articles has another connection to the first category: non-Western understandings of what could be considered the fundamentals of the serial method. Thilmans article describes rotation, interpolation and substitution alongside Schoenbergs traditional three mirror forms (inversion, retrograde, and retrograde inversion). 22 Denisov also observes that, in serial compositions, the row can be divided into segments which can be used independently, but will always maintain their connection to the original series. 23 Both of these articles present as primary serial techniques a number of row manipulations which are fundamentally non-Schoenbergian. The third category of Russian writings on serialism which pre-date Kurbatskayas book involves monographs with individual chapters devoted to the subject. While a thorough survey encompassing all such books is beyond the scope of the present dissertation, I will mention just one representative example. Natalya Gulyanitskayas Introduction to Contemporary Harmony (1984) has chapters devoted to the following topics: chords [akkordika], modality [modalnost ], tonality [tonalnost ], atonality [atonalnost ], and serialism [seriynost ]. At fifty-three pages in length, the chapter on serialism in this monograph represents the longest treatment of the topic we have discussed thus far. Most of the chapter, however, treads over already-covered ground. This chapter is divided into eight sections with the following titles: 1. Preliminary Information [Predvaritelnye svedeniya] 2. The Series and Its Purpose in the Composition [Seriya i yeyo naznacheniye v kompozitsii ] 3. The Structure and Ordering of the Series [Struktura i sistematizatsiya seriy] 4. Serial Forms and Transposition [Seriyne form i transpozitsii ] 5. Questions of Texture and Form [Vopros faktur i form ]
22 This information is summarized from Schmelz 2009, 43. Schmelz speculates that these operations may come from a familiarity with Kreneks writings on serialism as well as his Lamentations of Jeremiah. 23 Schmelz 2009, 149.

30 6. 12-Tone Harmony [12-tonovaya garmoniya] 7. Combinatoriality of the Units of Harmonic Language in Serial Compositions [Kombinatorika edinits garmonicheskogo yazka v seriynkh kompozitsiyakh] 8. Methodological Recommendations [Metodicheskiye rekomendatsii ] Compared with the writings we have surveyed to this point, Gulyanitskayas chapter is most notable for including a section on combinatoriality. The preceding survey, albeit incomplete, gives us some sense of the state of Russian research in serialism prior to Kurbatskayas monograph. In the foreword to her book, Kurbatskaya identifies Weberns 1905 C#-C-E string quartet as the earliest experience of the serial organization of musical material. 24 As Example 2.1 shows, Weberns quartet contains no obvious twelve-tone rows, does not appear to be organized around any sort of ordered completion of the total chromatic, and was written at least sixteen years before Schoenbergs first completely serial movement. 25 This excerpt, at least, appears to be better described as organized around transformations of the opening motto, C#-C-E (labeled x in Example 2.1), and might find itself comfortably at home in an introductory course on post-tonal analysis as an example of a composition based primarily on transformations of [014]. While it may seem strange to a Western theorist to call this movement serial, Kurbatskayas statement must be understood as representing a more generalized definition of serialism. This is reflected in a number of entries in the MES for words which Western theorists would generally use more or less interchangeably: serialism [serialnost ], serial technique [seriynaya tekhnika], and dodecaphony [dodekafoniya]. All these definitions were written by Yuri Kholopov. As these terms appear in Kurbatskayas book, I will delay

24 25

Kurbatskaya 1996, 3. Whittall 2008, 31.

31 EXAMPLE 2.1: WEBERN, STRING QUARTET (1905), MM. 1-13

specific discussion of their definitions for the moment, being satisfied for the time being with the mere observation that these terms are all defined differently from one another in the MES. There is an additional Russian term, mentioned in Chapter 1, for which no exact Western counterpart exists, encompassing all of the terms in the previous paragraph, and others: dvenadtsatitonovost, which is most closely translated as twelve-toneness. In an essay analyzing aspects of Volkonskys Musica Stricta, Schmelz credits Kholopov with

32 coining this term, which is used to describe atonal music that sounds twelve-tone. 26 In the Musical Encyclopedic Dictionary, Kholopov defines the term in a somewhat more global sense: Twelve-toneness (dodecatonicism [dodekatonika], twelve-tone harmony [dvenadtsatitonovaya garmoniya]) a kind of twelve-step sonic system where each sound can function as an independent element. Twelve-toneness is connected with many new types of techniques in the music of the 20th century symmetrical harmonies (for example, in I. F. Stravinsky, O. Messiaen), synthetic chords (A. Roslavets), dodecaphony, tropes (J. M. Hauer), 12-tone rows (for example in D. D. Shostakovich, R. K. Shchedrin), 12-tone fields, chords, free atonality and other models of the 12-tone field. 27 Kurbatskaya, a former Kholopov student, adopts this term in its broader sense. 28 In her book, twelve-toneness is an over-arching concept of which she defines twelve specific varieties. As defined in Chapter 2, these twelve categories are as follows: 1. Free atonality [svobodnaya atonalnost ] 2. Technique of tonal centers [tekhnika zvukovogo tsentra] 3. Technique of synthetic chords [tekhnika sintetakkordov] 4. Twelve-note chords [12-zvukovye akkord ] 5. Technique of twelve-tone rows [tekhnika 12-tonovkh ryadov] 6. Technique of twelve-tone fields [tekhnika 12-tonovkh poley] 7. Technique of tropes [tekhnika tropov] 8. Serial technique [seriynaya tekhnika] 9. Dodecaphony [dodekafoniya] 10. Microserialism [mikroseriynost ] 11. Total serialism [serializm] 12. Serialism [seriynost ] The general organization of Kurbatskayas chapter seems to owe a debt to George Perles Serial Composition and Atonality (the second edition of which is cited by Kurbatskaya) in regard to two specific issues. The manner in which she proceeds through the definitions of her twelve categories of twelve-toneness is not unlike the

Schmelz 2004, 325. Kholopov 1990, 164. 28 Kurbatskaya and Kholopov also authored a book together which presents analyses of selected pieces by Pierre Boulez and Edison Denisov. See Kurbatskaya and Kholopov 1998.
26 27

33 organization of the second and third chapters of Perles book. Specifically, the order in which Kurbatskaya presents her first three topics is the same as in Perles book, even citing several of the same examples, such as Schoenbergs Erwartung, Skryabins Piano Sonata No. 7 and Roslavets Three Compositions for Piano. 29 Furthermore, Perles manner of dividing his chapter Motivic Functions of the Set 30 into subsections titled The Set As a Theme, The Set As a Melodic Prototype, and Segmentation is echoed to some degree by Kurbatskaya, who, as we shall see, takes care to note whether the series is treated as a melodic device or not, and whether or not the series in treated as a single, whole unit, or as a collection of smaller parts.

***

While a translation of the entire text of Kurbatskayas Chapter 2, On the System of Terminology [O sisteme terminologii ], is provided in Appendix 1, a summary with general notes and commentary is found here. As the preceding discussion shows, the purposes of Kurbatskayas chapter are first to define twelve-toneness, and second to define the twelve different categories that fall under its umbrella. Her definition of twelve-toneness does not differ significantly different from Kholopovs: Twelve-toneness is a property of the musical material according to which each of the twelve tones of the chromatic scale can be used as

Erwartung is discussed by Perle on p. 19 and by Kurbatskaya on p. 32, though Kurbatskaya only mentions the work by title (without a detailed musical example), having already directed the reader to Perles book in her first footnote of the chapter. Both authors give identical examples from Skryabins Piano Sonata No. 7, found on p. 43 in Perles book and p. 34 in Kurbatskayas. Finally, Perle discusses Roslavets Three Compositions on pp. 43-44, and Kurbatskaya cites the same work (again, without a detailed musical example) on p. 34. 30 The reader is reminded that Perle uses the word set for what we have been calling a series.
29

34 independent units of the musical fabric. 31 The phrase musical fabric [tkan ] is one that Kurbatskaya uses frequently in this chapter, often in relation to whether or not a particular compositional device is responsible for the entire musical fabric [vsya tkan ], or every note on the musical surface. 32 In the chapters introductory section, Kurbatskaya links the concept of twelvetoneness with another term from Kholopovs MES: dodecatonicism. The way the term is presented in both Kurbatskayas chapter and Kholopovs definition suggests that these two authors consider this word to be synonymous with twelve-toneness: it is enclosed in parentheses, with no explanation, after the word twelve-toneness. Kurbatskaya writes, twelve-toneness (dodecatonicism, twelve-tone harmony) is the system of thinking based on the autonomy of each of the twelve pitch classes,33 associating dodecatonicism and twelve-tone harmony with twelve-toneness, in exactly the same way as does Kholopovs definition (quoted earlier). In an essay titled After Prokofiev, Schmelz provides a bit more detail on the origin of this term. Schmelzs research traces this word to a 1965 article by Soviet musicologist Arnold Sokhor titled On the Nature and Expressive Possibilities of Diatonicism. Sokhors new word was intended to account for the expanded diatonic features of Prokofievs music, which, by having a similar appearance to the word dodecaphony, allowed Sokhor to portray Prokofiev as a fundamentally tonal composer whose works offered either a substitute or complement to Schoenbergs system. 34 Kholopov and Kurbatskayas apparent connection of this word to twelve-toneness could possibly indicate that these authors allow for an understanding of twelve-toneness in a tonal context, as well.
Kurbatskaya 1996, 32. By the musical surface, I am referring to all the notes in the score, without any sort of reduction. 33 Kurbatskaya 1996, 33. 34 Schmelz 2008, 515.
31 32

35 Kurbatskayas twelve categories of twelve-toneness group themselves as follows: #3 (technique of synthetic chords) is a version of 35 #2 (technique of tonal centers); #9 (dodecaphony) and #10 (microserialism) are each a type of 36 #8 (serial technique). We can also group together the categories which are related to the specific compositional procedure of an individual composer: #2 (technique of tonal centers) refers to the technique of Skyrabin, #3 (technique of synthetic chords) describes the technique of Nikolai Roslavets, #6 (technique of twelve-tone fields) references Nikolai Obukhovs concept of total harmony, and #7 (technique of tropes) is based on the method of Hauer. The definitions of several of Kurbatskayas categories are self-evident to the Western reader. Specifically, her definition of free atonality, twelve-note chords, and total serialism are not significantly different than their Western counterparts. Even so, a few brief comments on these concepts are in order. In the section on free atonality, Kurbatskaya mentions the hemitonicmethod of symmetrical interval groups. 37 The hemitonic method is described in Kholopov and Tsenovas book on Denisov, though it appears to date back to Valentina Kholopovas 1973 article Toward One Chromatic Principle in TwentiethCentury Music. 38 A full description of the theory of hemitonic groups is neither necessary for, nor within the scope of the present dissertation, but can be summarized as an analytical approach which considers combinations of ic1 with itself or with other interval classes. Kholopov and Tsenova thus define two categories of hemigroups (short

Kurbatskaya 1996, 34. Ibid., 37, 38. 37 Ibid., 32. 38 Kholopova 1973, 331-344.
35 36

36 for hemitonic groups): 1. a semitone with a semitone (coupling: 1+1+) and 2. a semitone with other intervals (1+n). 39 This section also lists, by title or opus number, five different pieces which can be considered as representative examples of music featuring a freely atonal type of twelvetoneness. Of these five pieces, Kurbatskaya only provides in her text notation and an explanation for one: Weberns Five Pieces for Orchestra Op. 10, no. 3. Her example is reprinted here as Example 2.2.

EXAMPLE 2.2: EXCERPT FROM WEBERN OP. 10, NO. 3 (FROM KURBATSKAYA)

39 Kholopov and Tsenova 2002, 112. This section is not present in the original Russian version of this book. Rather, it appears to have been inserted from Kholopovs contribution to a Festschrift created in honor of the composers 65th birthday. See Kholopov 1995, 84-94.

37 Unfortunately, Kurbatskayas verbal explanation of this particular musical example leaves much to be desired. Her comments are as follows: the opening section of Piece No. 3 (Op. 10) rests on a thirteen-note complex (eleven different pitch classes), including homogeneous groups which are distributed between a melody and a chordal background. 40 This example seems to be getting at the concept of (near) pitch-class complementation. The bottom staff of the example shows the two primary elements of the musical fabric: a six-note chord labeled background [fon], and a group of seven quarter-notes, played by violin and horn, labeled melody [melodiya]. The melody contains seven pitch-classes that are almost entirely separate from the six pitch-classes in the accompanying background (the pitch classes D and Ab/G# are common to both groups). Though these two musical elements do not form the full aggregate when taken together, the idea of primarily non-intersecting pitch-class sets appears to be at the heart of Kurbatskayas example. In addition, the bottom staff of the example shows the generative interval classes of both the background chord and the melody. Beginning with the background chord and progressing to the melody, Kurbatskaya labels the eleven distinct pitch classes, clearly revealing the common pitch classes and makes the absence of pitch class G obvious. Notes 1-6 are stemmed together and labeled with a 4 above the beam, which, of course is the interval class formed by G# and E. Within this ic4, this hexachord is broken up into the interval patterns < 1 4 > and < 4 2 >. Notes 7-10 are stemmed together and labeled as forming ic5, which is further divided into the interval pattern < 1 5 1 >. This interval pattern is nearly continued, as the final three notes in
40

Kurbatskaya 1996, 32.

38 the example form the pattern < 1 5 >. With this diagram, Kurbatskaya is trying to make the point that the background is built primarily out of ic4 and the melody is built out of a combination of ic1 and ic5. Total serialism is another of Kurbatskayas categories which resonates with Western musicians. Following Schmelz, I have translated Kurbatskayas word seriynost as serialism, even though the suffix -nost suggests that serialness might be a more literal translation. I have translated Kurbatskayas term for this technique, serializm, as total serialism, in order to avoid any confusion with the Western word serialism. 41 Based on the definitions given, it is clear that Kurbatskayas seriynost is meant in a more wide-ranging sense than is serializm. As such, I will refer to these two terms by their familiar Western counterparts: serialism and total serialism, respectively. Category #10, microserialism, deserves some special attention. It is this category that allows Kurbatskaya to make the claim noted earlier that Weberns C#-C-E quartet represents the earliest instance of serial organization. 42 She defines a microseries as a series of fewer than twelve pitch classes. 43 In addition to the Webern quartet, she also cites two Bransle dances from Stravinskys Agon, 44 In Memoriam Dylan Thomas, and Brittens Symphony-Concerto for cello and orchestra. Perhaps the most notable aspect of Kurbatskayas discussion of microserialism is her advancement of three distinct, but related, terms. The three terms microseries [mikroserii], incomplete series [nepolnaya seriya], and subseries [subseriya] all relate to the

Schmelz 2004, 353 n. 49. See note 24. 43 Kurbatskaya 1996, 38. 44 Kurbatskaya refers to two Bransle dances from the ballet Agon. It is curious that Kurbatskaya only mentions two Bransle dances when there are, in fact, three such dances in Agon: Bransle Simple, Bransle Gay, and Bransle Double. From the musical example she gives of the series of two Bransle dances, it is clear that she is referring to the Bransle Simple and the Bransle Gay.
41 42

39 ordering of a small number of pitch classes, 45 but all have slightly different meanings. Of the three, a microseries appears to be the only one which is a complete unit in itself. Kurbatskaya says that an incomplete series is part [chast ] of a complete twelve-tone series, which functions independently [of the complete series] in a specific section of the work. 46 On the other hand, a subseries is a segment of a series, which functions as a series on a micro-level. 47 At a glance, these terms appear to be synonymous, but Kurbatskaya seems to treat the words part and segment [segment ] differently. In Kurbatskayas use, the former has more of a top-down connotation, while the latter implies a more bottom-up perspective. Specifically, the concept of an incomplete series views the whole series as the given entity, and the very existence of the part of it which is used independently depends on the existence of the series as a whole. The concept of a subseries, however, takes the segment as the generative unit, from which a larger (twelve-note, or otherwise) series is built, through various transformations. Kurbatskaya cites no particular examples of incomplete series, although the third movement of Denisovs The Sun of the Incas is a clear example of this, as the entire movement is based entirely on transformations of the first hexachord of the row established in the first movement of the cantata. As an example of a subseries, Kurbatskaya cites the modus classicus of the technique: the row from Weberns Concerto for Nine Instruments Op. 24. She also cites the row from Lutosawskis Funeral Music in this category. 48 As previously mentioned, the most important aspect of Kurbatskayas categories lies in the careful distinction she makes between four concepts that appear to be essentially synonymous to Western readers. These concepts are her fifth, eighth, ninth,
Kurbatskaya 1996, 38. Ibid. 47 Ibid. 48 Ibid., 39.
45 46

40 and twelfth categories: technique of twelve-tone rows, serial technique, dodecaphony, and serialism, respectively. It is in the definitions of these four types of twelve-toneness that Kurbatskayas debt to Kholopov is the most clear: there are no major differences between Kholopovs definitions of these terms in the MES and Kurbatskayas definitions in her monograph. Underlying her discussion of these four concepts is a cautious differentiation of the following two pairs of terms: twelve-tone composition and dodecaphony, and series [seriya] and row [ryad]. Twelve-tone composition is meant as a general term, referring to a system of thought, while dodecaphony is a more specific method of composition. 49 Kurbatskaya likens this to the German language distinction between Zwlftonmusik and Zwlftontechnik. Kurbatskayas differentiation between series and row is based on a similar idea. A series is an organizational principle which determines the appearance of relationships in the work. 50 In contrast, a row is primarily a surface-level phenomenon: a row is a horizontally presented series. 51 The fifth category, the technique of twelve-tone rows, begins first with a definition of row. In Kurbatskayas words, a row is the horizontal occurrence of non-repeating pitch classeswhich is used as a melodic construction and is not the sole source of the musical fabric. 52 In a footnote, she notes the curiosity that melodies meeting her criteria have appeared in pre-20th century music as well, citing the subject from the B minor fugue in Book 1 of J.S. Bachs Well-Tempered Clavier and the introductory theme from Liszts Faust Symphony. While it is clear that she is not trying to imply that these pieces are based on the technique of twelve-tone rows, these pieces
Ibid., 37. Ibid. 51 Ibid., 36. 52 Ibid., 34.
49 50

41 meet the criteria established in Kurbatskayas definition of this technique: the row is found as a melodic construction, and, although transformations of that row may be used (particularly in Bachs B minor fugue), the row is not the sole source of the musical material. But she does cite three bona fide examples of the technique of twelve-tone rows: the last of Alban Bergs Altenberg Lieder Op. 4, Rodion Shchedrins Polyphonic Notebook and Shostakovichs Symphony No. 14. The inclusion of these particular examples demonstrates the significance of her claim that the row is not the sole source of all pitchclass material. The importance of this claim cannot be overstated for an understanding of the serial practices of composers like Volkonsky and Denisov. As mentioned earlier, Kurbatskaya defines dodecaphony (category #9) as a specific type of serial technique (category #8). Schmelzs Soviet serial bulls-eye diagram, shown in Example 2.3, is helpful in understanding the relationship between these two categories. Dodecaphony, then, is a specific type of serial technique where the series contains exactly twelve pitch classes.

EXAMPLE 2.3: PETER SCHMELZS SOVIET SERIAL BULLS-EYE DIAGRAM 53

53

Reprinted from Schmelz 2009, 135.

42 Distinguishing between the technique of twelve-tone rows and serial technique requires the invocation of Kurbatskayas distinction between row and series. Essentially, one can arrive at Kurbatskayas definition of serial technique by starting with the definition of technique of twelve-tone rows, changing twelve-tone row to series, and adding the requirement that the series be the source of the entire pitch fabric. As Kurbatskaya says, serial technique is one of the forms of contemporary musical composition where the entire musical fabric is derived from an invariant [invarianta] series and its continuous recurrences (in basic and derived forms). 54 Kurbatskaya closes her chapter by discussing serialism [seriynost ] as category #12. As she defines it here, serialism is really a different sort of category than the preceding eleven. It is clear that she considers serialism a wider-ranging phenomenon than any of the other categories she has discussed. She claims that serialism is one of the universal categories of musical thought in the 20th century, applied to a wide range of genetically related phenomena. 55 She gives a two-part definition for serialism, the first part of which is not strikingly different from her definition for serial technique. But the second part of the definition is especially important for the present study of multiplerow serialism in the music of Denisov: serialism is the principle of musical thought by a series (or multiple series) [seriyami ]. 56 While she does not directly discuss her use of the plural form of series here, it is clear that she is referring to two different ideas. First, she is allowing for the inclusion of total serialism, in which there may be multiple series governing multiple musical parameters. Her own example shows the twelve-tone pitch series, the twelve-unit durational series, the five-unit dynamic series, and the five-unit

Kurbatskaya 1996, 36. Ibid., 40. 56 Ibid.

54 55

43 articulation series from Oliver Messiaens le de feu. Secondly, by using the phrase multiple series here, she is allowing (or at least, not prohibiting) the use of two or more distinct series governing the same musical parameter. This chapter of Kurbatskayas book provides insight into the Russian understanding of serial composition and, more generally, twelve-toneness. The fact that her work resonates with (and was, in fact, written under the influence of) that of the prominent Russian theorist Yuri Kholopov, and the fact that its conceptual origins are based on Denisovs writings indicates that Kurbatskayas point of view is not dismissible as idiosyncratic or invalid. As the next section of this dissertation will demonstrate, it is not only intellectually worthwhile to categorize different types of twelve-toneness in a Russian context but it is analytically useful, as well.

***

ANALYSIS: ANDREY VOLKONSKYS MUSICA STRICTA (1956-1957) Andrey Volkonsky was one of the young composers who developed an interest in serialism during the Thaw. He and Denisov were students at the Moscow Conservatory during the same time period. However, Volkonsky was given an academic leave and eventually expelled from the Conservatory. While it is an oversimplification to say that he was expelled simply because he possessed scores of modern Western music, it does seem that this played some role in his removal from the school, though this was not the official reason given by the Conservatory administrators. 57

Volkonskys biography, including details regarding his expulsion, is thoroughly treated in Schmelz 2009. See Schmelz 2009, 68-77.
57

44 Composed shortly after the death of Stalin, Volkonskys Musica Stricta is generally accepted as the first piece of Soviet serial music. 58 Some general observations about the compositional techniques in each of the four movements will reveal that this piece actually provides examples of several of Kurbatskayas categories of twelve-toneness. The first movement is organized around transformations of an unordered pc set, and therefore does not fit into standard Western definitions of serialism, 59 but does fit in Kurbatskayas second category, the technique of tonal centers. On the other hand, the second, third and fourth movements do fit Western definitions of serialism, as they are based on the transformation of ordered twelve-tone rows. All three of those movements derive their pitch material from multiple, simultaneous twelve-tone rows: the second movement features four distinct rows used in counterpoint, while the third and fourth movement each use two, and all three movements make use of non-serial atonal material.

MOVEMENT 1: ANDANTINO In the opening movement, the A-B-A form (see Example 2.4) can be interpreted as arising from a conflict between two of Kurbatskayas categories of twelve-toneness: the technique of tonal centers and the technique of twelve-tone rows.

58 Schmelz 2009, 81. Schmelz has discussed Musica Stricta in his dissertation, book, and two separate articles. See Schmelz 2002, 84-99; 2004, 323-326; 2005, 160-179; and 2009, 81-89. 59 In a slightly broader sense than Kurbatskaya, the term freely atonal is often used by Western theorists to define note groups by their intervallic and/or pitch-class content, rather than by their order.

45 EXAMPLE 2.4: MUSICA STRICTA (MVT. 1) FORM

A B A

m. 1 [0126]
technique of tonal centers

10 [0126]

17 21 31 [0126] dissolves into [014], [013], etc. [0126]

technique of 12-tone rows technique of tonal centers

The A and A sections of the first movement are saturated with different forms of set class 4-5, [0126]: literally, every single note within these two sections can be explained by set-class membership, with the exception of an Eb in the last chord of m. 13. 60 In fact, the different forms of 4-5 are actually all transpositions of the unordered pc set presented in m. 1, {0128}. It is this {0128} tetrachord that functions as the tonal center of these two sections, and, as Kurbatskayas definition requires, the entire musical fabric is worked out by its regular repetitions in specific interval relationships. 61

EXAMPLE 2.5: MUSICA STRICTA (MVT. 1) MM. 1-9

60 This note could be a misprint. As it is, it forms [0136] with the other notes of the chord. However, were this note an E-natural, rather than Eb, it would form [0126]. 61 Kurbatskaya 1996, 33.

46 The opening nine measures are shown in Example 2.5, with the initial {0128} labeled as X, and all other occurrences of 4-5 labeled as transpositions of X. Over the course of the two A-sections (mm. 1-16, mm. 31-35), Volkonsky uses eleven of the twelve possible transpositions of X; T1(X) is the only transposition that is never used. The B-section (mm. 17-30) of this movement begins with more transformations of 4-5, but gradually, that set class loses hold of the pitch structure. The set-class structure of the B-section is much less consistent, and is dominated by a seemingly freer interaction between trichords, highlighting 3-3, [014]. Example 2.6 shows this interplay of trichords in mm. 21-22.

EXAMPLE 2.6: MUSICA STRICTA (MVT. 1) MM. 21-22

Throughout this section, the use of melodic aggregates appears to be the primary organizing force. As Example 2.7 shows, the first move away from the transpositions of X comes in m. 20, where a [0134] is sandwiched between T3(X) and T0(X). Here, the [0134] represents the complementary pitch classes to the combined {0123458e} of T3(X) and T0(X). 62 By adding this [0134], Volkonsky effectively unites two transformations of

62

As the interval-class vector of 4-5 [210111] indicates, T3 is the only transposition level that can be applied to X without resulting in any pitch-class duplication.

47 X into a single twelve-tone row. As such, it is possible to interpret this melodic line as representing a transition between the use of the technique of tonal centers to the technique of twelve-tone rows.

EXAMPLE 2.7: MUSICA STRICTA (MVT. 1) MM. 19-20

From this point until the end of the B-section, each measure contains one or two complete rows. There are four exceptions, which occur in m. 21, 25, 26 and 30. All of these exceptions involve the doubling or repetition of a single pitch class and the omission of another. Referring back to Example 2.6, the first two beats of m. 21 in the right hand present a near-aggregate, with Bb omitted, and C occurring twice. Example 2.8 shows a similar case (m. 30), where Ab appears twice, but A-natural is missing.

EXAMPLE 2.8: MUSICA STRICTA (MVT. 1) M. 30

48 Volkonskys use of the technique creates no fewer than fourteen different rows, but none of these rows are repeated in this movement, or in any of the other three movements of Musica Stricta. Volkonsky appears to be using the technique of twelvetone rows (in exactly the manner Kurbatskaya describes) in order to freely generate a wide variety of musical material, in contrast with the method of transpositions of a single tonal center.

MOVEMENT 2: ALLEGRETTO In the second movement, Volkonsky uses four different twelve-tone rows, and he always uses them in a horizontal manner, with vertical sonorities resulting from combinations of rows (and row forms) rather than individual row segments or partitions. The movements form evolves as the result of the process of juxtaposing these different rows, as shown in Example 2.9.

EXAMPLE 2.9: MUSICA STRICTA (MVT. 2) FORM

m. 1 AP2

5 BP3 CP0 DP5

7 AP2 ARP2

9 CP8 BP0 DP8

10 11 12 13 ARP2 BP4 AI2 DP1 CP8

14

15 16 17 19 AI2 DPt ARIt ARIt CP9 AP2 BP4

49 I have labeled the four different rows using the letters A-D, followed by a row-form label. 63 The square brackets on the diagram indicate the presence of freely atonal pitch material, which I will discuss later. These four rows are shown in Example 2.10.

EXAMPLE 2.10: MUSICA STRICTA (MVT. 2) ROWS A-D ROW A:

INT:

ROW B:

INT:

ROW C:

INT:

ROW D:

INT:

As suggested by the diagram in Example 2.9, the four rows are not presented with equal emphasis. Clearly, Row A is the primary row of the movement it is the only row that is presented in counterpoint with another form of itself. Rows B, C and D serve to provide contrasting material to Row A, while allowing each horizontal voice to
All told, movements 2-4 of Musica Stricta use eight different twelve-tone rows. I will label each of these with a capital letter, A-H.
63

50 still be serial. The four rows themselves bear little resemblance to each other in terms of their respective INTs. The most significant such connection seems to be that Rows B and D both contain two consecutive occurrences of ic5.

EXAMPLE 2.11: MUSICA STRICTA (MVT. 2) IC 5 SEGMENTS IN ROWS B AND D

Volkonsky makes this connection audible on the musical surface near the very beginning of the movement. In mm. 4-5, BP3 and DP4 both share the trichord {279}. For BP3, these pitch classes occupy order positions < 3 4 5 >, while in DP4, they occupy order positions < 1 2 3 >. Both of these trichords are presented with the same rhythm, one immediately after the other. At no other time in the movement do BPx< 3 4 5 > and DPy< 1 2 3 > actually state the same pitch classes, although they usually occur with the same

51 rhythmic profile, and in close succession. The four occurrences of these [027] segments of Rows B and D are shown in Example 2.11. It should also be noted that the initial statement (mm. 4-5) is immediately preceded by freely atonal material, stating exactly the same pitch classes, removed from the context of any twelve-tone row (Example 2.12). These three fast statements of {279} serve to attune the listeners ears to these short 5cycles that will occur at various points throughout the movement, separating the Row A material from the groups of Rows B, C and D. 64

EXAMPLE 2.12: MUSICA STRICTA (MVT. 2) MM. 3-4

The kind of inter-row invariance exemplified by the shared {279} described above also appears to guide Volkonskys choice of row-forms throughout the piece, though not in any systematic way. In the four large formal sections (refer to Example 2.9), the following inter-row invariances appear: Section 1: AP2 and CP0 share {4579} BP3 and DP5 share {346} and {279} CP0 and DP5 share {0te} Section 2: AP2, BP0, and CP8 share {013} AP2 and CP8 share {0139te} ARP2 and CP8 share {567} Section 3: ARP2 and CP8 share {567} (as in Section 2) ARP2 and BP4 share {457} ARP2 and CP8 share {013} (as in Section 2) AI2 and DP1 share {09e} AI2 and CP8 share {9et} BP4, CP8, and DP1 share {39t}
64

For more on 5-cycles and cycle theory in general, see Headlam 1996.

52 BP4 and CP8 share {239te} CP8 and DP1 share {678} Section 4: AI2 and DPt share {1345} ARIt and BP4 share {3578} ARIt and CP9 share {124te} With the exception of the opening three measures, forms of Row A are always presented in pairs. The pairing of Row A forms cycles through AP2, ARP2, AI2, ARIt, and finally returning to AP2. In each Row A section, there are two row forms: one in the right hand part of the piano, and the other in the left. These two row forms always involve one old and one new form of the row. Volkonsky begins with AP2 in the opening three measures. In the next Row A section (mm. 6-8), AP2 remains, but this time it is accompanied by a new form, ARP2. In mm. 11-12, ARP2 remains, accompanied by AI2. AI2 remains in mm. 15-16, with the addition of ARIt, and finally, in the final measures of the movement, AP2 returns as the lower voice to the still-present ARIt. The choice of forms of Row A is based on the opening pitch of each form; the forms of Row A alternate between starting on D and F#. The assignment of row form to linear voice follows a pattern best described as a form of invertible counterpoint. Each new row form is presented by the pianists left hand and is moved into the right hand when the next new row form enters. This change of voicing is shown in Example 2.13.

EXAMPLE 2.13: FORMS OF ROW A IN TWO VOICES RH (old): LH (new): mm. ---AP2 1-3 AP2 ARP2 6-8 ARP2 AI2 11-12 AI2 ARIt 15-16 ARIt AP2 19-21

53

Rows B, C and D are presented in a similar manner. As before, each row is restricted to a single linear voice: high, middle or low. As such, the texture of the movement is constantly changing from the two-voiced Row A sections to the threevoice multiple-row sections. Each time that these three rows are used simultaneously, they are presented in a different vertical alignment. Because there are only four sections in the movement that use these three rows, Volkonsky is only able to present these rows in four of the six possible vertical arrangements. These arrangements are shown in Example 2.14.

EXAMPLE 2.14: FORMS OF ROWS B, C AND D IN THREE VOICES High: Middle: Low: mm. BP3 CP0 DP5 4-5 CP8 BP0 DP8 9-10 BP4 DP1 65 CP8 13-14 DPt CP9 BP4 16-17

Freely atonal music is interspersed throughout the movement, between each of the Row A sections and multiple-row sections (specifically, in mm. 3-4, 6, 8, 10, 12-13, 14, 16-17, 19, and 21-22). As I have previously suggested, these brief figures are based on short cycles of fourths/fifths, clearly derived from the ordered interval content of Rows B and D. 66 These gestures function as transitional signals for the listener, set apart
The first pitch of this D-row form is notated as D-natural, but the remainder of the row is clearly from the form beginning on Db. 66 These little fragments cannot easily be categorized as 5-cycles in the Headlam 1996 sense of the term, as the intervals are sometimes presented as perfect fourths/fifths and sometimes as diminished or augmented fourths/fifths, which leads to a variety of [027]s, [016]s, and [015]s. Despite differences in the specific interval content, the overall sense of quartal-ness remains throughout each of these figures.
65

54 from the surrounding music by virtue of their accented articulation. As such, the second movement juxtaposes dodecaphony, multiple-row dodecaphony, and freely atonal, cyclic elements.

MOVEMENT 3: LENTO RUBATO The third movement of Musica Stricta functions as an interlude between two predominantly contrapuntal movements. Volkonsky uses two different twelve-tone rows in this movement, but neither of them bears any strong resemblance to the rows of either surrounding movement. These rows are shown in Example 2.15.

EXAMPLE 2.15: MUSICA STRICTA (MVT. 3) ROWS E-F ROW E:

INT:

ROW F:

INT:

Volkonsky makes use of only one form of each row, EP5 and FP6, and those two forms alternate throughout the movement (one per measure). Each measure is set apart from the one which follows it by a pause, the approximate duration of which is specified each time (sehr kurz, kurz, mig lang, or lang). These pauses give the music a fragmented texture, in which each measure is heard as a complete, finished entity.

55 This movement features a similar conflict between multiple-row dodecaphony and free atonality as we saw in the second movement. However, this conflict manifests itself in a different way in the third movement. Noting that there are only four measures (mm. 14-16, 18) of non-row-based music in this movement, one might be quick to claim that the dodecaphonic material dominates the texture. However, in the dodecaphonic measures, the strictness of twelve-tone presentation is varied throughout. The first five measures are presented with few or no order-position exchanges, but beginning in m. 6, the orderly presentation of pitch classes dissolves to the point of no longer being dodecaphonic in any true sense of the word. In this way, this movement seems to blur the distinction between what Kurbatskaya calls a row and a series. A logical, though ultimately un-answerable question is this: how much order-position manipulation can one apply to a row before it is no longer recognizable as a melodic construct? While I do not assume to be able to answer this question, it seems clear that, over the course of this movement, what began as a clearly presented horizontal twelve-tone row is gradually hidden beneath the musical surface. At certain points during the movement (such as the four measures mentioned earlier in this paragraph), the row is so completely hidden that it no longer appears to exert any control over the pitch organization. The row has been transformed into a series, and the series has been dissolved into freely atonal material. Example 2.16 shows how Volkonsky sets up this dodecaphonic dissolution in m. 6 by means of contrapuntally inverting a partition scheme of FP6 which was presented in m. 2.

56 EXAMPLE 2.16: MUSICA STRICTA (MVT. 3) MM. 2, 6

In m. 2, the right hand presents FP6 order positions < 0 1 2 4 5 7 8 > while the left hand presents < 3 6 9 t e >. The two parts interlock rhythmically in such a way that, with the exception of the simultaneous presentation of 6 and 7, all the pitches occur in precisely their serial order. Measure 11 maintains this partition scheme, but in opposite voices: the right hand now presents < 3 6 9 t e > and the left hand presents { 0 1 2 4 5 7 8 }. 67 But the actual order of presentation is changed drastically. 0-3 are presented at the beginning of the measure, followed by the right hands < 6 9 t >, the left hands < 5 7 >, and the right hands 4 occurs at the same time as the left hands 7. The measure concludes with 8, played in the left hand. The consistency of partitioning between m. 2 and m. 6 allows us to understand that the series is still present, but that its original ordering is slowly being obliterated on the musical surface. Similar partition schemes can explain the serial underpinnings of the music from mm. 7-10. Measures 7 and 9 are both based on EP5, but both are missing a single pitchclass. In m. 7, D# is missing, and in m. 9, C is missing. Example 2.17 shows how the order-position patterns introduced in m. 7 are maintained in m. 9.

67

The order positions 4 5 and 7 are internally rotated.

57 EXAMPLE 2.17: MUSICA STRICTA (MVT. 3) MM. 7, 9

Despite the relative freedom of ordering demonstrated in measure 7, it is clear that the order position sets {0 1 9}, {3 5 6 7}, {t e} and {2 5} play a role in connecting these two measures. Example 2.18 shows an analogous situation with the two statements of FP6 in mm. 8 and 10.

EXAMPLE 2.18: MUSICA STRICTA (MVT. 3) MM. 8, 10

From mm. 7-20, there are only four measures that present either E- or F-row with minimal order-position manipulations. Measure 12 is a fairly clear statement of FP6, with order position 6 delayed to occur simultaneously with 9, which connects the interval

58 [11] of {4 5} in the left hand with the interval [11] of {6 9} in the right hand. Measures 11, 17 and 21 present two six-note polychords (the first of which is familiar from m. 1) as shown in Example 2.19.

EXAMPLE 2.19: MUSICA STRICTA (MVT. 3) M. 11

The sonority {23569t} occupies order positions 0-6 of the row form EP5, and is realized as an Eb-minor triad above a D-minor triad. In the first hexachord, both hands present segmental trichords from EP5, producing these two members of set class [037]. The second hexachord, however, does not present such segmental trichords. As Example 2.19 shows, the right hand presents order positions {6 7 9} while the left hand presents {8 t e}, essentially exchanging order positions 8 and 9. While the left hand still produces a [037], the right hand produces a contrasting set-class, [026]. Despite the fact that [026] is not a subset of Row E, it is a prominent subset of Row F. In this way, Volkonsky is able to use order position manipulations to create a connection between the otherwise distantly related Rows E and F.

MOVEMENT 4: ALLEGRO MARCATO The final movement of Musica Stricta is, in many ways, similar to the second movement. Both movements are primarily contrapuntal, with different rows and row

59 forms presented in an almost exclusively linear fashion, and formal divisions are marked by the juxtaposition of serial and non-serial techniques. The major difference in form is that, in the fourth movement, the freely atonal sections are expanded into large formal sections, rather than just acting as transitional signals between formal sections.

EXAMPLE 2.20: MUSICA STRICTA (MVT. 4) ROWS G-H ROW G:

INT:

ROW H:

INT:

Made up of the two new twelve-tone rows shown in Example 2.20, the fourth movement is cast in an A-B-A form similar to that of the first movement (see Example 2.21). The serial material is treated in a manner contrapuntally similar to the second movement. Beginning with GP1 in the upper voice and HPt in the lower voice, the opening 13 measures state GP1 three times, moving from the upper voice, to the lower, and back to the upper voice. In invertible counterpoint, GP1 swaps places with the Hrows, HPt, HP5, and HPt again. This is shown in Example 2.22.

60 EXAMPLE 2.21: MUSICA STRICTA (MVT. 4) FORM

A
m. 1 GP1 HPt 5 HP5 GP1 9 10 GP6 GP1 GP2 HPt 13 15 17 20 25 GI1 GRP7 HPt HP0 non-row HP7 non-row HRP6 HRP2 HP7 clusters

B
28 HPt HP9 31 34 HRP6 HRP2 HP9 HPt GP6 37 GP6 HPt HP9 42 HPt HP9 GP6 44 48 HP9 GP6 HPt

A
55 GP0 57 HP9 HPt 61 64 HRP4 non-row-based material HPt 71 [0157]

EXAMPLE 2.22: FORMS OF ROWS G AND H IN MM. 1-13 High: Low: mm. GP1 HPt 1-4 HP5 GP1 5-8 GP1 HPt 10-13 68

The B-section also features a section of triple counterpoint, similar to that of the multi-row sections of the second movement. In this case, however, the row forms do not change from statement to statement: only their vertical arrangement changes.
68

At this point, a third voice (middle) enters on GP2.

61 As in the multiple-row sections of the second movement, this section accounts for four of the six possible vertical arrangements of these three row forms. The method Volkonsky uses to present these four vertical arrangements is virtually identical to that which he used in the second movement.

EXAMPLE 2.23: TRIPLE COUNTERPOINT IN MM. 28-54 High: Middle: Low: mm.: HPt HP9 ---28-33 HP9 HPt GP6 34-36 GP6 HPt HP9 37-41 HPt HP9 GP6 42-47 HP9 GP6 HPt 48-54

As Example 2.23 shows, from the first to the second set of three row forms, one row form remains in the same voice, which the other two switch places. 69 The next two transformations involve moving the row form in the highest voice to the lowest voice, while the other two row forms move up one position.

EXAMPLE 2.24: MUSICA STRICTA (MVT. 4) MM. 19-23

69

To be precise, the stationary voice in the analogous part of the second movement was found in the lowest voice, not the middle voice as is the case here.

62 It is within the A- and A-sections that the non-row-based material occurs, toward the end of each section. In the first A-section, the end of m. 19 marks the spot at which non-row-based music begins to take control of the musical surface. As shown in Example 2.24, this material makes use of octave doublings and prominent occurrences of ic3, an interval class that does not occur in either Rows G or H. A varied restatement of this material, with right- and left-hands inverted, followed by a new, freely atonal bass melody and fortissimo statement of [0157], also forms the conclusion of the movement, and the entire piece. These measures are shown in Example 2.25.

EXAMPLE 2.25: MUSICA STRICTA (MVT. 4) MM. 67-74

Although there are no prominent usages of {2489} in this or earlier movements, [0157] is the final discrete tetrachordal subset of Row H. In fact, < 8 2 9 4 > does occur at order positions 8-e of HI7. But, as this row form has not occurred at any prior point

63 in this movement, it does not seem logical to claim that the movement ends on a segment of this row form. The use of [0157] here gives a sense of finality to the phrase without being used in a row-based context. The final fourteen measures, in effect, return the piece to the world of free atonality, while still maintaining an underlying connection to Row H. In Musica Stricta, we have seen a somewhat idiosyncratic first attempt at serial composition. Volkonsky uses not only a variety of row forms, but eight completely different twelve-tone rows, and juxtaposes them with non-serial musical material. Despite the fact that six of the eight rows used in Musica Stricta have some sort of combinatorial properties, Volkonsky never takes advantage of these possibilities. 70 Furthermore, it is only in the second movement that he seems concerned with making other smaller invariances between rows salient features of the musical surface. That Volkonsky thought he was writing an orthodox twelve-tone composition 71 reveals that his understanding of Schoenbergian/Webernian serialism was incomplete. However, his influence on his fellow students at the Moscow Conservatory, and on future generations of Soviet composers, has in some ways formed a distinctly Russian brand of serialism. Kurbatskayas twelve categories of twelve-toneness permit us to view Volkonskys varying compositional decisions through a lens which allows for a finer distinction between differing degrees of serial-ness, rather than being forced to decide whether the piece is or is not serial. While it is convenient for Western theorists to dismiss Volkonskys serial missteps as the uncertain efforts of a young composer, we
Excluding the trivial case between any row-forms Px and RPx, the following combinatorial relationships are present in the eight rows of Musica Stricta: Row A is combinatorial at APx and AI(x+1); Row B is not combinatorial; Row C is combinatorial at CPx and CRI(x+9); Row D is combinatorial at DPx and DI(x+3); Row E is not combinatorial; Row F is combinatorial at FPx and FRI(x+6); Row G is combinatorial at GPx and GI(x+1); Row H is combinatorial at HPx and HI(x+7). 71 Schmelz 2005, 171.
70

64 shall see that the case is not always so clear. In the later chapters of this dissertation, we shall return to this idea in the context of three serial works by Edison Denisov, all composed after he had acquired a reasonably full and well rounded understanding of classical (i.e., Schoenbergian) dodecaphonic techniques.

65 CHAPTER 3 ON MULTIPLE-ROW SERIALISM GENERAL ISSUES OF MULTIPLE-ROW SERIALISM Let us begin with a fundamental question: what constitutes a multiple-row scenario? While the answer to this question may seem self-evident, we shall see that it is not so straight forward, even within the works of the Second Viennese School. Though Schoenberg spoke against the use of multiple rows (see Chapter 2), he apparently rejected the idea on a theoretical level, rather than on a practical compositional level. By the time he wrote his Variations for Orchestra Op. 31, Schoenberg had evolved his principal technique of developing variation for twelve-tone pieces that is based on different partition schemes deriving internal, non-adjacent, but relatively ordered, motivic segments from a row. 1 While these derived motives have different interval patterns, they are usually presented in a row context, with a more-or-less clear derivation. However, in his book Schoenbergs Serial Odyssey, Ethan Haimo shows that several of Schoenbergs works actually use multiple rows. The String Quartet No. 3, like several of its predecessors, uses more than one set. However, unlike Opp. 26 and 29 where the subsidiary sets were generated by partitioning out elements from a succession of local statements of the principal set, here Schoenberg begins the composition with clear statements of the different sets. (Schoenbergs next composition, Von Heute auf Morgen, Op. 32, also uses more than one set.) In the present instance, Schoenberg used three sets or, more precisely, one principal set with two ordering variants. 2 Haimo speculates that Schoenbergs use of multiple rows in the third movement of the Wind Quintet Op. 36 was probably something quite mundane: having used rotation extensively in the second movement, Schoenberg felt the need for even more melodic

1 2

See Haimo 1990, 20, 22-26. Haimos term for these partition schemes is isomorphic partitioning. Ibid., 150.

66 material to provide variety. The extraction of a new twelve-tone set from the original set must have seemed like a plausible option. 3 In this last sentence, Haimo makes an important observation about the nature of these multiple rows they are extracted from the original row, and therefore may be interpreted as falling within a somewhat larger concept of the row class. Rather than developing this direction (as did his student, Alban Berg, as described below), according to Haimo, the composition of the String Quartet No. 3 and the Variations for Orchestra, however, Haimo suggests that Schoenberg learned the compositional techniques that would permit him to draw themes from the set with sufficient flexibility that he would have no need of multiple orderings. 4 In place of the use of multiple rows, then, partition schemes applied to a single row plays an important role throughout the variations of Schoenbergs Op. 31. 5 In the fifth variation, the row is decomposed into six semitone dyads, 6 which, as shown by Tuukka Ilomki, makes it impossible to indisputably decipher Schoenbergs progression of row forms. 7 This brings Ilomki to an interesting question about this particular variation: is the variation based on rows or is it based on dyads? 8 For our purposes, the answer to this question in regard to the specific case of Schoenbergs Variations is not particularly crucial. But it does provide an interesting case study of a portion of a work by Schoenberg in which it is unclear how the row relates to the musical surface. Certainly, many such passages can be found in Schoenbergs later works.

Ibid., 117. Ibid., 150-151. Haimos terminology here is a bit different than mine. When he refers to multiple orderings he is referring to multiple orderings of the total chromatic, or multiple twelve-tone rows. 5 Ibid., 169, 171, 175-176, 178-180. 6 Ilomki 2007, 31. 7 Ilomki demonstrates that the same set of ordered semitone dyads can always be drawn from at least two row forms. See Ilomki 2007, 34. 8 Ilomki 2007, 41.
3 4

67 In Bergs music, row partitions are pulled out and used independently of the original row context, and newly derived twelve-tone rows appear. One major issue in Berg scholarship is the nature of the derivations of these new rows from a principal or main row. 9 It is well known that the three secondary rows used in the Lyric Suite are derived successively by order position exchanges from the preceding row: the allinterval row used in the first movement generates a second row by exchanging the pitch classes at two particular order positions, that new second row generates a third row in a similar fashion, and the third row generates a fourth row in a similar (but somewhat more complex) manner. 10 The derivational pathway from the first to the fourth row is clear, but the INTs of the first and fourth rows bear little resemblance to each other. 11 This process begs the following question: given a parent row and its derived offspring, how dissimilar can the two be before the relationship between them is irrelevant? We can use definitions of degrees of similarity between two rows to consider this question. Ilomkis 2008 dissertation presents a number of different ways of measuring similarity between distinct twelve-tone rows. 12 These different similarity measures highlight different properties of the rows themselves: 1) similarity may be based on the coincidence of specific pitch classes at or near specific order positions in two rows; 2) similarity may be based on shared subsegments between two rows; 3) similarity may be based on the ordered series of intervals of two rows, and so forth. While these similarity measures all enable the analyst to assign a discrete value representing how similar two
Reich 1936, Perle 1959, Jarman 1979, Hall 1985 and Headlam 1985, 1996 all contribute to a discussion of the relative independence and role of these multiple rows. 10 Headlam 1996, 248. 11 This is even more important in Lulu, where a number of rows are derived from a single Basic Row but are used independently of each other, representing individual characters in Leitmotif fashion. See Headlam 1985. 12 I am using the word similarity in a slightly informal manner. To be sure, some of the so-called similarity measures found in the music theoretic literature actually measure dissimilarity. See Buchler 1997, 31 and Ilomki 2008, 35-36.
9

68 rows are, they do not provide an answer to the question in the previous paragraph. In this context, aspects of Ilomkis impressive work will be discussed in more detail later in this chapter. In the end, it seems impossible (or at least difficult) to provide a definitive answer to the question of multiple rows in any sort of general way. While we may be willing to use the label multiple-row piece to describe pieces like Bergs Lyric Suite and Volkonskys Musica Stricta, it seems clear that the nature of the multiple rows in each of these pieces is substantially different from the other. Whether the pitch fabric of a piece is taken from two distinct rows or from two rows, one of which is derived from the other, the characteristics of multiple-row pieces therefore must be addressed on a caseby-case basis. In beginning to analyze a piece which takes its pitch material from more than one twelve-tone row, it is necessary to examine the relationship(s) between the rows themselves. In the course of studying these relationships, the analyst is confronted with three broad categories of issues specific to multiple-row serialism: issues of structure, function and unity.

ISSUES OF STRUCTURE Structural issues of multiple-row serialism involve relationships between the rows themselves, removed from the context of the specific composition. Do the two (or more) different rows share any common properties, such as interval successions, interval-class content, and so forth? I have already mentioned Ilomkis work on the similarity of twelve-tone rows, but it is worth examining certain aspects of it more closely here.

69 In his dissertation, Ilomki presents seventeen different similarity measures for relating any of the 12! twelve-tone rows. 13 In so doing, he identifies five different conceptions of twelve-tone rows, and argues convincingly that each of these conceptions brings some aspect to the fore. 14 Ilomkis five conceptions are the following: the twelve-tone row as a vector, as a set of ordered pairs (which is, of course, influenced by Babbitt), as a set of subsegments, as a set of subsets, and as a sequence of pitch-class intervals. Of these five, the last appears to be the most useful for the purposes of the present study of music based on multiple row classes. Simply put, if we are interested in determining what the abstract relationship is between two distinct row classes, it seems most appropriate to consider the elements of a twelve-tone row class that remain invariant (or are varied in a predictable manner) among all members of that row class. 15 When comparing two (or more) distinct row classes, we should be most interested in a method of comparison that allows us to make observations about the entire row class, rather than making a comparison between a single member of each row class, and claiming that comparison holds true for all members of each class. 16 As the present dissertation defines a row class as a set of up to forty-eight twelve-tone rows

Despite the fact that I have been consistently using the word row to mean row class in this dissertation, I will not follow that convention in the paragraphs discussing Ilomkis work. As Ilomkis goal is to measure the similarity between any two ordered sequences of twelve non-repeating pitch classes, his notion of row refers to any one of the 12! orderings of the twelve pitch classes. See Ilomki 2008, 67, and Babbitt 1960, 248. 14 Ilomki 2008, 92. 15 Otherwise, we find ourselves in the potentially awkward position of being forced to find a single row form that can be used as a representative for the entire row class. Of course, this is the case with the most common approaches to pc-set theory. The list of prime forms found as an appendix in Forte 1973 (among a myriad of other sources) does exactly this: it uses a particular member of a given set class as the representative of the entire set class. It is this member which is often used as the name of the set class, unless the Forte names 3-1, 3-2, etc. are used. It is certainly possible to decide upon such a representative member of any given row class in a particular composition, in a variety of ways. We could choose the first member of that row class used, or the most commonly used member, and so forth. 16 Ilomki addresses this concern briefly when discussing the spaces formed by different conceptions of twelve-tone rows. His solution is to define the distance between any two finite sets of objects (row classes, in this case) as the distance between their closest members. See Ilomki 2008, 76-77.
13

70 related by the canonical twelve-tone operations (transposition, inversion, retrogression, and the combination of those operations), the INT is the only element of all members of a row class which is varied in a predictable manner for each member, and can be used to distinguish any row class X from any row class Y. Ilomki rightly notes that conceiving of a twelve-tone row as a succession of ordered pitch-class intervals means that we abandon the permutational approach 17 pioneered by Babbitt, which has been fundamental in twelve-tone theory. While I do not wish to suggest that a permutational perspective is not valid in multiple-row music, it is my belief that viewing a row as an ordered succession of intervals allows the analyst the ability to conceptualize all members of a given row class in an easier, more intuitive manner, thus making this approach useful in distinguishing between multiple row classes. As such, when examining the structural relationship between two (or more) twelve-tone rows in a particular piece, we will focus primarily on the similarities between the ordered succession of intervals and interval classes of the rows in question. Later in this chapter, I will define one specific analytical tool which will aid us in this kind of comparison.

ISSUES OF FUNCTION Functional issues of multiple-row serialism involve questions of how the different rows are used within a particular piece. Regarding a multiple-row piece like Bergs Lulu, the function of the different rows seems clear: the rows function in a Leitmotif-like manner, with each row being linked to a particular character. 18 In this case, the rows serve a specific dramatic function. But for pieces of absolute music, the
17 18

Ilomki 2008, 230. Headlam 1985, 198.

71 function of multiple rows is less clear. It is necessary to examine how the rows themselves interact during the piece. Are they presented simultaneously? Are they juxtaposed? Does one row seem to be primary and the other(s) serve as contrasting material, perhaps analogous to traditional themes in tonal music? A consideration of Soviet composers is especially relevant in this context. In a 2004 essay, Schmelz notes that [o]nly in Russia was there a pervasive tendency to construct pieces with multiple twelve-tone rows that in fact obeyed none of the traditional Schoenbergian laws of twelve-tone music (or if they did, did so only selectively). 19 This does not mean that Soviet serial music was entirely unorthodox, however. In Chapter 2, we saw that Volkonskys serial language involved the use of multiple rows and the combination of dodecaphonic music with freely atonal material. But the specific dodecaphonic techniques he applied were not so far removed from those of Schoenberg: the generation of a large amount of pitch material from the transformation of these rows, the use of partition schemes, and so forth. Volkonskys early serial efforts were profoundly influential to his peers at the Moscow Conservatory: Schnittke and Denisov, in particular. As Denisov later recalled, [p]recisely these compositions [Musica Stricta and Suite of Mirrors, another of Volkonskys early twelve-tone works] were the ones that exerted a very great influence on us. On all of us. 20 In fact, it was not only the young generation of Moscow composers in the 1960s who felt Volkonskys influence. His influence was felt, at least indirectly, by Dmitri Shostakovich. Schmelz discusses Shostkovichs very personal 21 implementation of serial techniques in the late 1960s, and shows their connections to the serial techniques
Schmelz 2004, 326. Though Schmelzs comments here refer most directly to the technique of twelvetone rows, they are still relevant in the present context of Soviet serialism in general. 20 Quoted in Schmelz 2009, 67. 21 Schmelz 2004, 305.
19

72 used earlier by the younger generation of Soviet composers: Prt, Schnittke, Volkonsky and Denisov. Schmelz points out that Shostakovichs use of twelve-note rows in his Seven Verses of Aleksandr Blok Op. 127 (1967) demonstrates a use of a single tone row as a melody. In Schmelzs words, the row retains its identity as a single, unchanging melodic figure, 22 and states that the use of the row (which occurs, untransposed and unvaried, in only one of the seven movements of Op. 127) functions as a catalyst of harmonic instability and atonality, 23 consistent with the unclear harmonic language of the work [which] reflects the apocalyptic text of the poem, which presents images that are part nocturnal hallucination, [and] part ecstatic vision 24 In Shostakovichs Sonata for Violin Op. 134 (1968), String Quartet No. 12 Op. 133 (1968) and Symphony No. 14 Op. 135 (1969), the composer utilizes multiple twelve-tone rows to create an effect of long-term shifting instability. 25 But in all these cases, Shostakovichs row usage falls more into Kurbatskayas category of technique of twelve-tone rows and not the dodecaphony found in the works of the Second Viennese composers. The rows are occasionally repeated in slightly varied form, but in general, they each have a fundamental rhythmic profile, as well as a common melodic contour. In short, these rows are used mostly as melodic or motivic constructions. 26 This is markedly different from Ernst Kreneks comments that the row not be used as a theme, but it is fully consistent with Kurbatskayas definition of the technique of twelve-tone rows. 27

Ibid., 306-307. Ibid., 307. 24 Ibid., 306. 25 Ibid., 309. 26 Ibid. 27 Krenek 1940, 3. Kreneks treatise is one to which Arvo Prt (in Tallinn, Estonia) evidently had access, as well as a circle of student composers in Kiev. See Schmelz 2009, 51.
22 23

73 In Chapter 2, we also saw multiple rows used for a different functional purpose. The second movement of Musica Stricta uses four distinct rows in a form-generative manner. To summarize, the movements form evolves from the juxtaposition of a single, primary row with a complex of three secondary rows. Certainly, the question of whether multiple rows function in a piece primarily as contrasting themes or as generators of form is not always a clear either/or situation. It is not difficult to imagine a piece of music in which contrasting themes serve the purpose of also defining large-scale form. The sonata form-like structure of Denisovs Concerto for Guitar and Orchestra (1991) provides a representative example. Example 3.1 shows a form chart of this single-movement concerto, taken from Elena Baraschs article The Concertos of Edison Denisov. 28 The Primary and Secondary Rows of the Concerto are shown in Example 3.2.

EXAMPLE 3.1: CONCERTO FOR GUITAR AND ORCHESTRA FORM Exposition

m. 1 primary theme Primary Row

38 secondary theme

68 primary theme

Development I

Reprise

Development II

Coda

85 Secondary Row
28

193 238 Primary Row

283

This essay actually contains two form charts for the Concerto: one with only the largest structural divisions indicated, and a more detailed one, with smaller formal divisions indicated as well. The diagram presented here is a combination of the two: it is essentially the larger-scale form chart, but with a few of the more specific details of the second chart (measure numbers, and smaller divisions in the exposition, in particular). See Barasch 1999, 361 and 366. The indications Primary Row and Secondary Row are not Baraschs terms, but the expositional divisions primary theme and secondary theme are hers (glavnaya partiya and pobochnaya partiya, respectively).

74 EXAMPLE 3.2: CONCERTO FOR GUITAR AND ORCHESTRA ROWS PRIMARY ROW

SECONDARY ROW

It must first be noted that neither the Primary nor the Secondary Row accounts for the entire pitch fabric; there are significant portions of the Concerto which do not appear to be drawn directly from either row. As such, the presence of the words Primary Row or Secondary Row in Example 3.1 should be taken as summarizing the material of the section in question, rather than being a specific indication of exactly where a particular row is found. The Primary Row is presented at the Concertos opening, as the first statement by the solo guitar. This row recurs several times during the Concerto, but is treated primarily as a static melody: rhythmically fixed, and never inverted or presented in retrograde (but it is transposed at least once, at m. 136). The final statement of the Primary Row is shown as Example 3.3.

EXAMPLE 3.3: CONCERTO FOR GUITAR AND ORCHESTRA, GUITAR, MM. 349-352

75 The Secondary Row forms the bulk of the pitch material of the Development I section, where a variety of different forms of the row are used. The opening statement of the Secondary Row, divided among twelve different instruments, is shown as Example 3.4.

EXAMPLE 3.4: CONCERTO FOR GUITAR AND ORCHESTRA, BPt

MM. 130-132

These two rows allow us the opportunity to view a potential type of interaction between the two functional categories identified earlier. The Primary Row functions

76 largely as a recurring melody, as a sort of ritornello which bookends the entire Concerto, and even invades the presentation of the Secondary Rows material. The Secondary Row, on the other hand, is not presented as a theme, per se, but rather, its presence indicates the beginning of a new formal section. Often, structural and functional issues of multiple-row serialism go hand in hand. In Volkonsky and Denisovs multiple-row serialism, when two rows function as contrasting themes in a work, those rows are usually differentiated by having different intervallic characteristics. The same might be true of two rows used in a form-generative manner, as we will see in the analysis of Denisovs Octet for Winds in Chapter 6.

ISSUES OF UNITY Finally, the issues of unity deal with the way in which a composer does (or does not) achieve a unified musical structure when that structure is based upon a variety of generative sources. Given that different rows have different characteristics (that may or may not be measurable in the sense described by Ilomki), does the use of different rows destroy the unity of a composition? Though the answer to this question might vary from composer to composer, or even from composition to composition, it seems that, as evidenced by his quotations found in Chapter 2, Schoenbergs answer to this question would likely be a resounding yes. As it turns out, Denisov himself appears to have had varying thoughts on the topic at different points during his career. In Denisovs article Dodecaphony and the Problems of Modern Compositional Techniques, the composer writes the following: Sometime in addressing dodecaphonic techniques, the following question arises: why is the composer limited to use only one series in a work, rather than using several different series? First, the presence in a work of a new series does not enrich it as all possible sonic combinations can be

77 obtained by the application of the same series. Secondly, the presence of two or more series introduces a certain stylistic differentiation into the work, disrupting thestrength of a single unified thematic basis [however, we see examples of the simultaneous use of several series (A. Berg, I. Stravinsky)]. Usually, even when creating large works, composers are limited to one series, thus giving some monothematicism to the cycle, and if they apply several different series (such as A. Bergs Lulu), they prefer to use so-called derived series, where the series is brought out in some way from the original. 29 Though published in 1969, this essay was written in 1963, while Denisov was in the relatively early phases of his second conservatory. As mentioned in Chapter 1, the serial works Denisov composed prior to 1964 are much closer to Schoenbergian ideals than are his later works. At some point between this 1963 article and Denisovs 1966 Laments and Five Stories of Mr. Keuner (both of which use multiple twelve-tone rows), the composer seems to have made an about-face on the disruption inherent in multiplerow serial music. As we shall see in the following chapters, the rows in Denisovs multiple-row pieces typically have sharply contrasting intervallic material. In the third movement of the Sonata for Alto Saxophone and Piano, for instance, the primary row 30 is based on T7 transpositions of Dmitri Shostakovichs D-S-C-H motto, < 2 3 0 e > and < 9 t 7 6 >, while the secondary row is almost entirely transformations of set-class [037]. In the second movement of the Octet for Winds, Row A uses mostly ic1 and ic2, while Row B is made up of the palindromic INT < 6 5 6 e 6 e 6 e 6 5 6 >, which forms four transformations of [016]. As shown in Example 3.2, in the Concerto for Guitar and Orchestra, the Primary Row is derived from transformations of [026], while the Secondary Row has a much less uniform INT.

Denisov 1969, 518. I do consider the first row of this movement to be primary, because it, by itself, formed the basis of the pitch material from the first movement as well.
29 30

78 In all three cases, we see a similar pattern: one row is of highly regular intervallic construction (in all cases, these regular rows are made up of only three interval classes), and the other row has a wider range of interval (or interval-class) content. There is a sense in which these rows seem to be paired in such a way that they complement each other. In the passage quoted earlier, Denisov says, the presence of two or more series introduces a certain stylistic differentiation into the work. 31 Perhaps Denisov determined that this thought did not have to be viewed in the negative context in which it is presented in his essay. This kind of stylistic differentiation does not necessarily indicate a less unified musical structure.

*** INTS, INT-CS, AND BIPS The remainder of this chapter will focus on the introduction of a particular means of examining the structural relationship between multiple rows. The third movement of Denisovs Sonata for Alto Saxophone and Piano derives much of its pitch material from transformations of two distinct twelve-tone rows. Example 3.5 shows the row from the first movement. As previously noted, this row (Row A), with its frequently occurring ics 1 and 2, begins with Shostakovichs famous D-S-C-H (D-Eb-C-B) motto. The second hexachord of the row begins with the same motto at T7. In the third movement, this row is joined by a second twelve-tone row, featuring triadic and seventhchord subsets. This row (Row B) is shown as Example 3.6.

31

Denisov 1969, 518.

79 EXAMPLE 3.5: SONATA FOR SAXOPHONE, ROW A

BIP: 11111122335 BIP-Vector: [622010]

EXAMPLE 3.6: SONATA FOR SAXOPHONE, ROW B

BIP: 33333344455 BIP-Vector: [006320] In Examples 3.5 and 3.6, each rows INT is given beneath the row. Below the INT for each row, I have added an INT-C, which is the ordered succession of interval classes in the row. While the INT-C is one level of abstraction beyond the INT of a row, it offers the advantage of representing both the transposed and inverted forms of the row simultaneously. As is well known, given an INT for row form x, one can produce the INT of a row form related by inversion to x by substituting each entry of xs INT with its mod 12 complement. Because each interval and its mod 12 complement are members of the same interval class, it should be obvious that the INTC of row form x is identical to the INT-C of all row forms related to x by transposition and inversion. Furthermore, reversing the order of all entries of row form xs INT-C yields the INT-C for all row forms related to x by inversion and retrograde inversion. However, it is important to note that the INT-C carries the potential pitfall of not necessarily defining a single row class. While an INT can unambiguously define a row

80 class, it is possible for two distinct row classes to share an INT-C. 32 Despite this possibility, the existence of non-unique INT-Cs does not concern us presently. The information in the INT-Cs of Examples 3.5 and 3.6 is also displayed in unordered fashion by using Allen Fortes BIPs, or basic interval patterns. 33 The BIPVectors in these examples are not something Forte explicitly defines, but their construction simply displays the information of each BIP in the same format as Fortes interval-class vectors. 34 A quick inspection of the two BIP-Vectors shows little in common between the two rows, in terms of unordered interval-class content. However, even though the interval classes contained in these two rows are mostly different, their relative distributions are not so different; each BIP-Vector has one 6, and at least one 2. One might also entertain the possibility of understanding the 3 in Row Bs BIP-Vector as having been broken into 2+1 in Row As vector. If we consider different ways of describing the interval content of that row, it is easy to see the decreasing level of specificity from INT to INT-C to BIP. An INT-C and a BIP show the same information, but respectively with or without specifying the order of the interval-classes of a row. If we desire to fill out our imaginary continuum, we might consider investigating partial orderings of the interval or interval-class content of a row. Looking at the INT-C of Row A, we can list that rows interval-class successions as protocol pairs, following the convention established in Lewin (1976) 35 :

32 This property defines a twelve-tone analog to the Z-related sets of pc set theory. Where Z-related sets are defined as those that share a common interval-class vector, but are not members of the same set class, these Z-related rows share a common INT-C, but are not members of the same row class. An examination of the properties of these Z-related rows could be an interesting topic for future research. 33 Forte 1973, 63-73. 34 This idea is adopted in Ilomki 2008, 235. 35 Musical applications of partial orderings are also discussed in Starr 1980, Morris 1987 and 2001, and Ilomki 2008.

81 Row A = {(1,1), (1,2), (1,3), (2,1), (2,5), (3,1), (5,1)} While this set of protocol pairs accounts for every pair of successive interval classes in Row A, it is clear that this set does not constitute a true partial ordering. In a brief but thorough explanation of the mathematical concept of partially ordered sets, Ilomki defines a partial ordering as a binary relation that is reflexive, transitive and antisymmetric. 36 In verbal form, the protocol pair (1,2) states that ic1 precedes ic2. The set of protocol pairs that makes up Row A includes the contradictory pairs (1,2) and (2,1), which violates the requirement of anti-symmetry: if ic1 precedes ic2, then ic2 does not precede ic1. 37

EXAMPLE 3.7: BERG, LYRIC SUITE (PRIMARY ROW)

{(2,3), (3,4), (4,1), (5,2), (6,5), (7,6), (8,9), (9,t), (t,7), (e,8)}

All-interval rows provide the only occasion in which such an account of all of a rows interval successions in a row can result in a proper partial ordering. The INT of the original row from Bergs Lyric Suite provides a suitable example (see Example 3.7). In this case, the set of protocol pairs is completely anti-symmetric; there are no contradictions. This set of protocol pairs can be considered a true partial ordering because all the intervals are distinct. It should be clear without example that, considered
Ilomki 2008, 284. Lewin 1987 defines a precedence-oriented system as one in which there is no pair of nodes (N, N) such that N both precedes and follows N. See Lewin 1987, 210. Clearly, this is not possible with twelvetone rows examined in terms of interval class successions; there are 11 positions in an INT-C, but they must be filled with only 6 interval classes.
36 37

82 in terms of interval classes, no twelve-tone row (including the Lyric Suite row, and any other all-interval row) can be described completely by partial orderings.

THE ADJACENCY MATRIX AND INTERVAL-CLASS SUCCESSION GRAPHS The inability of interval-class successions to produce mathematically correct partial orderings does not necessarily indicate that a similar concept might not be useful. Example 3.8 uses an adjacency matrix to show the succession of adjacent interval classes from the Sonata for Alto Saxophones Row A.

EXAMPLE 3.8: ADJACENCY MATRIX FOR SONATA FOR SAXOPHONE, ROW A 1 1 2 3 5 1 1 1 1 2 1 0 0 0 3 1 0 0 0 5 0 1 0 0

The matrix contains the interval classes of the row as both column and row headers. Every cell in the interior of the matrix is filled either with a 1 or 0, depending on whether a particular interval-class succession is present in the row. If the intervalclass succession (i,j ) occurs in the row then a 1 is placed in the cell in the i th row and j th column. For example, the matrix tells us that ic1 is followed by ics 1, 2 and 3 in Row A,

83 but never by ic5. 38 The adjacency matrix provides a convenient way of accounting for the interval-class successions of a twelve-tone row class. At this point, it is worth reminding the reader that the adjacency matrix has been filled based on the presence or absence of a particular interval-class succession in a rows INT-C. This means that the matrix only accounts for the interval-class succession of the transposed and inverted forms of the row. In order to account for the retrograde and retrograde-inverted forms of the row, one must simply swap all 0s for 1s and 1s for 0s. This does not create any serious problems, as the INT-C will refer to the same row class (or classes) whether it is read forward or backward. Looking at the adjacency matrix in Example 3.8, we can easily see that ic2 is followed by either ic1 or ic5, but it is always preceded by ic1. In the matrixs row (2,j ), 1s are found in the cells (2,1) and (2,5), but not (2,3). Likewise, in the matrixs column (i,2), the cell (1,2) is the only cell that contains a 1. Example 3.9 shows a visual representation of these relationships. EXAMPLE 3.9: SONATA FOR SAXOPHONE, ROW A, SUCCESSIONS OF IC2

ic2

ic1

ic5

In this example, a graph has been drawn to show the information about the matrixs row (2,j) and column (i,2). As the somewhat clumsy prose of the preceding paragraph demonstrates, the information in an adjacency matrix might be clear, but difficult to
38 One could easily imagine a different kind of adjacency matrix, which uses not just the numbers 0 and 1, but actually fills the cell in the i th row and j th column with the actual number of occurrences of intervalclass succession (i,j ) in the row.

84 communicate efficiently in verbal form. The graph of Example 3.9 provides a more visually attractive alternative. 39 Although the meaning of this graph is intuitively clear, it is necessary to consider graph terminology from a somewhat more formal perspective. Appendix 2 provides an introduction to the basic vocabulary of graph theory and demonstrates how that vocabulary will be used in the present dissertation. I call the kind of graph depicted in Example 3.9 an interval-class succession graph, or ICSG. In this graph, each node is labeled with one of the interval classes from the rows INT-C, and the directed arrows show the possible interval classes that can occur before or after ic2. Every ICSG is a connected graph. 40 Clearly, the graph in Example 3.9 is a preliminary step toward a complete visualization of all the interval-class successions of Row A from the Sonata for Alto Saxophone. It does not tell us all the relationships between ics 1, 2 and 5, and it does not even mention ic3. However, this information could prove to be useful, as it is a subgraph of the full graph used to represent the entire set of interval-class successions of Row A all of the relationships depicted in Example 3.9 will be present in the full graph, but there will be additional nodes and arrows as well. As we shall see, subgraphs such as this one prove to be of special interest. Example 3.10 shows the complete graph for Row A. This ICSG demonstrates nearly the full range of interval-class successions in Row A. However, it will not go unnoticed that there are a handful of successions which may appear to be missing from this graph, which I will address presently. First, in the interest of maintaining visual

On this topic, Robert Morris 2007 has noted that presenting musical structures by the means of graphs often has an intuitive appeal that verbal or mathematical notation does not. 40 See Appendix 2 for the definition of connected graph and other graph theoretic terms.
39

85 EXAMPLE 3.10: SONATA FOR ALTO SAXOPHONE, ROW A, ICSG

ic1 ic2 ic3 ic5

clarity, I have not drawn an arrow from ic1 to itself, despite the fact that the rows INT has two semitones in a row. While this looped arrow could certainly be added to the graph, I have chosen only to show successions of different interval classes. 41 Second, I have not taken into account the so-called cyclic interval or wrap-around interval of the row: the interval which occurs between the last pitch-class of the row and the first, if the row were presented in a looped or rotated form. Finally, we must be aware that the arrows only indicate the interval-class successions in transposed and inverted forms of the row. The directions of the arrows must be switched to accurately account for retrograde and retrograde-inverted row forms. With an understanding of these apparent omissions, ICSGs are useful to gain a quick understanding of the interval-class structure of a row. Furthermore, they can also be used to show structural relationships between rows that might otherwise go unnoticed.

COMPARISON AND CATEGORIZATION OF ICSGS There are several ways of comparing ICSGs. Example 3.11 shows the ICSGs for rows used by Webern in three of his early twelve-tone pieces: the rows from his Op.

41 In a similar manner, my ICSGs do not show multiple arrows connecting the same pair of nodes. While ic1 is followed by ic2 twice in Row A, I have not drawn two separate arrows from ic1 to ic2. A graph which contains such multiple arrows is called a multigraph and a graph which contains loops is called a pseudograph. See Harary 1972, 10.

86 17/iii, Op. 18/iii, and Op. 21. Clearly, these three rows produce identical ICSGs. Identity is, of course, the closest possible relationship between two graphs. 42

EXAMPLE 3.11: THREE WEBERN ROWS AND THEIR ICSGS WEBERN, OP. 17/III
1 3 4 6

WEBERN, OP. 18/III

1 3 4 6

WEBERN, SYMPHONIE, OP. 21

1 3 4 6

When dealing with non-identical graphs, the concept of isomorphic graphs is useful. In terms of the underlying structure of each graph, the contents of each node of an ICSG are largely irrelevant. To compare different graphs abstractly, we are interested primarily in the relationships between the nodes: which nodes are connected by arrows, and which direction or directions the arrows point. Frank Harary, author of what is perhaps the classic graph theory text, states that two graphs G and H are isomorphicif there exists a one-to-one correspondence between their [node] sets

42 While it may seem trivial, it is important to note that these graphs are identical rather than the same graph. Being derived from three different musical situations, the three graphs of Example 3.11 are three distinct graphs with the same structure of nodes and arrows and the same node labels.

87 which preserves adjacency. 43 Such isomorphisms form a second equivalence relation on graphs.

EXAMPLE 3.12: WEBERN, KINDERSTCK (1924) ROW AND ICSG

1 3 4 5

Example 3.12 shows the row and ICSG for Weberns earliest serial composition, the unpublished Kinderstck from 1924. Because the labels of the nodes are not the same, we cannot say that the Kinderstck rows ICSG is identical to any of the ICSGs of the rows in Example 3.11. However, this ICSG is isomorphic to all of the ICSGs of the rows from Op. 17/iii, Op. 18/iii, and Op. 21. In this particular case, the visual representation of the ICSGs makes this isomorphism easy to see: the only difference between the graphs of Example 3.11 and Example 3.12 is that the node on the far right has as its label a 6 in the former case and a 5 in the latter. But changing the visual layout of a graph does not change the structure of the graph. Example 3.13 shows the Kinderstck rows ICSG presented in three different but equivalent manners. Both the representations in A and B show an isomorphic relationship with the graphs of Example 3.11, despite the rearrangement of nodes 3, 4, and 5. The relationship between the graphs of Example 3.11 and the representation shown at letter C is somewhat less obvious, but the isomorphism can be seen by considering the

43

Harary 1972, 10.

88 EXAMPLE 3.13: DIFFERENT REPRESENTATIONS OF KINDERSTCK ICSG A.

1 3 4 5

B.
1 4 5 3

C.
3 5 1 4

indegree and outdegree 44 of each of the graphs nodes. Even with the nodes shuffled in this way, the isomorphism is fairly easy to discern. For the purposes of the graphs in this dissertation, I will use either a verbal description of the isomorphism when it is not immediately apparent, or I will re-draw one of the graphs in a way to make the isomorphism visually clear. The idea of graph isomorphisms is not new to music theorists; it is probably most familiar from the writings of David Lewin, dealing with transformation graphs and Klumpenhouwer networks, in particular. 45 When considering a given system of nodes and arrows, Lewin draws a distinction between networks and graphs, 46 based on whether or not the underlying node/arrow system 47 is labeled or not; a network is a labeled graph. Although I have not been using his terminology explicitly, it should be clear that I am invoking the same distinction when I consider isomorphic graphs. 48

See Appendix 2. See, especially, Lewin 1987, Chapter 9, and Lewin 1990. 46 In this paragraph, I will enclose Lewins terms network and graph in quotation marks in order to avoid confusion with my slightly different definition of graph. 47 Lewin 1987, 193. 48 For the formalism behind Lewins network and graph distinctions, see Lewin 1987, 195-196. In graph-theoretic terms, a network is a graph (with or without labeled nodes) that is associated with a function that provides a numerical value to each of the arrows in the graph. See Harary 1972, 50. Clearly, this is the case for Lewins transformational networks (as well as Klumpenhouwer networks, see Lewin 1990).
44 45

89 Beyond identity and graph isomorphism, the rows used in the three pieces from Weberns Op. 17 provide an example of a third kind of relationship between ICSGs. Example 3.14 shows the three rows with their ICSGs.

EXAMPLE 3.14: WEBERN, DREI VOLKSTEXTE (OP. 17) ROWS AND ICSGS A. OP. 17/I
1 2 3 4 5

B. OP. 17/II
1 3 4 5 6

C. OP. 17/III
1 3 4 6

These three ICSGs are not identical, and they cannot be considered isomorphic because they do not share a common number of nodes and arrows. But, looking at the structure of each graph carefully reveals that the ICSG in Example 3.14C is a subgraph of the ICSG in Example 3.14B: all the nodes and arrows of Op. 17/iiis ICSG are present in Op. 17/iis ICSG, but the latter contains additional nodes and arrows. Conversely, the ICSG in Example 3.14B is a supergraph of the ICSG in Example 3.14C. There are two types of subgraph relationships: abstract and literal. If we consider these three graphs without their respective node labels, we can see that the
As none of the graphs in the present dissertation involve labeled arrows, I will use the word graph for all cases, specifying whether the graph is labeled, or unlabeled (or abstract) if necessary. See Appendix 2.

90 ICSG in Example 3.10C is an abstract subgraph of the ICSG in Example 3.14A. Because the relationship described in the previous paragraph is true based not only on the relationships between nodes and arrows, but also based on the specific label assigned to each node, that relationships is a literal subgraph relationship. 49 Two of the three graphs in Example 3.14 also provide an example of a particular kind of graph: the star graph. A star graph is a graph in which exactly one node is adjacent to or from all other nodes, but none of the other nodes are adjacent to or from each other. In other words, a star graph contains a single nexus node which is the only node to or from which all remaining nodes are adjacent. The ICSGs from Example 3.14B and 3.14C are both star graphs. 50 This definition allows us to make a similar, but more specific, observation to the type made by Kathryn Bailey about Weberns general intervallic preferences in his earliest twelve-tone works (Op. 17, Op. 18 and the unpublished Klavierstck and Kinderstck). Bailey discusses the abundance of ic1 in these early rows, while stating that many of the later rows rely less on semitones between adjacent notes as symmetry and invariance become the significant concerns. 51 A quick glance at the INT-Cs of

It is worth noting that Kathryn Bailey notes that the first of the three Op. 17 Lieder consists of twelvenote fields rather than linear statements (Bailey 1991, 14), and that it is not really constructed from a row. It presents a succession of twelve-note fields (Bailey 1991, 33). Interestingly, even though the row shown in Example 3.14A (which is the row Bailey assigns to this movement) only appears on the musical surface in a more or less strict ordering during mm. 1-3, the majority of the vocal line forms a non-serial ICSG which is, itself, isomorphic to the ICSG shown in Example 3.14A. The only portion of the vocal line of this piece which does not form such an isomorphic ICSG is at the pieces climax, mm. 10-12. 50 This definition is based on that which is found in Harary 1972, 17, and Wilson 1975, 18. Both authors define a star graph as a graph in which the nodes can be divided into two non-intersecting sets, V1 and V2, in such a way that every edge in the graph joins a node from V1 to a node in V2. This is called a bipartite graph. A special type of bipartite graph exists when every node in V1 is adjacent to every node in V2 (a complete bipartite graph). The definition of a star graph, then, is a complete bipartite graph in which either V1 or V2 is of cardinality 1. The examples provided by Harary and Wilson reveal that both authors are referring to a star graph as a type of (non-directed) ordinary graph. Because my definition of star graph applies to directed graphs, it might be more accurate, if less clear, to say that a star (di)graph is a directed graph whose underlying ordinary graph is a star graph. 51 Bailey 1991, 13-15.
49

91 Weberns rows supports this observation, but the use of ICSG technology allows us to put his early use of ic1 into a larger context. Rather than simply counting the number of ic1s in a particular row, we can use the ICSG to examine how these interval classes are used with the other interval classes of the row. By noticing that the ICSGs for the rows used in Op. 17/ii and iii are both star graphs, and that the node labeled ic1 is, in both cases, the nexus node, it becomes clear that, in these rows, Webern uses this abundance of ic1s in a very specific way. In addition to these two rows, the rows from Op. 18/iii, the Klavierstck, and the Kinderstck all form ICSGs which are star graphs, and they all have ic1 as the nexus node. This adds a deeper level of understanding to Baileys note about the preference for ic1 between adjacent pitch classes in Weberns early rows. Bailey also implies that sequential intervallic construction plays a significant role in these early twelve-tone works. 52 The preponderance of a single type of ICSG supports this claim. While star graphs are not found in many of the ICSGs for Weberns later dodecaphonic works, the ICSGs of the rows for the Op. 20 String Trio and the Op. 21 Symphony are both also star graphs. So, where Bailey sees Op. 20 as the culmination of these early experiments, 53 the analytical use of ICSGs allows us to see a point of coincidence and an additional thread of continuity between these early rows and Weberns later twelve-tone rows.

***

In this chapter, I have attempted to engage three general issues of multiple-row serialism. To begin, we have seen some of the ways in which issues of structure,
52 53

Ibid., 16. Ibid.

92 function and unity play a role in the understanding of compositions which use more than one single row as the foundation of their pitch material. This led to a discussion of some of the ways one might choose to explore the relationships between the rows of a piece, and it provided a new analytical device, the interval-class succession graph, which can be used to this end. The chapters to follow make up the second large part of this study: the analytical discussion of three pieces by Edison Denisov. First, we will examine this composers use of ICSG-equivalent and ICSG-isomorphic row in the first of his Five Etudes for Bassoon. Second, we will examine the choice of rows with notably different ICSGs in his Sonata for Alto Saxophone and Piano and his Octet for Winds.

93 CHAPTER 4 ANALYSIS: FIVE ETUDES FOR SOLO BASSOON (NO. 1) In this chapter, I will examine Denisovs use of multiple rows in the first of his Five Etudes for Solo Bassoon. 1 I will begin by making some general observations about the pieces composition, form, and row content. The analysis to follow will then revisit the three broad categories of multiple row serialism discussed in Chapter 3. First, I will examine the structural relationships between the ten rows which appear in this piece, using an ICSG-based method to describe the similarity in unordered interval-class content among these rows. Second, I will discuss the functional use of the rows, focusing particularly on row-chaining procedures, which Denisov refers to as serial modulation. Finally, I will turn my attention to questions of unity in this composition, which is based on a wide array of rows and uses little repetition of individual row forms.

GENERAL OBSERVATIONS Composed in 1983, these etudes are dedicated to the famous Russian bassoonist Valeriy Popov. 2 As one might expect by the title, the Five Etudes appear to have been composed with a pedagogical purpose in mind. Jeffrey Lyman notes that [i]n listening to the Five Etudes, one realizes that the standard etude books of the nineteenth century

I am aware of only two writings about the Five Etudes, neither of which deals with the work in any length. The first is Jon Beebes annotated bibliography Music for Unaccompanied Solo Bassoon (Beebe 1990), which has three paragraphs of descriptions of performance concerns of the piece (range, duration, some brief analysis). The second is Jeffrey Lymans article After Shostakovich, What Next? New Russian Soviet Music for Bassoon, from the International Double Reed Society Journal (Lyman 1996). This article gives a paragraph-long description of the overall shape of each of the Etudes, which would be suitable for recital program notes. 2 Beebe 1990, 24. The published score (Deutscher Verlag fr Musik, Leipzig) does not indicate a dedicatee. According to the website of the Moscow Conservatory, Popov studied at the Conservatory during the early 1960s, and was appointed to teach there in 1971. He has been on the wind and percussion faculty at the Conservatory ever since, and, in 2007, he was named the head of that department. http://www.mosconsv.ru/english/teachers/about.phtml?65 accessed October 16, 2009.
1

94 could not have been far from the composers mind during their creation. 3 In fact, Denisovs widow, Ekaterina Denissova-Bruggeman, has written a brief note about this piece based on the composers own explanation of his work. 4 This note indicates that he had the student bassoonist in mind when composing this piece. Because, to the best of my knowledge, this particular note is not available in any published form, I will quote it here in full: This work was composed at the request of the publisher Deutsche Verlag fr Musik and is dedicated to the Russian bassoonist Valeriy Popov. The publisher Deutsche Verlag fr Musik made an entire series of pedagogical books for various solo instruments. To this end, they commissioned compositions from various 20th-century composers in order to have works of [not only] high musical quality but also aesthetic variety. Their idea was also to introduce a maximum number of innovative techniques for each instrument. This is why Denisov composed several works of the same design for Deutsche Verlag fr Musik in the mid-1980s: Winter Landscape (1987) for harp, Two Pieces for Solo Flute (1983), a piece for vibraphone, 5 Dead Leaves (1980) for harpsichord, and Five Etudes for Bassoon (1983). The composer writes: I think that this type of collection, devoted specifically to a chosen instrument and its new technical procedures, is very useful. There should be as many of these [collections] as possible. It is evident that they are indispensable for performers. But [they are] also [indispensable] for composers who, through this work, have the opportunity to study the instruments in greater depth, exploring their potential technique and expression. Furthermore, as the composer knows that his music will be published with the works of other composers (often already well known, and even famous), he is well obliged to apply [his best effort] to not appear weaker and be ashamed of his work. 6

Lyman 1996, 55. Ekaterina Denissova-Bruggeman, email correspondence with the author, February 6, 2008. 5 Though she does not give a specific title here, Denissova-Bruggeman is referring to Black Clouds (1984) for vibraphone. This piece is more commonly known as the first of Denisovs Three Pieces for Percussion, published in 1989 by Deutsche Verlag fr Musik. 6 Unpublished note Cinq Etudes pour basson seul, 1983, from the composers personal archive, in possession of Ekaterina Denissova-Bruggeman. The note was originally written in Russian by Mme. Denissova-Bruggeman. I am deeply grateful to Mme. Denissova-Bruggeman not only for sharing this note with me (email correspondence, October 16, 2009), but also for translating it from Russian to French. The translation from French to English is my own.
3 4

95 Jon Beebe characterizes the Five Etudes as atonal and non-serial, with free use of twelve-tone rows, and states that the first movement, in particular, is based on continuous development of two motives[: o]ne consisting of a P4 and a tritone, the other a chromatic neighbor-note figure. 7 Beebes observation is accurate, yet incomplete. The pitch fabric of this Etude is derived from no fewer than ten distinct twelve-tone rows, and is a clear example of Kurbatskayas technique of twelve-tone rows, as discussed in Chapter 2. The two motives cited by Beebe are actually built into the structure of the rows used: of the ten rows used in this brief, 50-measure Etude, only three of them are not made up exclusively of ics 5, 6 and 1. The three remaining rows each make use of these interval classes as well, but with the addition of a few occurrences of ics 2 and/or 4. We shall examine the structure of each row shortly.

EXAMPLE 4.1: FIVE ETUDES FOR BASSOON (NO. 1) ROW CHART

1 2 3 4 5 6 7 8 9 10 11 12

AP9 -----------------------------AP8 ---------------DP7 -----------------------------------FP1 BP1 ----------------------------CP0 ---------------------------EP5 ----------13 14 15 16 17 18 19 20 21 22 23 24

AP1 -----------NR ------ GP7 --------- NR ----------------------------------------------(FP1)-------------CP6 ------------------------CP6 -------(EP5) 25 26 27 28 29 30 31 32 33 34 35 36

CI8 -------------------------------CP8 -----HP8 --------- NR ----------------------------------CP1 ------------------------ CP1 -------------------------37 38 39 40 41 42 43 44 45 46 47 48 49 50

IP9 ------------------------------------------------- NR --- FP8 ----------------------------JP0 --------------------------------------- NR

Beebe 1990, 24.

96 Example 4.1 shows the organization of rows and row forms in the first Etude. In this diagram, each of the ten row forms is assigned a letter, A-J, and the label NR is an abbreviation for no row: a succession of notes which, with or without pitch-class duplication, does not present all twelve tones. Although this piece contains no simultaneous pitches, I have listed the succession of row forms in Example 4.1 along two lines. Listing the row forms as such is intended to show that one of the major organizational principles of this Etude is the chaining of various rows and row forms. Thus, in Example 4.1, when two rows appear to be indicated simultaneously, row chaining is occurring: the last few pitches of one row are simultaneously functioning as the first few pitches of another row. I will have more to say about this feature shortly. The ten rows used are shown in Example 4.2, along with their respective ICSGs.

EXAMPLE 4.2: FIVE ETUDES FOR BASSOON (NO. 1) ROWS A-J AND THEIR ICSGS ROW A:
1 5 6

BIP-Vector: [300053] ROW B:

BIP-Vector: [500033]

97

EXAMPLE 4.2, CONTD

ROW C:
1 5 6

BIP-Vector: [300035] ROW D:

BIP-Vector: [500033] ROW E:

1 5 6

BIP-Vector: [500024] ROW F:

1 4 5 6

BIP-Vector: [300134] ROW G:

1 5 6

BIP-Vector: [500024]

98

EXAMPLE 4.2, CONTD

ROW H:

BIP-Vector: [210143] ROW I:

2 5 6

BIP-Vector: [010064] ROW J:

BIP-Vector: [300035] The three primary formal divisions in this Etude occur at mm. 12, 25 and 35; these divisions are articulated by changes in texture, register, dynamics and rhythm. However, the formal divisions generally do not coincide with the changing row forms. The opening section, from mm. 1-11, presents a number of somewhat isolated gestures, separated by silences of various durations. The bassoons large leaps force the instrument to cover a wide range: from C2 to Ab4. After m. 11, registral expansion (A4 in mm. 12, 15 and 19; D5 in m. 20; and finally E5 in m. 21) is accompanied by increased rhythmic activity and much shorter silences between individual gestures. This section also features the introduction of a small number of new interval classes and interval-class

99 successions, most notably the introduction of ic4 in m. 13. The arrival of E5 is followed by a rapid descent and decrescendo, forcing the third section (mm. 25-34) to begin with apparent uncertainty, with a return to the isolated gestures of the opening section. This time, however, the gestures themselves are shorter: mostly single staccato notes, or twoor three-note slurs. The final section (mm. 35-50) presents a series of five short phrases, played successively louder. Each phrase begins with a repeated staccato pitch, played in irregular rhythms, followed by a series of slurred notes.

ISSUES OF STRUCTURE As previously mentioned, ics 1, 5 and 6 are at the heart of the intervallic structure of nearly all of the twelve-tone rows found in this Etude. Example 4.2 reveals that only Rows F, H and I involve additional interval classes: Row F has a single occurrence of ic 4, Row I has a single occurrence of ic 2, and Row H has single occurrences of both ics 2 and 4. To be more precise, one might desire to measure the similarity between each pair of rows. In Chapter 3, I briefly discussed Tuukka Ilomkis work on the similarity of twelve-tone rows, but unfortunately, none of the methods proposed by Ilomki serve my present purposes. Of his seventeen similarity measures, two are based on intervallic structure: intervallic distance, and intervallic displacement. The former is concerned with the ordered pattern of intervals in a row form, and I am primarily concerned with the unordered interval-class content of the rows in the Etude. While the latter is concerned with examining row forms as sets of unordered intervals, it only returns numerical values when the row forms being compared share the same set of unordered intervals. 8
8

Ilomki 2008, 230-241.

100 Furthermore, neither of these similarity measures examines the interval classes of a twelve-tone row. Anticipating this issue, Ilomki notes that all set-class similarity measures based on the interval-class contents of set classes could beadapted with little effort. 9 To pick just one such similarity measure, Damon Scott and Eric Isaacsons Angle can be adapted to measure the similarity of the unordered interval-class content of two twelve-tone rows. 10 In its original conception, the Angle measurement is designed to measure the similarity between two pc sets by considering the entries of their interval-class vectors as coordinates indicating a location in a six-dimensional space. The similarity between two pc sets, then, is determined by measuring the angle between the locations of both of the pc sets. Results for Angle are returned in degrees, ranging from 0o to 90o, with larger Angles indicating a greater dissimilarity between the two pc-sets. 11 In adapting this measure for use with twelve-tone rows, it is necessary to use a rows BIP-Vector in place of the traditional pc set interval-class vector. 12 Of the ten rows in Denisovs Etude, three pairs have identical BIP-Vectors (Rows B and D, Rows C and J, and Rows E and G). Figure 4.1 shows the Angle measurements for each pair of rows. The information in this table provides concrete support to my earlier casual observation that these rows are generally quite similar not only in interval-class content, but also in relative distribution. Considering all the results in Figure 4.1, the average
Ibid., 236. Scott and Isaacson note that their Angle measurement has a very high correlation (over 0.90) with some favorite measures, namely Lewins REL2 and Castrns %Rel and his RECREL, suggesting that their measurement is, at least, as good a choice as any of these other methods of measuring pc-set similarity. See Scott and Isaacson 1998, 118. 11 Scott and Isaacson 1998, 107-108. The formula for calculating Angle is given on p. 110. 12 Because all twelve-tone rows are members of set-class 12-1, they all have the same interval-class vector [12 12 12 12 12 6]. As such, any calculations using the interval-class vector will return meaningless values, and, thus, the BIP-Vector must be used.
9 10

101 Angle for all the rows of Denisovs Etude is 26.9o. The largest values are 51.1o and 55.0o, both of which are pairs involving Row I. As can be gleaned by looking at the BIPVectors shown in Example 4.2, Rows H and I appear to have the least in common with the other eight rows, and correspondingly, the highest Angles are all measurements involving one of these two rows. 13

BASSOON (NO. 1)

FIGURE 4.1: ANGLE MEASUREMENTS FOR ROW PAIRS IN FIVE ETUDES FOR ROW PAIR Row A Row B/D Row A Row C/J Row A Row E/G Row A Row F Row A Row H Row A Row I Row B/D Row C/J Row B/D Row E/G Row B/D Row F Row B/D Row H Row B/D Row I Row C/J Row E/G Row C/J Row F Row C/J Row H Row C/J Row I Row E/G Row F Row E/G Row H Row E/G Row I Row F Row H Row F Row I Row H Row I ANGLE 24.9o 24.9o 32.7o 21.9o 16.5o 28.4o 24.9o 12.2o 21.9o 31.9o 51.1o 21.2o 11.6o 25.3o 37.3o 21.2o 36.6o 55.0o 19.8o 37.9o 24.1o

We can also seek to discover structural connections between the various rows in the Etude by examining the ICSGs of those rows, which are given in Example 4.2. Some inter-row relationships can be seen easily with these graphs. Most obviously, Rows C, E,

13

However, Rows H and I are not terribly dissimilar from each other, as suggested by an Angle of 24.1o.

102 and J all yield identical ICSGs. Rows C and J, therefore, can be said to have the closest relationship of any pair of rows: not only are their BIP-Vectors identical, but their ICSGs are identical as well. Furthermore, Example 4.2 also reveals that the prime forms of Rows C and J have nearly the same INT (and, thus, INT-C). The last seven entries of the INTs of these two rows are identical, and, in fact, the first four are identical, except for the reversal of the second and fourth entries. 14 These INT relationships are shown in Example 4.3.

EXAMPLE 4.3: INTS OF ROWS C AND J (PRIME FORMS) Row C: 6 5 6 e 6 e 6 e 5 6 7 Row J: 6 e 6 5 6 e 6 e 5 6 7

The identical ICSGs of Rows C, E and J are also literal subgraphs of the graph of Row F. Removing the node which contains ic4 (and both its associated arrows) from Row Fs ICSG results in the ICSGs of Rows C, E and J. Examination of the ICSGs also reveals a close connection between these three Rows and Row A: though Row As ICSG is not identical to the ICSG shared by Rows C, E and J, it is related by isomorphism. Though it is the least visually similar to all the other ICSGs, Row Hs ICSG is seen to be an abstract supergraph of the ICSGs of Row A and Rows C, E and J. Example 4.4 shows the ICSGs of Rows H and C with the specific interval-class content of their nodes removed. Removing the nodes b and c (and their associated arrows) in Row Hs ICSG (which represent ics 2 and 4, respectively) makes this abstract inclusion relationship clear: node a x, d z and e y. These
14

Though distinctly different in purpose, this swapping recalls Fortes R1 relation. See Forte 1973, 47-48.

103 graphs are each one arrow shy of being a complete graph, in which every possible path is allowed.

EXAMPLE 4.4: ABSTRACT INCLUSION RELATIONSHIP BETWEEN ICSGS OF ROWS H AND ROWS A, C, E AND J ROW H: ROWS A, C, E AND J:

In a row with three distinct interval classes, there are six possible interval-class successions. Example 4.2 shows that Row A uses five of those six possible successions: ic1 ic5; ic5 ic1; ic5 ic6; ic6 ic5; ic6 ic1. The only interval-class succession lacking from Row A is the succession ic1 ic6. However, this interval-class succession is only lacking from Row A in a structural sense. Example 4.5 shows the music for mm. 1-3, the first occurrence of Row A. As this example shows, a single pitch-class repetition on the musical surface results in the completion of Row As ICSG.

EXAMPLE 4.5: FIVE ETUDES FOR BASSOON (NO. 1) MM. 1-3 (ROW A)

In this example, the pitch C4 is found in two places: at order position 5 and again following order position 6 (B3). By repeating C4 in this way, Denisov uses the one missing interval-class succession: the ic1 between order positions 5 and 6 (and again

104 between order position 6 and the repeated 5) is now followed by an ic6 between the repeated order position 5 and 7. In this light, we can see how the chromatic neighbornote figure 15 observed by Beebe serves a deeper purpose: the exhaustion of all possible interval-class successions involving ics 1, 5 and 6.

EXAMPLE 4.6: FIVE ETUDES FOR BASSOON (NO. 1) MM. 5-6 (ROW A)

After its initial presentation, Row A returns in mm. 5-6, as seen in Example 4.6. In this realization of Row A, we also find that the repetition of order position 5 serves to present the single interval-class succession which is lacking in the Rows unrealized structure: that of ic1 ic6. A statement of Row B occurs between these two Row A passages. Shown in Example 4.7, this presentation of Row B involves no pitch repetition. But such a pitch repetition is unnecessary in this case: Row Bs ICSG (see Example 4.2) indicates that all six possible interval-class successions are present in the unaltered row.

EXAMPLE 4.7: FIVE ETUDES FOR BASSOON (NO. 1) MM. 3-5 (ROW B)

15

Beebe 1990, 24.

105 Continuing in the same vein, let us now examine the first appearance of Row C, which uses as its initial pitch the final pitch of AP8 (Example 4.5). This statement of CP0 is shown in Example 4.8.

EXAMPLE 4.8: FIVE ETUDES FOR BASSOON (NO. 1) MM. 6-9 (ROW C)

Row Cs ICSG (see Example 4.2) shows that, again, five of the six possible interval-class successions are present in the structure of the row. In this case (and also the case of Rows E and J, since they share identical ICSGs), the missing interval-class succession is ic 5 ic 1. But, in a manner similar to that shown in Examples 4.5 and 4.6, the repetition of order position 3 (after 4) forms an ic5 with the Bb of order position 5, which in turn is followed by an ic1 between 5 and 6. These examples suggest that the different rows used in the opening section of Denisovs Etude (mm. 1-11) all are realized in particular ways to bring the full range of possible interval-class successions to the fore. 16

ISSUES OF FUNCTION The primary functional issue in this Etude is the way in which the different rows and row forms are linked from one to the next. In his 1970 essay on Weberns Piano

Although I have not explicitly discussed Row Ds role in this A-section, it is obvious from its ICSG shown in Example 4.2 that Row D, in its natural state presents all six possible interval-class successions.
16

106 Variations Op. 27, Denisov uses the term serial modulation (seriynaya modulyatsiya) to describe Weberns practice of row chaining, where the last note or notes of one row form becomes the first note or notes of a new row form. 17 I will adopt this term here, as it seems especially appropriate for Denisovs method of linking together not only different row forms, but even different rows. I will also use the term modulating area to refer to the notes shared between two row forms. Throughout the Etude, every time two row forms are presented successively, without any intervening non-row material, Denisov uses the technique of serial modulation to connect one row form to the next. These connections are shown by the overlapping lines of Example 4.1. The size of the modulating area is not entirely consistent throughout the Etude; some pairs of rows are linked by three pitches, others are linked by four, and still others are linked by a single shared pitch. Denisovs selection and presentation of specific row forms appears to be guided by invariant pitch-class segments. Clearly, the common notes in the modulating area are one type of invariance, but each successive pair of row forms features additional invariant dyads, trichords, tetrachords, or pentachords. Figure 4.2 presents a list of all successive pairs of row forms in this Etude to highlight these invariances. In this list, the common notes found in the row forms modulating area are enclosed in circles, and the other invariant pitch-class segments are indicated with brackets. 18 While several different invariance schemes are employed throughout the Etude, it is clear that the maintenance of tetrachordal content is the most common method of
Denisov writes, By coupling the row forms by means of one or more common notes, a transition from one pitch level to the next is carried out, or in other words, serial modulation is achieved. Denisov 1970, 51 (italics original). 18 In many cases, there exist a variety of ways of identifying invariant pitch-class segments. For example, in the second row of Figure 4.2 (mm. 3-6), we could also bracket the trichord {017}, and the dyads {56}, {te} and {34}. In these cases, I have chosen to prefer the most regular partitioning: if invariant pitchclass segment of equal size can be identified, I will label those segments.
17

107 FIGURE 4.2: INVARIANCES IN ROW-FORM SUCCESSIONS IN FIVE ETUDES FOR

BASSOON (NO. 1)
mm. 1-5

Row A: AP9 9 3 t 4 5 0 e 6 7 2 8 1

Row B: BP1 1 0 7 6 5 e t 4 3 8 2 9 Row A: AP8 8 2 9 3 4 e t 5 6 1 7 0 Row C: CP0 0 6 e 5 4 t 9 3 2 7 1 8 Row D: DP7 7 1 8 2 3 t 9 4 5 6 0 e Row E: EP5 5 6 0 e t 4 3 9 8 1 7 2 Row F: FP1 1 7 2 8 4 3 9 t e 6 0 5 Row C: CP6 6 0 5 e t 4 3 9 8 1 7 2 Row A: AP1 1 7 2 8 9 4 3 t e 6 0 5 Row C: CP1 1 7 0 6 5 e t 4 3 8 2 9 Row H: HP8 8 2 9 t 4 e 3 5 6 1 7 0 Row J: JP0 0 6 5 e 4 t 9 3 2 7 1 8

mm. 3-6

Row B: BP1 1 0 7 6 5 e t 4 3 8 2 9

mm. 5-9

Row A: AP8 8 2 9 3 4 e t 5 6 1 7 0

mm. 6-11

Row C: CP0 0 6 e 5 4 t 9 3 2 7 1 8

mm. 8-13

Row D: DP7 7 1 8 2 3 t 9 4 5 6 0 e

mm. 11-14

Row E: EP5 5 6 0 e t 4 3 9 8 1 7 2

mm. 12-16

Row F: FP1 1 7 2 8 4 3 9 t e 6 0 5

mm. 14-17 mm. 16-18 mm. 25-29 mm. 27-30 mm. 30-33

Row C: CP6 6 0 5 e t 4 3 9 8 1 7 2 Row C: CI8 8 2 9 3 4 t e 5 6 1 7 0 Row C: CP1 1 7 0 6 5 e t 4 3 8 2 9

mm. 45-50

Row F: FP8 8 2 9 3 e t 4 5 6 1 7 0

108 connecting the various successive row forms in this piece. Nine of the thirteen rowform pairs 19 shown in Figure 4.2 are related by tetrachordal invariance. To be even more specific, all of the row-form pairs with shared tetrachordal content feature the same relationship between those shared tetrachords. In every case, the tetrachords are related in the symmetrical manner shown in Example 4.9.

EXAMPLE 4.9: SYMMETRY OF INVARIANT TETRACHORDS Row form X: {ABCD} {EFGH} {IJKL}

Row form Y:

{IJKL} {EFGH} {ABCD}

These tetrachordally invariant row-form pairs are found exclusively in the first, second and third formal sections of the piece; tetrachordal invariance is not a feature of the final section. In fact, it is precisely this tetrachordal invariance that runs as a common thread between the first two sections. The first section makes use of this relationship in its second pair of row forms (BP1 and AP8 ) and again in its last pair of row forms (CP0 and DP7). But as row form DP7 is ending, the pitch-class segment < 5 6 0 e > allows this row form to modulate to row form EP5. This particular modulation features the largest modulating area of any successive row forms throughout the Etude : an entire tetrachord is shared between DP7 and EP5. This shared tetrachord makes up the final four pitches of the Etudes first section. As the second section moves through row form EP5 to FP1, from FP1 to CP6, from CP6 to AP1, and from AP1 back to CP6, each pair of rows utilizes the same three
Although there are only eleven distinct row pairs listed in Figure 4.2, two of these pairs are used twice each (CP6 and AP1; CI8 and CPe).
19

109 sets of tetrachords: {056e}{349t}{1278}. As Example 4.9 shows, every pair of tetrachordally invariant row forms switches the placement of the first and third tetrachords while keeping the second tetrachord in the same place. As such, from the beginning of the second section of the Etude until the introduction of non-row-based material at m. 18, the tetrachords {056e} and {1278} are found alternating between placement at the beginning and end of each row form, while {349t} is continually found at order positions 4-7 of each row form. The exchange of tetrachord placement is shown in Example 4.10.

EXAMPLE 4.10: LOCATION OF TETRACHORDS {056E}{349T}{1278} IN MM. 6-18 Row CP0, m. 6: 0 6 e 5 4 t 9 3 2 7 1 8 Row DP7, m. 8: 7 1 8 2 3 t 9 4 5 6 0 e Row EP5, m. 11: 5 6 0 e t 4 3 9 8 1 7 2 Row FP1, m. 12: 1 7 2 8 4 3 9 t e 6 0 5 Row CP6, m. 14: 6 0 5 e t 4 3 9 8 1 7 2 Row AP1, m. 16: 1 7 2 8 9 4 3 t e 6 0 5 Row CP6, m. 17: 6 0 5 e t 4 3 9 8 1 7 2

A brief, eight-note pitch segment occurs between m. 18s row form CP6 and m. 19s row form GP7. This segment, labeled as X, is shown as Example 4.11, with a bit of the C- and G-rows surrounding it.

110 EXAMPLE 4.11: FIVE ETUDES FOR BASSOON (NO. 1) MM. 18-19 X

Initially, the segment X sounds like a repetition of a small portion of CP6, featuring the same < 4 3 4 9 8 > motive from the beginning of m. 18, with pc9 placed an octave higher. This pitch segment can be seen as a portion of yet another ordered aggregate, overlapping with the surrounding rows. However, this overlap produces a pitch-class duplication which prevents listing this as a row. Although we have seen repetitions within rows earlier in this Etude, the duplication here is of a different sort. Prior to m. 18, all of the repeated pitches have occurred as part of a neighbor-note motive: they have been repetitions of a given rows order position segment < n n+1 n >. In the present case, pc7 is presented two times: once during the modulating area between CP6 and X, and once in the modulating area between X and GP7. This passage also features a similar type of invariance as was found in the earlier portions of the Etude. While neither the invariance pattern between CP6 and X nor the invariance pattern between X and GP7 have the consistent sharing of pitch-class segments of uniform sizes that we saw in mm. 6-18, they do share the same kind of symmetrical relationship between those same row forms. The relationship between these three units is shown in Example 4.12. In this example, repeated pitch classes are shown in parentheses.

111 EXAMPLE 4.12: FIVE ETUDES FOR BASSOON (NO. 1) MM. 17-20 (CP6, X, GP7) CP6, mm. 17-18: 6 0 5 e t 4 3 (4) 9 8 1 7 2

X, mm. 18-19: 1 7 2 4 3 (4) 9 8 e t 6 (7)(6)(7) 0 5

GP7, mm. 19-20: 7 6 (7) 0 5 e t (e) 4 3 (4) 9 8 1 2 (1)

Each pair of pitch-class units (CP6 and X; X and GP7) shares three invariant subsets, two of which are shared among all three units. In moving from one unit to the next, the invariant pitch-class segments feature a similar kind of symmetrical exchange as we saw in Example 4.9. Even though, in this case, the invariant segments are not of the same size, nor do they account for every pitch class in the three units, the principle remains the same among these invariant segments: the first invariant segment becomes the last, the last becomes the first, and the invariant segment between these two remains in the middle. The next phrase, from the anacrusis to m. 21 through m. 24, represents the longest non-row-based passage of the entire Etude. But even this passage shares a connection with the row form (GP7) that precedes it. As if deconstructing the invariant relationships established in the first 20 measures of the piece, Denisov chooses an ordering of pitch classes which highlights several small sets (mostly dyads) from GP7. Example 4.13A shows the relevant division of GP7, with each small set labeled with a letter from a-e. Example 4.13B shows the placement of those small sets within the nonrow-based passage.

112 EXAMPLE 4.13A: FIVE ETUDES FOR BASSOON (NO. 1) SMALL SETS OF GP7 a b c d e GP7, mm. 19-20: 7 6 0 5 e t 4 3 9 8 1 2 EXAMPLE 4.13B: FIVE ETUDES FOR BASSOON (NO. 1) MM. 20-24 d c b a b d c e d e c e d mm. 20-24: 8 9 4 3 t e 6 5 0 t e 8 9 7 5 3 4 2 1 8 9 7 1 2 3 4 2 1 0 t 8 9

ISSUES OF UNITY In a fifty-measure composition, does the presence of ten distinct twelve-tone rows allow for sufficient unity? The following section will serve to conclude the present chapter by addressing this issue. While it seems unlikely that a relatively short piece of music could project a unified musical whole when it is built on the structure of ten different rows, both the nature of the rows themselves and the motivic construction of the musical surface act to contradict this hypothesis. The web of inter-row relationships outlined in the earlier section of this chapter titled Issues of Structure certainly provides ample evidence that these rows, though distinct from each other, have a great deal in common. At the heart of this is the relationships made clear by the ICSGs of each row, and, even more directly, the shared unordered interval-class content (ics 1, 5 and 6) of the majority of the rows used. Regarding the motivic construction of this Etude, we have already seen that at least one commentator has noted two unifying motives; Jon Beebe has identified the frequent uses of ics 5 and 6, as well as the chromatic lower-neighbor figure as the basis of the pitch organization of the piece. While I have labored to show that these motives fall directly out of the pieces twelve-tone organization, the presence of an underlying

113 fabric of rows certainly does not render these motivic connections moot. The chromatic lower-neighbor figure is an obvious audible motive, used throughout the piece, most frequently found as pitch-class segment < 4 3 4 >, though it is not restricted to this pitch level. I have discussed the lower-neighbor motive earlier, in the context of < n n+1 n > order position repetitions in several row forms throughout the piece. Beebes consideration of ics 5 and 6 seems to miss the mark slightly, however. The melodic motion of this Etude is not a haphazard string of ics 5 and 6, but rather the combination of the two intervals to form members of set class 3-5, [016].

EXAMPLE 4.14: FIVE ETUDES FOR BASSOON (NO. 1) MM. 36-50

Example 4.14 shows the music for the closing section of the Etude. In this example, I have labeled the occurrences of this pc set throughout. The overlapping brackets indicate that nearly every pitch in this fifteen-measure section can be explained

114 as a member of multiple members of set class 3-5. This is somewhat reminiscent of the way set class 4-5, [0126], saturated the musical surface of the first movement from Volkonskys Musica Stricta, as seen in Chapter 2. By using rows of similar construction, and by bringing these structural relationships to the fore by using them to create salient motives, Denisov is able to provide his Etude with two levels of unity: the deep-level unity given by similarity of the structural material (the interval-class content of the various rows) and the surface-level unity given by instantly audible and identifiable motives. This observation leads to a potential question: why consider this piece to be rowbased at all? Couldnt one analyze this piece without recourse to a discussion of any twelve-tone rows? Certainly, it is easy enough to interpret this Etude as a realization of a kind of game of intervallic manipulations. While it is not my intention to hypothesize about the precise manner in which Denisov might or might not have gone about composing this piece, it is not unreasonable to imagine a composer setting up the system of limitations in which Denisov appears to be operating in the majority of this Etude: using only three interval classes, complete the aggregate with minimal pitch-class repetition. The distinction here is subtle, and ultimately is little more than a chicken-and-egg dispute. Is the piece based on the transformations of several rows which feature ics 1, 5 and 6? Or is the piece based on manipulations of ics 1, 5 and 6 which happen to form several ordered aggregates? While I suspect that both interpretations have some validity, it seems unlikely that the use of a variety of twelve-tone rows was not a conscious part of the compositional act. Given that there are 1,392 twelve-tone rows that use only ics 1, 5 and 6, it is improbable that Denisov would ever stumble upon a different member of a

115 previously used row class. While he does not repeat the same row class very often, members of Rows A and C are both used multiple times in this Etude. Furthermore, the third Etude in this set of Five Etudes is based exclusively on transformations of Row C. This evidence makes it difficult for me to dismiss the underlying row structure as merely incidental, or the result of intervallic manipulations.

***

In this chapter, we have seen Denisovs use of highly similar rows in the first of his Five Etudes for Solo Bassoon. I have demonstrated the nature of the relationships between these rows by using the ICSG and adapting Scott and Isaacsons similarity measure, Angle. I have also demonstrated Denisovs technique of serial modulation between the different row forms in this Etude, and I have tried to show how the underlying structural connections between the different rows manifest themselves as motives on the surface of the music. In the next chapter, we shall see a different kind of multiple-row piece: Denisovs Sonata for Alto Saxophone and Piano. Where the Etude made use of ten closely related rows, the Sonata makes use of two widely contrasting rows.

116 CHAPTER 5 ANALYSIS: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3) From Denisovs use of multiple twelve-tone rows with highly similar intervallic construction in the Five Etudes for Solo Bassoon, we now turn our attention to the analysis of a piece in which the multiple rows have contrasting intervallic material. In this chapter, I will discuss the third movement of Denisovs 1970 Sonata for Alto Saxophone and Piano, which is probably his most frequently performed and recorded work, and a staple of the modern saxophone repertoire. 1 This piece takes most of its pitch material from two distinct twelve-tone row classes, but also contains a fair amount of non-rowderived material. We begin with some general observations about Denisovs writing for the saxophone and discuss five existing published studies of the piece. The ensuing analysis has two main objectives: to examine Denisovs use of serial modulation in the Sonata, and to demonstrate how the interval-class succession graph can inform an understanding of the relationship between the two rows, and also the relationship between the dodecaphonic and non-dodecaphonic sections of the composition. The Sonata is the second of six works including saxophone written by Denisov during his career, and the first to use the instrument in a soloistic capacity. The 1966 Five Stories of Mr. Keuner for tenor and seven instrumentalists uses the alto saxophone as well, but the instruments role in this piece is almost exclusively supportive, filling the role of low woodwind the pieces instrumentation, which is similar to that of Stravinskys The Soldiers Tale. 2 Furthermore, the saxophone part does not involve any significant technical challenges for the performer.

1 2

Helton 2000, 16. Denisov and Stravinskys pieces each make use of one high and one low representative from woodwind, brass, and string families (though Denisovs piece uses a piano in place of a high string instrument). The instrumentation for Five Stories of Mr. Keuner is as follows: Eb clarinet, alto saxophone,

117 The Sonata marks a new phase in Denisovs interest in the saxophone, in particular, as a solo instrument. In the years following 1970, the composer wrote Two Pieces for Alto Saxophone and Piano (1974), the Concerto Piccolo (1977) for a solo saxophonist playing successively soprano, alto, and tenor saxophones with the accompaniment of six percussionists, a transcription for alto saxophone of his Concerto for Viola (1986), a Quintet for Four Saxophones and Piano (1991), and the Sonata for Alto Saxophone and Violoncello (1994). This new phase appears to have been brought about by Denisovs interaction with the famous French saxophonist, Jean-Marie Londeix, to whom the work is dedicated. Through the encouragement of Dmitri Kabalevsky, Londeix gave six performances and several public lectures in Moscow during March, 1970. 3 During this visit, Londeix apparently made quite an impression on the Soviet music world in general, and on Denisov specifically. Before this visit, the saxophone was not taken seriously in the Soviet Union as an instrument of serious artistic merit, and, in fact, it was not even taught in the Moscow Conservatory. 4 Denisov was so impressed by Londeixs remarkable performance abilities that he asked the performer to make a recording demonstrating the unique sonic capabilities of his instrument, which he did upon his return to France. It appears to have been this recording that helped Denisov to discover a number of extended saxophone techniques (quarter tones, multiphonics, timbrealtering fingerings, etc.) which were incorporated into the Sonata. 5 The impact that this work has had for the saxophone community is immeasurable. Londiex himself has noted that Denisovs Sonata has become one of the
trumpet, trombone, piano, double bass and percussion. Stravinskys The Soldiers Tale uses the following instrumentation: clarinet, bassoon, trumpet, trombone, violin, double bass and percussion. 3 Londeix 2000, 222. 4 Umble 2000, 100. See Umble 2000, 97-101 for a more detailed account of the impact that Londeixs visit to Moscow had on the contemporary music scene in general. 5 Ibid., 101.

118 most significant works in the saxophones repertoire. 6 It is interesting, then, that Denisov himself has said that he does not consider this work to be one of his most successful pieces. 7 Ekaterina Denissova-Bruggeman has written a brief note about the Sonata, based on her late husbands own thoughts about the work, which would be suitable for use as program notes for a performance of the composition. As was the case with the note about the Five Etudes for Solo Bassoon, which we saw at the beginning of the previous chapter, this note comes from the composers personal archive, maintained by his widow, Ekaterina Denissova-Bruggeman. Because, to the best of my knowledge, this note is not available in any published form, 8 I will quote it here in full: The Sonata for Saxophone and Piano was written at the request of the French saxophonist Jean-Marie Londeix and is dedicated to him. Denisov had a great deal of love for the saxophone, because it possesses considerable expressive and technical possibilities, both in the domain of jazz and classical [music]. Furthermore, Denisov had great esteem for the saxophones very large, sonorous, and dynamic amplitude. Sonata form is not present here. It is more a cycle of three movements, which have an analogous meaning to the three-part cycle of the 19th century sonata. The first movement has the function of the allegro of the sonata-cycle. Very energetic, it alludes to a jazz improvisation. The syncopations, the uneven rhythms found in the swing [style], disrupt the regular beat. The meters are constantly changing: 6/32, 4/32, 11/32, 17/32, etc. The principal theme is constructed in the manner of jazz: the melodic line is almost improvised, syncopated. The piano accompaniment is also typical of jazz: the chords which occur in the gaps left by the soloist sound isolated and heavily accented. The second movement is a monologue for solo saxophone. It is only at the end [of the movement] that the piano emits some very delicate
Londeix 2000, 222. Helton 2000, 16. James Umble also notes that [t]wenty-five years later the composer paradoxically stated he didnt totally identify with this work, as he found it too aggressive. See Umble 2000, 223, n. 332. 8 There are similarities between this note and two other sources: the liner notes Denisov wrote for Claude Delangles recording of the piece, on the CD The Russian Saxophone, BIS CD-765 (1996), and the discussion of the work found in Denisov and Shulgin 1998, 208-211. It seems that Denissova-Bruggeman has compiled the present note from at least those two sources.
6 7

119 chords, bringing a new sound. This movement is very free in its development, which gives the impression of improvised music. It also acts as a long prelude to the third movement. In the finale, all the musical material becomes imprinted with the jazz element that takes center stage at the beginning, when the piano imitates the movement of the boogie woogie. Here, a concertante and virtuosic style dominates the writing. Denisovs Sonata is very popular among saxophonists throughout the world. It appears as a required work in the majority of saxophone competitions, and has been recorded about fifteen times. 9 EXISTING STUDIES OF THE SONATA Denisovs most frequently performed piece is also the piece about which the most has been published. Four essays have been published dealing with particular aspects of the Sonata, all in English, and all primarily concerned with performance issues. In addition, there is a short discussion of the piece in Kholopov and Tsenovas book on Denisov, and I will discuss that as well. The first of these articles is the brief essay by the works dedicatee, Jean-Marie Londeix, published as part of James Umbles important reference book about the renowned saxophonist, Jean-Marie Londeix: Master of the Modern Saxophone. The introduction to this essay provides some background on the composition of the Sonata, most of which has been cited earlier in the present chapter. Following this, Londeix describes some of the technical concerns found in each movement of the composition. Most of these observations are either purely descriptive or provide some sort of seemingly self-evident advice to the performer: [t]he atonal and rather chromatic language [of the first movement] is Boulezian in its general character, especially in regard
9

Unpublished note Sonata pour saxophone alto et piano, 1970, from the composers personal archive. The note was written in Russian by the composers widow, and I am deeply grateful to Mme. DenissovaBruggeman not only for sharing this note with me (email correspondence, October 9, 2009), but also for translating it from Russian to French. The translation from French to English is my own.

120 to accentuation and dynamics, and the suggestion that the performer should clearly and consistently present each dynamic gradation throughout each movement. 10 Londeix describes the third movement as a jazz trio for saxophone, piano, and string bass, with the piano assuming the two latter roles. 11 He also mentions a number of jazz characteristics found in this movement: a strong rhythmic pulse, dominant seventh and blues chords, the pizzicato of the string bass at the outset 12 Londeix does provide a description of the form of the movement. He argues for an interpretation that divides the piece into four large sections (with divisions at m. 21, 43 and 53) and a coda (beginning at m. 76), which he refers to as a stretto recalling elements from the first two movements. 13 The author refers to the Sonatas first and third movements as being based on a twelve-tone row, and identifies the two transpositions of the D-S-C-H motive which begin each of the rows hexachords. In a footnote, he cites the suggestion of one of his students that these two four-note groups (D-Eb-C-B and A-Bb-G-F#) can be re-ordered diatonically (G-A-Bb-B-C-D-Eb-F#) to demonstrate a kind of altered blues scale. 14 This observation, unfortunately, is not explored any further. It is unclear, for example, why G has been selected as tonic in this scale, since that particular pitch class does not appear to have any significant structural importance throughout either the first or the third movement.; if anything, the consistent presence of the pitch class C at the beginning of the pianos ostinato in m. 1 and the pitch class D at the beginning of the saxophones repeated statements of the primary row immediately following suggests that

Umble 2000, 223. Ibid., 224. 12 Ibid. 13 Ibid., 225. 14 Ibid., 223.
10 11

121 a more likely tonal hearing of the beginning of the third movement, at least, would likely identify either one of those two notes as the primary pitch center. Regardless, Londeix correctly identifies the primary row of the movement (and of the Sonata as a whole), but he neglects to even mention the presence of a second row. Instead he refers to the dominant seventh and blues chords 15 used in the third movement, without mentioning that those purported jazz elements have a deeper level of organization as part of a twelve-tone row, which is constructed precisely to highlight those specific characteristics. Denise Dabneys 1995 article seeks to provide diverse analytical insights by taking a multicultural approach to Denisovs Sonata. 16 The over-arching argument in this essay is that Denisov represents the music of a different culture in each of the movements of the composition. The first movement is labeled Russian because it suggests a possible element of anarchy by using violent rhythms, extreme dynamics, [and] textural and dissonant sonorities. 17 Furthermore, she claims that the repeatednote patterns depictSoviet oppression by representing machine gun firing. 18 The second movement, on the other hand is labeled East Asian because of the use of quarter-tones and unmeasured time, and the final movement is deemed American because of the influence of jazz music. 19 While these labels may or may not be helpful in evoking a particular approach to performance, they are likely intended to be understood in a rather informal, casual manner, and therefore I will not explore them in any detail here.

Ibid., 224. Dabney 1995, 10. 17 Ibid., 12. 18 Ibid. 19 Ibid., 13.
15 16

122 Dabney notes in passing that the third movement begins with a 12 tone row in the saxophone part, 20 but while she makes frequent reference to a three-note motive which is heard throughout the [first] movement, 21 she does not recognize the motives source in the first three notes of Shostakovich's D-S-C-H motto, which begins the twelve-tone row that governs much of the pitch-class material of the movement. In 1999, Joren Cain published an article comparing stylistic aspects of Denisovs Sonata for Alto Saxophone and Piano with his Sonata for Alto Saxophone and Violoncello (1994). Cain is the only author surveyed thus far who not only notes the presence of jazz-like triads and seventh chords, but also recognizes their origin from a second row used in the third movement. Additionally, he makes at least a passing reference to the interaction between the two rows, noting how the two instruments begin to exchange the rows between mm. 53-69. 22 Unfortunately, the articles primary goal does not allow Cain to explore the movement in greater depth. Finally, Jonathan Heltons 2000 article presents some more detailed analytical observations of the pitch structure of the work. Like Londeix, he also proposes an outline of the movements form, interpreted in three large sections with an additional coda (with divisions at m. 21, 53 and 76). 23 While he also fails to identify the second row as the source of many of the third movements jazz harmonies 24 and intermittent chordal material 25 he goes to great lengths to analyze the structure of the pieces primary row. He notes that the rows INT does not contain any occurrences of ics 4 or 6, and shows how these interval classes can be formed by simply exchanging the notes
Ibid. Ibid., 11. 22 Cain 1999, 37. 23 Helton 2000, 31. 24 Ibid., 30. 25 Ibid., 33.
20 21

123 found at order positions 4 and 5, 7 and 8 (to form ic4), or 7 and 8 simultaneously with 9 and t. 26 Helton makes the claim that these particular order position manipulations do occur in the piece, 27 although he never identifies precisely where in the piece they occur. In their book on Denisov, Kholopov and Tsenova devote three pages to the Sonata. Again, the comments provided are largely descriptive rather than analytical, but the authors make special note of the two rows present in the movement, and the intervallic dissimilarity between them. They mention that Denisovs serial method generally involves working with the series as [one would work with] a theme, allowing for its fragmentation, incomplete statement and repetition of separate segments. 28 From this thematic perspective, which resonates with Kurbatskayas definition of the technique of twelve-tone rows, they are able to draw an analogy to sonata form for this movement. The two rows are comparable to primary and secondary themes, presented in mm. 1-20, developed in mm. 21-53, and recapitulated beginning in m. 53, where the character of the main part is imparted to the material of the subsidiary second series. 29 All five of the writings discussed here mention jazz influences in the third movement only on a rather superficial level. While this might appear to be evidence of incomplete analysis, it actually resonates with Denisovs own writing on jazz music. In 1968 he published the article New Music and Jazz in the multi-lingual journal The World of Music. In this article, he attempts to show ways in which jazz music (and folk
Ibid., 17-18. These are not exhaustive lists of the order position exchanges that can produce ics 4 or 6. For example, exchanging order positions 2 and 3 or 5 and 6 also will produce ic4 in this row. 27 Ibid. 28 Kholopov and Tsenova 1993, 119; 2000, 159. 29 Ibid. Helton also suggests a tentative sonata form reading of this movement, based on the recurrence of particular row chaining devices. See Helton 2000, 38.
26

124 music) might provide new musical structures for use in concert music. As he says in the introduction to the article, we should not become snobs and, merely because many people regard folk music and jazz as primitive forms of art, shut the door on their natural penetration into serious music. 30 But throughout the article, Denisov remains relatively vague on the specifics of the jazz influences that can be adapted for this purpose. He mentions a prevalence of the use of ostinato as a formal device in the art music of the 20th century (citing specifically Stravinsky, Shostakovich and Orff), and suggests that jazz provides us with new forms of ostinato, not all of which (especially those in the finest works of avantgarde [sic] jazz) have yet been investigated by composers. 31 But Denisovs discussion of these forms of ostinato ends here; at no point is the topic engaged in any more detail. While it is clear that, for Denisov, the use of ostinato can be seen as a jazz influence, it is unclear whether the ostinato walking bass sections of the third movement of the Sonata are among these new forms of ostinato or not. Additionally, he discusses connections between the general idea of improvisation and various components of musical form which are left unfixed. 32 In a sense, Denisov appears to be primarily concerned with broad, general characteristics of jazz music, rather than with specific features of a particular jazz style or performer. As a result, the third movement of Denisovs Sonata is able to sound jazzy in a general sense, while still incorporating a number of typical characteristics of his own music, and thus being unmistakably by Denisov.

Denisov 1968, 31. Ibid., 31-32. 32 Ibid., 33. Denisov experimented with aleatoric techniques in a number of works, perhaps most famously the 1965 Crescendo and Diminuendo for Harpsichord and Twelve Strings.
30 31

125

FORM Example 5.1 shows the form of the third movement. In this diagram, all the row forms from this movement are labeled, and the non-row-based sections are identified by the use of boxes, with some of the most prominent pc sets labeled inside these boxes. The layout of the diagram shows that this movement primarily features a linear, threevoice texture made of saxophone, piano right hand, and piano left hand. This texture, of course, corresponds with Londeixs interpretation of the movement as a jazz trio for saxophone, piano, and string bass, with the piano assuming the two latter roles. 33

EXAMPLE 5.1: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), FORM * = incomplete row form = out-of-order row form = non-serial material
OSTINATO 1 1 2 3 4 5 6 7 8 9 10 11 12

Sax: AP2 ------------------------- AP4 --- AP2 AP4 -- AP6 --------- AP8 -------------- APt --------- AP0 ----- AP2 ---------RH: BP2 --------- BRI1 ----- BP9 -------- BRI8 ------- BP4 ---------------LH: *AP0 (+Eb) --------------------------------------------------------------------------------------------------------------------------------

13

14

15

16

17

18

19

20

Sax: AI7 ----- AI5 -------------- AI3 ---------------------------------------------------------- AI1 --- *AIe --- [0369] [0258] RH: BPe ---------- BRI3 ----------------*BRI9 *BRI7 *BRI5 *BRI3 [048][037][014] *BP1 *BPe *BP9 *BP7 [036][037][014] LH: (AP0) ---------------- [016] ------- 4 3 2 1 0 e

BP *BP BP *BP BP BRI

24 25

BP4

OSTINATO 2 21 22

23

26

27

28

Sax: AI4 ------------------------AI0 ---------- *ARP9 --------- ARIe ------------------------------ AIe ------------ AIt ------RH: AI4 -- AI2 --------AI0 -- APe -- AP1 ---AP4 ---------- AI7 ----- AI5 --- ARP0 --- ARPt ----------LH: *AP6 -----------------------*BRP1 -- BRP7 ---------- BP3 --------- BRI2 ---- BPt --------------- BRI9 -----------

29
RH: ARP8 - ARP6 33 See note 11. Sax: (AIt) ----------

30

31

32

33

34

35

36

------ARP4 --- AP2 ------- AP4 -- BRI7 LH: BP5 ---------- *AI2 -------------------------- AI2 -----------

BP - - -BP - - - -

*AP1 --- BP7 ------- BI4 -------------- BRP5 ------------- AP5 ----BI1 --------------------------BP4 --------------------------1 0

126

EXAMPLE 5.1, CONTD

37 38 39 40 41 42
[0147] [0258]

43

44
*BRI7 ------------- [0125] BRIt ------------- [0147] BRIe ------------- [036]

Sax: (AP5) --- BP7 ----------- BRI6 ----------- BP7 ------ BRI6 -------RH: BI0 ---- BRP1 - BI0 -------BP9 ---------- BI9 -------LH: *BP9 --*BP5 - BRI4 -BP5 - BI3 ----------- BP3 -------

WHISPER CHORUS 45
Sax: BP3 --------RH: BP3 --------LH: BP3 ---------

46
*AP2 --*AP2 --*AP2 ---

47
AP2 ----AP2 ----AP2 -----

48

49
*AP4 --------*AP4 ----------- BI2 *AP4 ----------- BI2

50
*AP4 *AP4 *BRP6 *AP4

51

52

*AP4 -- AI7 --- AP7 ---*AP4 --- AI7 -- AP7 ---*AP4 --- AI7 -- AP7 ----

OSTINATO 3 53 54 55 56 57 58

Sax: (AP7) --BI7 ---------------*BRP8 --------*AP9 ----- AP4 ----- AP6 ----- AP8 ------*BP2 [0147] [036] [0258] [0147] -------------------*BP2 *BP4 *BP2 RH: LH: *BP2 --------- *BP2 --------------------------- BP2 -------------------------------------------------- *BP2 --- BP0 ------

59

60

61

62

63

64

Sax: (AP8) -------------- APt ----------------------------- *BP5 -------AI7 ------------------------------------------RH: *BP9 [0369][0147][0258] [037][0258] BP2 ------ [0147]----------------------------------LH: [0258]------------------ *BP2 -------*BP9 ---------------------------------------------------------------- BP2 -----

65

66

67

68

69 (1st half)

AIe -------------------------------------------------------Sax: *AIt ------RH: [0147] [0258][037] [0147] ------------------------------------- BP8 ------------LH: BP3 ---------------------------------------- [0258] -------------[0358] ------ [0258] BP0 -------------

CODA 69 (2nd half)

Sax: *ARIt -------------RH: [016] LH: [016]

70

71

72

73

74

ARI6 -------- AP8 ---------------------------- APt ---------------------------------------------*AI7 -----*AIe [016] -------------[016] *AI 5 9

AI

75

76

77

78

79

80
[0369] *AIt ------------------[0147][0258] ----------------[0258] -------------------------

Sax: [0369] -------------------------------------------------------------------- *AI0 RH: *AIe ------------- [0369] -------------------------------------------------------------------[0369] -------------------------------------------------------------------- 3 LH: *AI8 [016]

BP BP4

81

82

83

84

85

Sax: [0369] --------------------------------------RH: [016] ----------------------------- [0127] [0147] LH: 2 [016] ----------------------------- [0147] [0258]

AP

[015] ----------------------------------------[016] -----------------------------------------

127

As seen in Example 5.1, my interpretation of the form of this movement is similar to that of Londeix and Helton. Both authors place structural divisions at m. 21, 53 and 76, but Londeix offers an additional division as well, at m. 43. I propose a formal model that casts the movement as a succession of three ostinati, the second and third of which are separated by a series of soft unison gestures I call a whisper chorus, 34 as these unison figures are reminiscent of a big band shout chorus, but at a piano dynamic. Like Londeix and Helton, I recognize the presence of a coda in this movement, but I argue that the coda begins in the second half of m. 69, rather than at m. 76, due to the reprise of material from the first movement. Measure 76, however, is a significant point of articulation within this coda, as it introduces the [0369] sonority which has not played a role in the piece prior to this point. I shall have more to say about this [0369] sonority shortly.

TECHNIQUES OF SERIAL MODULATION IN DENISOVS SONATA One of the primary organizational forces of the third movement of Denisovs Sonata for Alto Saxophone and Piano is his use of serial modulation. In the previous chapter, we saw serial modulation used as a method of moving between different rows, but in the Sonata Denisovs modulation techniques are limited primarily to connecting different forms of the same row. As we shall see presently, there are four primary techniques of serial modulation found in the third movement. The most widely known technique of serial modulation is undoubtedly David Lewins RICH operation. RICH is the operation which connects a twelve-tone row
34

Helton calls this section the jazz riff section. See Helton 2000, 35.

128 form to the retrograde-inverted form whose first two elements are order positions t and e of the original row form. 35 Additionally, Lewin defines TCH as the successive application of two RICH operations, and he defines MUCH as the operation which connects a pitch or pitch-class segment to the retrograde-inverted form of that segment which overlaps to the maximum possible extent. 36 Any ordered series of pitches (or pitch-classes) can be linked to another form of that series by RICH or TCH. Additional serial modulation techniques can be defined, based on specific properties of the series in question. One such property is found in the Sonatas Row A, as shown in Example 5.2.

EXAMPLE 5.2: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), ROW A ROW A:

As its INT shows, Row A begins and ends with the same interval. This property allows for chaining together two row forms that are related by transposition, which I call TRChain. 37

35 RICH is not restricted to twelve-tone rows. The operation can be applied to any pitch or pitch-class segment, regardless of whether or not that segment uses all twelve pitch classes. 36 See Lewin 1987, 180-183. 37 Care must be taken here not to confuse TR-Chain with Lewins TCH, as they both deal with transposition chains, but in different manners. Moseley 2009 also discusses something similar. His PCH is also a chain of transpostionally related row forms, but it only requires that the two row forms share a single pitch class. Moseley also introduces the labels RECH and ICH for retrograde-chain and inversion-chain, respectively. While these operations are useful in the repertoire Moseley investigates (as well as for the sake of theoretical completeness), they do not appear in Denisovs Sonata, and, as such, I will not deal with them here.

129

EXAMPLE 5.3: RICH AND TR-CHAIN OF ROW A

AP2

AP4

AP6

AP8

APt

AP0

AP2

ARI7

ARIe

ARI3

Example 5.3 shows the relationship between RICH and TR-Chain in Row A, starting with AP2, the first complete row form found in the movement. In this Example, the diagonal arrows represent RICH operations, and the horizontal arrows represent TRChains. RICH and TR-Chain are two of four different serial modulation techniques featured in the third movement of Denisovs Sonata. Their presence is especially noticable at the beginning of the movement. The saxophones opening gesture is a clear statement of row form AP2, repeated three times. Each repetition adds a portion of AP4 at the end of AP2, using AP2s last two pitches as the first two pitches of AP4, as a TChain. Once AP4 has been stated in its entirety, Denisov uses the last two notes of that row form as the first two pitches of AP6. Similarly, AP6 modulates to AP8 by T-Chain, which in turn modulates to APt, AP0, and finally back to AP2, completing a full T-Chain cycle. The beginning of this cycle is shown in Example 5.4.

130 EXAMPLE 5.4: 38 MM. 3-7

SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), SAXOPHONE,

During this process, the pianists right hand supports the saxophones TR-Chain with a succession of RICH operations on forms of Row B, moving through BP2, BRIt, BP9, BRI8, BP4, BRI3, and BPe before the pattern in m. 14. The beginning of this process is shown in Example 5.5. EXAMPLE 5.5: SONATA FOR ALTO SAXOPHONE AND PIANO (MT. 3), PIANO RIGHT HAND, MM. 7-13

The irregular beaming in this Example is Denisovs. Additionally, there is a clear misprint in the score in m. 6: the concert G4 on the sixth 16th note of the measure is notated as Ab4. This misprint has been corrected in Example 5.4.
38

131 The third technique of serial modulation found in this movement involves chaining row forms together based on sharing a single pitch class. I call this technique PC-Chain. As this technique requires no shared interval content, it is possible to link together any two types of row forms. Furthermore, PC-Chain is the only modulation technique which is used to link a form of Row A with a form of Row B. An occurrence of this type of modulation is found in the saxophone part between mm. 35-37, as shown in Example 5.6.

EXAMPLE 5.6: MM. 35-37

SONATA FOR ALTO SAXOPHONE AND PIANO (MT. 3), SAXOPHONE,

The fourth and final technique of serial modulation is one that involves the manipulation of one row form in order for the modulation to take place. I refer to this technique as MANIP-Chain. This occurs in two different ways: as the result of an order position exchange involving either the first or last two pitches of one row form, and as the result of omitting a particular order position of one row form. This occurs in mm. 25-26, where the pianists right hand modulates from AP4 to AI7 by switching the order of the first two pitches of AI7, as shown in Example 5.7. EXAMPLE 5.7: SONATA FOR ALTO SAXOPHONE AND PIANO (MT. 3), PIANO RIGHT HAND, MM. 25-26

132 Figure 5.1 shows the occurrences of each of these four techniques of serial modulation in the third movement of the Sonata.

FIGURE 5.1: FOUR TECHNIQUES OF SERIAL MODULATION RICH

mm. 7-14 (RH: BP2 BRI1 BP9 BRI8 BP4 BRI3 BPe) mm. 25-30 (LH: BRI7 BP3 BRI2 BPt BRI9 BP5) m. 27 (RH: AI5 ARP0) mm. 34-35 (Sax: BI4 BRP5) mm. 37-40 (Sax: BP7 BRI6 BP7* BRI6) mm. 38-39 (RH: BI0 BRP1 BI0*) mm. 38-39 (LH: BP5 BRI4 BP5*)
* = MUCH operation

TR-CHAIN
mm. 3-13 (Sax: AP2 AP4 AP6 AP8 APt AP0 AP2) mm. 13-18 (Sax: AI7 AI5 AI3 AI1 AIe) mm. 21-24 (Sax/RH: AI4 AI2) mm. 24-25 (RH: APe AP1) mm. 26-27 (RH: AI7 AI5) mm. 27-31 (RH: ARP0 ARPt ARP8 ARP6 ARP4)

PC-CHAIN
mm. 24-26 (RH: APe AP4) mm. 24-27 (Sax: ARP9 ARIe) mm. 35-38 (Sax: BRP5 AP5 BP7)

MANIP-CHAIN
mm. 30-31 (RH: omitted note, ARP4 AP2) mm. 25-28 (Sax: op exchange, ARIe AIe) mm. 25-27 (RH: op exchange, AP4 AI7)

THE ICSG IN DENISOVS SONATA The first and third of the Sonatas three movements are based on transformations of a twelve-tone row or rows. The first movement uses only a single row, which is brought back at the beginning of the third movement. In the third movement, however, this row is joined by a secondary row. As mentioned in Chapter 3, these two rows share little in common in terms of intervallic content. The two hexachords of Row A begin with transpositions of D-S-C-H, 39 which, when coupled together, create a weaving

The Sonata for Alto Saxophone was not the first piece Denisov wrote which used Shostakovichs musical monogram. A year earlier (1969), Denisov composed a piece titled DSCH for clarinet, trombone, cello and piano. On the surface, this piece appears to be a tribute to the young composers mentor, who played a major role in the early part of his career. See Kholopov and Tsenova 1993, 8-12; 2002, 4-9. However, Schmelz 2007 argues that this piece might be better understood as a stylistic and generational critique (p. 309) of Shostakovichs music, perhaps brought on by the disappointment the younger generation of Soviet composers felt when Shostakovich joined the Communist Party in 1960. In addition to
39

133 chromatic line, typical of a great deal of Denisovs music, both dodecaphonic and nondodecaphonic. 40 On the other hand, Row B makes use of no semitones or whole tones, featuring instead tonal intervals which are linked together in such a way that makes a variety of triads and seventh chords available to the composer. These rows and their respective ICSGs are shown in Example 5.8.

EXAMPLE 5.8: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), ROWS A & B ROW A:
1 2 3 5

ROW B:
3 4 5

While Example 5.8 clearly shows that Row Bs ICSG is an abstract subgraph of Row As ICSG, the subgraph relation also reveals an important connection between Row A and the pianists non-row left hand part at the beginning of the Ostinato 1 section: a literal subgraph of Row A occurs in the opening measures of the Sonatas third movement. The left hand alone begins the movement with the six-note ostinato pattern shown in Example 5.9, which supports the saxophones TR-Chain of forms of Row A.

Shostakovichs motto, DSCH features a quotation from his Eighth String Quartet, which Schmelz says is deformed (p. 309) when it is combined with Denisovs serial musical language. In this respect, DSCH can be seen as indicative of the tenuous relationship between Shostakovich and the next generation of unofficial composers (pp. 305-306). 40 As discussed in Chapter 1, this is an example of what Kholopov and Tsenova call smooth threads.

134 EXAMPLE 5.9, SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), M. 1 41

The INT and INT-C of the left hand part show the presence of Row As intervals, but neither series of interval successions is found in the row as such. Cain has identified this ostinato as the first five notes of row form AP0 with an interpolated Eb: < 0 1 Eb 2 3 4 >. 42 The addition of Eb to this partial row form allows for the presence of ic5 in this ostinato, and produces an ICSG which can be seen as a literal subgraph of Row As full graph. 43 As it turns out, Eb is the only non-duplicating pitch class that does not violate at least one of the paths of Row As ICSG when placed between Db and Bb. 44 Example 5.10 shows the piano part for mm. 17-19 of the third movement. This passage marks the first point in the movement where the pianists two hands appear to be working in conjunction with each other, rather than as two distinct voices. In the second half of m. 18, the two hands present a melodic line in parallel minor ninths. The INT of this line is < t 7 1 1 9 e >, which is not part of any form of either Row A or Row B. However, we can see by examining the INT-Cs of the two rows that the interval-class

41 The repeat sign is not found in the actual score, but I have added it to the notation in Example 5.9 to demonstrate the motives ostinato function. 42 See Cain 1999, 36. Cains actual words are as follows: the ostinato is based on the first five notes of the pieces tone row (with an added D-flat). Given that there is no form of Row A that contains the pitch-class segment < 3 t 9 e >, I can only assume that the reference to D-flat in Cains article is a misprint, and should be replaced by E-flat. 43 To confirm this, one need only to look back at Example 5.8 and notice that, if the node labeled 3 (and that nodes associated arrow) is removed, what remains is the graph shown in Example 5.9. 44 Ab is the only other pitch class that comes close. While the interpolation of Ab would not add any new interval classes to the ICSG, it would force the arrows from ic2 to ic5 and from ic5 to ic1 to be reversed.

135 EXAMPLE 5.10: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), PIANO PART, MM. 17-19
17

(ic 4 found only in RH)

segment < 2 5 1 1 3 1 > is present between order positions 3-9 of Row A. As such, this phrase produces an ICSG which is a literal subgraph of Row As subgraph: all the nodes of Row As ICSG are present in the mm. 18-19 ICSG, but without the double-ended arrow connecting ics 1 and 2. In the first half of Example 5.10, which features a sixteenth-note sequence of descending triads, the pianists right hand presents an ordered sequence from the second hexachord of four RI-forms of Row B, each relating by T10 to the preceding row form. The pianists left hand follows a similar pattern of T10-related P-forms of Row B. 45 But unlike the right hands passage, the left hand hexachords are presented out of order. Specifically, order positions 1 and 2 have been exchanged, as have order positions 4 and 5. The ICSG shown below the example shows that these order-position exchanges do not result in any significant alterations of Row Bs ICSG; the exchange of these particular order positions does not result in any new interval-class content. If, for

These partial T10 cycles seem to be motivated by the T2 cycle formed in mm. 3-13 by the saxophones TR-Chained forms of Row A.
45

136 example, order positions 0 and 1 were exchanged, a new interval-class succession would result: ic4 would be followed by ic5. At this point, it might be instructive and perhaps compositionally suggestive to examine what I believe to be the four-step derivational process by which a third row, Row C, is created. Example 5.11 shows the music of part of the Ostinato 2 section, mm. 34-42. EXAMPLE 5.11: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), MM. 34-42

INT-C: 4 3 4 3 4 3 5 35 3 3 4 3 3 3 4 5 3

137

EXAMPLE 5.11, CONTD.

5 3 3 4 3 3 3 4 3 3 3

INT-C: 3 3 3 3 4

3 3 5 3 3 5 3 3 5

5 4 3 3 5 5 3 5

4 3 3 5 3 3

3 3

3 3 3 3 3 3 3 35

With the exception of a few measures of non-row-based music and one brief statement of AP5 in the saxophone, this section is largely dominated by forms of Row B. This fulfills the first step of the derivational process by which Row C is created: establish Row B as an important aspect of the structural organization of the piece.

138 From the end of m. 35 until m. 41, the piano part presents a two-voice inversion canon, three sixteenth-notes apart. Throughout this canon, Row B briefly loses control, and the musical material is non-row-based in m. 37 and the first half of m. 38. However, each of the slurred segments here produces the same ICSG as Row B. This brief section thus maintains the same set of interval-class successions, but has removed the dodecaphonic requirement of aggregate completion without pitch-class duplication. A similar event takes place in the saxophone part from the end of m. 40 through m. 42. While the saxophone no longer presents different forms of Row B, it does present interval-class successions that have already been heard as part of that row. The saxophone uses ics 3 and 4 at the end of m. 40, ics 3 and 5 after an eighth-note rest, and finally ics 3, 4 and 5 after yet another eighth-note rest. 46 Both the piano part in mm. 3738 and the saxophone part in mm. 40-42 serve to fulfill the second step of the derivation of Row C: separate Row Bs ICSG from its serial context. After the whisper chorus, the second half of m. 53 finds the left hand beginning the third ostinato pattern, based on row form BP2. Of interest here are the cascading sixteenth-note triplet figures found variously in the saxophone and piano right hand parts, shown in Example 5.12. Some of these sixteenth-note triplet figures are portions of different forms of Row B, and some of them are not. Of those that are not forms of Row B, every figure here features only ics 3 and 4, forming the ICSG that will soon be associated with Row C. But in the present situation, this passage demonstrates the third step of the derivational process: eliminate one of the interval classes from Row B (and consequently, remove one of the nodes of Row Bs ICSG).

Notice that this final grouping does contain two illegal interval class successions: ic5 is followed by ic4 in two places.
46

139 EXAMPLE 5.12: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), MM. 53-68 (TRIPLET FIGURES ONLY)
1
m. 54

m. 56

m. 58

m. 59

m. 60

m. 61

m. 62

m. 62

m. 64

10

m. 65

11

m. 66

12

m. 66

13

m. 66

15 14 m. 67

m. 68

140 The final step of this process involves returning the new ICSG to a serial context. This step is precisely what occurs at mm. 76-78, shown in Example 5.13.

EXAMPLE 5.13: SONATA FOR ALTO SAXOPHONE AND PIANO (MVT. 3), MM. 76-78

Example 5.13 shows an excerpt from an extended passage, near the beginning of the Coda, at m. 76. This section is based entirely on transformations of set class [0369]. The linear voices in this passage can be seen to have ICSGs that are subgraphs of the ICSG of Row B: simple manipulations of ic3 and ic4. Though [0369] does not exist as an ordered subset of either Row A or Row B, it is presented here through repetitions of ic3, which is Row Bs most frequently occurring interval class, as well as one of only two interval classes that Rows A and B share. The T1, T11, and T10 forms of the pc set marked X on Example 5.13 result by connecting interval-class segments < 3 3 3 > by a single ic4. Repetition of this pattern results in a new row, as seen in the left hand part: < 0 9 6 3 e 8 5 2 t 7 4 1 >. Example 5.14 shows the INT and its ICSG. Simple

141 inspection shows that this ICSG is a literal subgraph of Row Bs ICSG. 47 Figure 5.2 summarizes the derivational procedure I have just demonstrated.

EXAMPLE 5.14: ROW C CREATED BY LEFT HAND IN MM. 76-78

FIGURE 5.2: DERIVATION OF ROW C STEP 1: Establish Row Bs importance in structural organization STEP 2: Maintain Row Bs ICSG, but remove the requirement of aggregate completion without pitch-class duplication STEP 3: Remove one interval class from this ICSG STEP 4: Re-apply the requirement of aggregate completion to this new ICSG. Throughout the present discussion, as well as in Example 5.14, I have referred to this row as Row C. But it must be remembered that Row C is of a notably different structural importance than Rows A and B. Where those two rows govern large sections of the movement, Row Cs influence seems to be limited to the section shown in Example 5.13. Thus, Row C appears to function in a manner more like many of the non-recurring rows of the Five Etudes for Solo Bassoon examined in Chapter 4.

***

In this chapter, I have used the third movement of Denisovs Sonata for Alto Saxophone and Piano to demonstrate an important extension to the use of the ICSG in
47 Returning the ICSG from Example 5.14 to a serial context shown in Example 5.13 involves a compositional choice there are, in fact, 185 rows which have this as their ICSG. Speculations on the reason why this particular row was chosen as Row C are beyond the scope of the present study.

142 analysis. After examining the existing writings about Denisovs best known composition, I have used the ICSG in the same way as I employed it in the preceding two chapters: to examine the structural relationships that exist between the movements two rows. In addition to this examination, we have seen how subgraph relationships can inform an understanding of the manner in which Row B relates to several of the nonrow-based passages of the Sonata. Finally, I have endeavored to explain the process by which Denisov appears to derive a third row for limited use in the Coda of this movement.

143 CHAPTER 6 ANALYSIS: OCTET FOR WINDS (MVT. 2) To this point, we have considered music based on multiple twelve-tone rows, and developed the ICSG. Then, Chapters 4 and 5 each presented analyses of multiplerow pieces from Denisovs oeuvre in order to demonstrate practical applications of the concepts. In the present chapter, I will turn my attention to the second movement of his Octet for Winds.

GENERAL OBSERVATIONS Because, to the best of my knowledge, the only published writing dealing with the Octet is the exchange between Denisov and Dmitri Shulgin, a brief note about the piece will provide some useful background on the general nature of the composition. 1 As with the notes from Chapters 4 and 5, this note was written by, Ekaterina DenissovaBruggeman, and is based on Denisovs own description of the piece. 2 The Octet for Winds was composed at the request of the clarinetist Sabine Meyer, for the ensemble that she founded. Regarding its form, it is a cycle in two movements, a configuration dear to Denisov, found in many of his works, beginning with the Wind Quintet (1969), Sonata for Solo Clarinet (1972), and later in the Sonata for Clarinet and Piano (1993) and the Symphony No. 2 for large orchestra (1996). The first movement is a type of extended introduction, while the second is based on entirely different thematic material. It makes use of the technique of groups, where an ostinato of a single note plays an important role. Denisov finds the opposition of the cantilena in the first movement and the pointillism in the second particularly interesting as a dramatic idea. [This opposition] makes an allusion to the pictorial contrast between lines and points. In many of Denisovs works, and particularly in his Octet, the lines of the first movement become the points of the second.
See Denisov and Shulgin 1998, 373-375. There are also some brief remarks on the Octet in the liner notes to Sabine Meyers recording of this piece (EMI 7235 5 57084 2). 2 Ekaterina Denissova-Bruggeman, email correspondence with the author, February 6, 2008.
1

144 Furthermore, in the singing first movement, the voices are never synchronized with each other. Here, the actual instrument group is very homogenous, [and] polyphony is present everywhere. Conversely, in the second movement, where one could easily believe that everything is scattered or even disorganized, the sounding material being very pointillistic, the tendency is actually toward synchronization. At the end, all the voices gather together in chords. However, the rhythmic pattern of his chords is so complex that one almost wonders if the composer did not have the malicious idea to test the ability of the musicians to play as an ensembleIn fact, septuplets within which the beats there are multiple, scattered points, and rhythmic figures like 9:8, 11:8, etc.all of this is singularly difficult to assemble. Thus, this work requires a large number of rehearsals. 3 The second movement of the Octet is organized around three primary rhythmic and textural gestures, featuring a series of different imitative textures. These gestures, defined by Kholopov and Tsenova and discussed in Chapter 1, are shooting, pointillistic bursts, and smooth threads. As a reminder to the reader, Examples 6.16.3 show instances of these three gestures as they occur in the Octet.

EXAMPLE 6.1: OCTET FOR WINDS (MVT. 2), MM. 1-3, HORN 2 SHOOTING

EXAMPLE 6.2: OCTET FOR WINDS (MVT. 2), M. 6, OBOE 1 POINTILLISTIC BURSTS

Unpublished note Octour vent, 1991 from the composers personal archive, in possession of Ekaterina Denissova-Bruggeman. The note was originally written in Russian by Mme. DenissovaBruggeman. I am deeply grateful to Mme. Denissova-Bruggeman not only for sharing this note with me (email correspondence, February 6, 2008), but also for translating it from Russian to French. The translation from French to English is my own.

145 EXAMPLE 6.3: OCTET FOR WINDS (MVT. 2), MM. 20-21, BASSOON 1 SMOOTH THREADS

As Denissova-Bruggemans note indicates, the movement ends with the eight players presenting complex shooting figures in rhythmic unison. But prior to that point, the bulk of the movement features imitative textures, with very little homorhythmic playing. In fact, through the first seventy measures, all shooting patterns are presented in strict four-, seven- or eight-voice canons, with imitative entries placed at constant quarter-note intervals. While shooting patterns are found fairly consistently throughout the movement, the pointillistic bursts and smooth threads appear to be used in different portions of the piece. Specifically, pointillistic bursts dominate the first thirty measures, are seemingly replaced by smooth threads from mm. 31-60, and return to the foreground during mm. 61-88. The final twelve measures (mm. 89-100) are made up almost entirely of homorhythmic shooting, with the exception of a single measure of smooth threads, at m. 98.

FORM Example 6.4 presents a form chart for the second movement. This chart shows the interaction of the three different gestures and the different transformations of the movements two twelve-tone rows. I shall have more to say about the rows themselves,

146 and their specific relation to the movements form shortly. But, for the sake of reference, Example 6.5 shows the two rows and their respective ICSGs. EXAMPLE 6.4: OCTET FOR WINDS (MVT. 2) FORM NR = non-row-based material
SECTION 1: Canon #1 1 2 3 4 5 6

Canon #2 7

Gesture: Shooting (S) ------------------------------------------------------------------------------ Pointillistic bursts (P) ---------------Rows: AP2 ---------------------------------------------------- AP6 -----------------------------------------ARI9 -------------------------------BPt/BRPt ------------------------------

Canon #3 9 10

11

12

13

Canon #4 14 15

16

17

Canon #5 18

Gesture: S --------------------------------------------------------------------------------------------------------------- P ---------------------AP1 ----------------------------------------------------Rows: APt ----------------------------------------ARI5 ------------- AP4 --------------------------------------------------------

SECTION 2: 19 20 21 22 Canon #6 23 24 25 Canon #7 26 27 28 29 30

Gesture: Smooth threads (T) ----------------------------P ---------------------------------------- Unison P ------Gesture: S -------------------------------------------AP3 --------------BP2 ------------Ob 1: AP4 -------------Ob 2: AI7 -------------AP3 -----------NR ------AP3 --------------BPt ---------Cl 1: AI9 -------------BP9 ---------Cl 2: AI9 ----------------------- AP3 ----------BP4 ---------Bsn 1: AI4 ----------------- AP3 -----------Bsn 2: ARP0 --------------- AP3 -----------BP2 ---------BP0 ---------Hn 1: AI2 --------------- AP3 ---------AP3 ---------NR -----------Hn 2: ARP6 -----------

BP2BP0
38

SECTION 3: Canon #8 31 32 33 34 35 36 37

Gesture: S --------------------------------------------------------------------- T ---------------------------------------------------------------Rows: BIt -------------------------------------------------------------------- (Bsn 1: AI5)------------------

147

EXAMPLE 6.4, CONTD

Canon #9 39 40 41 42 43 44
Gesture: S ------------------------------------------------------------------------------Gesture: T --------------------------------------------------------------------------Rows: Ob 1: BP4 ----------------------- NR ---------------------------------------Ob 2: BP0 --------------------------------- NR ----------------Cl 1: NR -----------------------------Cl 2: NR --------------------------------------NR -------------------Bsn 1: BP0 -------------------Bsn 2: BP0 -------------NR -----Hn 1: BI7 ----------------------

BP2 BP8 BP3

Canon #10 45 46 47 48

49

50

51

52

53

54

55

Gesture: S/T ------------------------------------------------------------------------------------------------------------------------------------Rows: BI8 ---------------------------------------------- BP4 --------------------

Canon #11 56 57 58 59 60

Gesture: S (with multiphonics) ----------------------------------------- (silence) Rows: BP2 ----------------------------------------------------------------

SECTION 4:

61

62

63

64

65

66

67

Gesture: Unison S ------------------------------------------------------------------------------------------------------------------------------Rows: BP2 ---------------------------------------------------------------------------------------------------------------------------------------

148

EXAMPLE 6.4, CONTD

Canon #12 Canon #13 68 69 70 71 72 73 74 75 76 77

Gesture: P -----------P/S -------------------P ------------------------------------------------------------Gesture: T -----------------------------------------------------------------------------NR --------------------------------NR ----- BP0 -----Ob 1: BPt -------------Ob 2: BPe -------------NR ---------------------------NR ----- BP0 -----BP7 -------- BP6 ---Cl 1: BI0 ------------------------- NR ---------------------------------------NR -------------------------BP7 ------------ BP6 -Cl 2: BI5 --------------------- BIt ----NR ------------------------NR ----- BP3 -----Bsn 1: BPe -------------Bsn 2: BIe ---------------NR -------------------------NR ----- BP3 -----BIt ----NR ------------- NR ------- BPt -----Hn 1: BI5 -------BIt ----NR -------------NR ------ BPt --Hn 2: BI5 ----

78

79

80

81

82

83

84

85

86

87

88

89

90

91

Gesture: P (some unison BP2 chords) ------Gesture: T ----------BI2 -----------Ob 1: BP9 ------------Ob 2: BI8 -------------NR -----------NR -----------Cl 1: BI6 -------------BP6 --------------------Cl 2: BI5 -------------NR --------------------Bsn 1: BI7 -------------NR --------------------Bsn 2: NR ------------NR --------------------Hn 1: BI8 -- BI2 ----BP2 ---------------------Hn 2: BI9 -------------

Unison P ---

Unison P -------------------------------T ------------------

92

93

94

95

96

97

98

99

100

Gesture: Unison S ---------------------------------------------------------------------------- T ---------- Unison S/Trill Rows: NR -------------------------------------------------------------------------------------- BP5 (Cl 1) NR ---------------

149 EXAMPLE 6.5: OCTET FOR WINDS (MVT. 2) ROWS A AND B, AND ICSGS ROW A:
1 2 3 5

ROW B:
1 5 6

In terms of large-scale form, most of the movements five large sections are separated by a quarter-note unison silence, with a three quarter-note silent duration between sections 3 and 4. Within sections, phrases are defined by points of imitation but with overlap blurring the phrase boundaries (in Example 6.4, the crossing arches indicate the overlap). Throughout the second movement, one can trace a process whereby a simple phrase structure is gradually altered. The following discussion of the movements form will focus on the development of the initial seven measures, paying particular attention to the interaction between shooting gestures, pointillistic bursts, and smooth threads as previously discussed. Section 1 (mm. 1-18) features a large sort of period structure. The first seven bars form a unit made of two canons, one utilizing shooting gestures, the other utilizing pointillistic bursts. As Example 6.4 shows, the shooting canon presents two separate forms of Row A, and the pointillistic canon presents several statements of a single form of Row B. Measures 8-18 follow a similar basic structure, providing a consequent phrase

150 to mm. 1-7: the section begins with shooting canon material based on forms of Row A, and concludes with canonically presented pointillistic bursts. This final statement of pointillistic bursts (labeled as Canon #5 in Example 6.4) is shown as Example 6.6.

EXAMPLE 6.6: OCTET FOR WINDS (MVT. 2) MM. 17-18, CANON #5

This section is not based on a particular row form, nor is the ICSG produced by any of the individual instruments related by identity, isomorphism or inclusion to the ICSGs of Rows A or B. However, ics 1 (presented as 11 or 13), 5 and 6 feature prominently in mm. 17-18, giving it at least a similar overall sound to the statements of BPt in mm. 6-7.

151 Section 2 (mm. 19-30) features the first stage of development of this initial phrase structure. While the first four measures introduce the lyrical band gesture in free imitation, the bulk of this section is made up of another alternation between a shooting canon based on a form of Row A (AP3) and a pointillistic canon based on several fragmentary forms of Row B. The section concludes with the first homorhythmic figures of the movement: a brief burst of pointillism followed by a long trill from the entire ensemble. Though it might be possible to understand Section 2 as a continuation of Section 1 because of the similar canonic structure, it seems clear that the primary purpose of Section 2 is the introduction of new gestural elements: the smooth threads which will be brought to the fore in Section 3, the idea of homorhythmic playing of somewhat unpredictable rhythmic patterns, and the full ensemble trill which will eventually conclude the entire piece. Section 3 (mm. 31-60) is divided into four subsections, mm. 31-38, 39-44, 45-55 and 56-60. Each of these sections recasts the structure of the different parts of Section 1, by allowing the smooth threads of the beginning of Section 2 to replace the pointillistic bursts of Section 1. The shooting canons remain constant, but are contrasted in each statement by increasingly long smooth threads. The angular pointillistic bursts have disappeared entirely. This pattern of shooting smooth threads plays out in three of the four subsections, and is expected when the fourth subsection begins in m. 56. But this expectation is not realized, as the addition of double-reed multiphonics bring Section 3 grinding to a rather alarming halt. At the beginning of Section 4 (mm. 61-77), the homorhythmic material which was introduced at the end of Section 2 returns. By this point in the piece, the earlier pointillistic bursts have completely lost the inclination to change pitch and, instead, form

152 the homorhythmic shooting gestures that play such an important role in Sections 4 and 5. These two sections have a recapitulatory function, bringing together all the primary gestures of the piece. After the opening homorhythmic passage, a freely imitative stretto brings the pointillistic bursts into close contact with smooth threads and a shooting canon as shown in Example 6.7.

EXAMPLE 6.7: OCTET FOR WINDS (MVT. 2) MM. 68-70

As the end of the piece approaches, we are confronted with the fiendishly difficult ensemble shooting patterns to which Denissova-Bruggeman referred in her writing about the piece. These shooting rhythms are interspersed with brief canonic smooth threads, before finally concluding with a long, forceful trill from the entire ensemble.

153 MULTIPLE ROWS IN THE OCTET As can be seen in Example 6.5, the two rows of Denisovs Octet for Winds share little in common in terms of interval-class content and subset content. Row A itself forms a weaving chromatic line, featuring primarily ics 1 and 2, with a single occurrence each of ic3 and ic5. On the other hand, Row B is derived entirely from [016] trichords, using exclusively ics 1, 5 and 6. 4 While the two rows both share ic1, the realization of those interval classes on the actual surface of the music is notably different: Row As five ic1s are always realized as 1, while the three ic1s in Row B are always realized as 11 or 13. This different treatment of the same interval class produces an audibly different effect. Despite the dearth of common interval-class content, the two rows share a similarity in the successions of their respective interval classes: Row Bs ICSG is an abstract subgraph of Row As ICSG. Because the visual presentation of the ICSGs in Example 6.5 does not make this relationship immediately obvious, the two graphs are redrawn in Example 6.8, with the nodes of Row Bs ICSG rearranged to make the connection clearer.

EXAMPLE 6.8: OCTET FOR WINDS (MVT. 2) ROW A AND B, ICSGS RE-DRAWN ROW A:
1 2 3 5

ROW B:
6 5 1

Not only are Row Bs discrete trichordal subsets members of set class 3-5 [016], but every trichordal subset from this row are members of set class 3-5.
4

154 The two rows share several notable features with Rows A and B of the Sonata for Alto Saxophone and Piano. Each pieces respective Row A is a winding chromatic line, made up of mostly ics 1 and 2, with a few occurrences of ics 3 and 5. They share very similar BIP-Vectors as well: [622010] is the BIP-Vector of the Sonatas Row A, and the Octet s Row A has BIP-Vector [631010]. Furthermore, their ICSGs are very closely related. Though not identical, they do share identical underlying ordinary graphs. 5 Although they can be visualized easily by imagining the ICSGs in Example 6.5 above and Example 5.9 with their arrows removed, Example 6.9 shows these two underlying graphs.

EXAMPLE 6.9: OCTET FOR WINDS, ROW A & SAXOPHONE SONATA, ROW A ICSGS AS NON-DIRECTED GRAPHS

OCTET, ROW A:
1 2 3 5

SONATA, ROW A:
1 2 3 5

In addition, each pieces respective Row B is made up of larger intervals. The Sonatas Row B is made up of leaps of ics 3, 4 and 5, and the Octets Row B is made up of leaps of 11, 13, and ics 5 and 6. While the trichordal subsets of the Sonatas Row B
5 As discussed in Appendix 2, an ordinary graph is a non-directed graph, or a graph without arrows. Although I have not explored the analytical uses of ordinary graphs in the present study, they can be used to show possible connections between interval classes, without the specificity of actual interval-class successions; an ordinary graph can only show that node x might follow and/or might be followed by node y.

155 are certainly not as regular as those of the Octets Row B, the Sonatas row does at least feature [037] subsets as three of its four discrete trichords (and three additional nondiscrete trichordal subsets, as well). The two rows yield isomorphic ICSGs, as shown in Example 6.10. Furthermore, both ICSGs are abstract subgraphs of their corresponding Row As.

EXAMPLE 6.10: OCTET FOR WINDS, ROW B & SAXOPHONE SONATA, ROW B ICSGS

OCTET, ROW B:
1 5 6

SONATA, ROW B:

Some comparable relationships can be seen between the rows of the Sonata, the Octet, and the Concerto for Guitar and Orchestra. The two rows from the Concerto are reprinted here as Example 6.11, accompanied by their INT-Cs and their ICSGs. From this Example, it can be seen that Row As ICSG is an abstract subgraph of that of Row B. 6 Clearly, the ICSG of the Concertos Row A is isomorphic to those of the Sonata and the Octets Row Bs. Composed in the same year as the Octet, the Concerto for Guitar also involves the use of a derived row: all the discrete trichords of Row A are members of set class 3-8 [026]. In both pieces, the regularity of construction of the derived row provides a
6

There are, in fact, six different mappings that can relate Row As ICSG to Row Bs ICSG. 1) 1 2, 4 4, 6 1; 2) 1 4, 4 2, 6 1; 3) 1 2, 4 6, 6 1; 4) 1 6, 4 2, 6 1; 5) 1 4, 4 6, 6 1; 6) 1 6, 4 4, 6 1.

156 contrast to the other, more intervallically varied row. As suggested in Chapter 3, this contrast can be viewed as analogous to the primary and secondary themes in a piece of EXAMPLE 6.11: CONCERTO FOR GUITAR ROWS A AND B, AND ICSGS Row A:

Row B:

tonal music. In the Concerto, Row A is treated as a static melody, with an apparent ritornello function, while Row B is used at specific formal junctions as a sort of transitional signal. In the Octet, the case is somewhat different. Initially, the rows are attached to specific gestures: Row A makes up the pitch material of the shooting canons, while Row B (or intervallic material related to Row B, as previously discussed) is used as pointillistic bursts. The consistent association between the rows and their respective gestures allows

157 us to view these rows as thematic, in the sense described by Kurbatskaya in her definition of the technique of twelve-tone rows. But these rows are used in two different ways: 1) through transformation, as the sources of pitch material of large parts of the musical fabric, and 2) in a form-generative manner. Regarding the first type of use, one can observe from the form chart shown in Example 6.4 that these two rows are used in a variety of transformations throughout the movement, and they are never used simultaneously. This allows for large sections of the movement to have the unity provided by the use of manipulated forms of a single row. As discussed in Chapter 2, Kurbatskaya makes a careful distinction between the concepts of row and series, where the former is treated somewhat like a melody, and the latter is treated like the background source of pitch material. In the Octet, Rows A and B walk a fine line between these two concepts: they are attached to a specific rhythmic/textural gestures in a manner similar to that which Kurbatskaya describes, but they also are used in a variety of transformations which recalls Kurbatskayas definition of serial technique. Secondly, the interaction of these two rows with each other and with the three primary gestures of the movement plays an important role in defining the compositions form. Overall, this piece traces a path from multiple-row serialism, to single-row serialism, to freely atonal. The first two sections of the piece establish a conflict between the two contrasting rows, each assigned to its own particular type of gesture. At the beginning of Section 2, smooth threads are heard for the first time, and each instrument presents a truncated form of Row A. These fragments of Row A are shown in Example 6.12. Eventually this gesture gives way to the familiar shooting canon, which features row form AP3. At this point in the piece, Row A seems to be asserting itself as the

158 primary row of the movement; Row A is not only continuing to dominate the shooting gestures, but also dominates the new gesture-type, smooth threads.

EXAMPLE 6.12: OCTET FOR WINDS (MVT. 2) MM. 19-22

This situation changes abruptly in Section 3. Canon #8, which begins this section, provides the expected return to the shooting gesture, but this time, the pitch material of this canon is taken from row form BPt. When the smooth threads return in m. 35, only the first bassoon presents a statement of a form of Row A; all the other instruments have moved to freely atonal material. However, given the presence of small intervals and frequent direction changes, this freely atonal music has at least a similar aural stamp to Row A. Following this last brief statement of AP5 by the bassoon, Row A disappears completely. Row B establishes its own dominance by taking over the shooting gestures and even involving itself in the smooth threads of mm. 42-44. This dominance appears to be confirmed in Section 4: from mm. 61-67, the entire ensemble presents a unison

159 pattern of irregular rhythms which project row form BP2, 7 a portion of which is shown
F F

in Example 6.13; pointillistic bursts return in mm. 68-69 and 74-77, with each instrument playing fragmented forms of Row B; a brief shooting canon based on BIt occurs in mm. 70-71.

EXAMPLE 6.13: OCTET FOR WINDS (MVT. 2) MM. 61-64

As was the case with Row A, once Row B appears to have established its primacy over the pitch material of the movement, the focus shifts. During the first six measures of Section 5, Row B gradually loses hold over the musical fabric, until it disappears

7 As Example 6.13 shows, mm. 61-67 only present order positions 4 e of BP . The first four pitches are 2 not shown here, but they are found in mm. 56-59 with the last note of horn 2s lyrical band (0), and the four-voice shooting canon in clarinet 1 (1), clarinet 2 (2), horn 1 (3), and horn 2 (4). The horn 2 pitch is carried through from the shooting canon (mm. 57-59) to the phrase shown in Example 6.13.
U U U U U U U U U U U U U U

160 entirely in m. 84. The remainder of the movement alternates between freely atonal homorhythmic shooting figures and smooth threads. The final set of smooth threads presents the interval classes involved in Row B, and even presents a fragmented version of row form BP5 in the first clarinet, but, before this row is allowed to reassert itself, a fortissimo statement of the unison shooting gesture shuts the door and ends the movement.

***

In this chapter, I have demonstrated the role played by three distinct rhythmic and textural gestures, defined by Yuri Kholopov and Valeriya Tsenova, in the second movement of Denisovs Octet for Winds. I have continued to use the ICSG developed in the earlier chapters of this dissertation, but have limited my use of it to a discussion of the relationships between the two rows of the Octet, and also to provide insight into some larger connections between the rows of the Octet, the Saxophone Sonata, and the Concerto for Guitar. These connections have revealed some general trends in Denisovs multiple-row serialism. Specifically, I have demonstrated the composers predilection for using pairs of rows with contrasting intervallic content, but with similar arrangements of interval-class successions. The remainder of this chapter focused on the interaction between the previously defined textural gestures and the changing row structure of the piece.

161 CHAPTER 7 CONCLUSIONS AND EXTENSIONS The primary goal of this study has been to formulate theories for and to present analyses from the multiple-row twelve-tone pieces in Edison Denisovs compositional output. Though multiple-row serialism makes up only a small segment of Denisovs oeuvre, it has been present in his music throughout his career. Furthermore, the use of multiple rows is noted by Peter Schmelz as a distinctly Russian trait. 1 The combination of the cultural and political separation of the young Russian composers, and Denisovs unique vision, resulted in works which simultaneously follow some of the Second Viennese Schools lead, but which move in new directions in the treatment of serial and non-serial elements. As a prelude to this discussion, it was first necessary to address some general issues. First, in order to gain an understanding of the Russian concept of serialism, Chapter 2 presented an annotated translation of a chapter from the first Russian monograph on serialism, Svetlana Kurbatskayas Serial Music: Questions of Theory, History and Aesthetics. The terms, definitions and concepts presented in Kurbatskayas work were then applied in an analysis of the first known piece of serial music to have been written in the Soviet Union, Andrey Volkonskys Musica Stricta. Second, Chapter 3 served two purposes: initially, to suggest three aspects of multiple-row serial music as fruitful points of inquiry for the analysis of such music, and subsequently to introduce the interval-class succession graph (ICSG). The first part of the chapter suggested that the analysis of multiple-row serialism should address structural and functional issues between the rows involved, and the ways in which an underlying unity might exist, despite the use of multiple rows. In the second part of
1

Schmelz 2004, 326.

162 Chapter 3 the ICSG was presented, with the intent of making clear its potential use in examining structural relationships between pairs of twelve-tone rows (or, indeed, any ordered pattern of pitch classes). The second part of this chapter, in combination with Appendix 2, presented terminology adapted from mathematical graph theory which can be used to describe the structure of the graphs formed by different twelve-tone rows. With these necessary preliminaries in place, the study presented analyses of three works from various points during Denisovs mature 2 period: chronologically, the third movement of the Sonata for Alto Saxophone and Piano (1970), the first of the Five Etudes for Solo Bassoon (1983), and the second movement of the Octet for Winds (1991). The Etude provided an appropriate opportunity to work through the tripartite approach to the analysis of music based on multiple twelve-tone rows. In the analysis of the Sonata, I used the ICSG to show not only the relationship between the two rows used in the third movement, but also to show the connections between the intervallic structure of the rows and some of the non-row-based sections of the movement. Finally, the analysis of the Octet focused primarily on the interaction between the row structure and the use of various rhythmic/textural gestures. Additionally, Chapter 6 endeavored to show commonalities between Denisovs multiple-row procedures in these two pieces, and in the Concerto for Guitar and Orchestra (1991), discussed briefly in Chapter 3. Specifically, we saw that Denisov tends to favor row pairs related by abstract inclusion which share relatively little common interval-class content.

2 Kholopov and Tsenova refer to the period from 1964-1977 as Denisovs individual style, and the time period from 1977-1996 as stabliziation, stating that the works prior to his 1964 cantata The Sun of the Incas represented a prenatal (predrodovomu) state in his compositional development. See Kholopov and Tsenova 1993, 51 and 58; 2002, 57 and 65.

163 FURTHER RESEARCH This dissertation has opened up several new avenues for potential exploration. First, the ICSG, as an analytical tool, is flexible enough to offer any number of extensions. The general concept of a succession graph can be adapted to account for successions in non-pitch domains. Example 7.1 shows the subject of an eight-voice canon (Canon #1, see Example 6.4) at the beginning of the second movement of Denisovs Octet.

EXAMPLE 7.1: OCTET FOR WINDS (MVT. 2) MM. 1-5, HORN 2 Phrase 1 Phrase 2

Beat Divisions:

10

8 Phrase 3

10

10

10

From a performers standpoint, the most difficult aspect of this passage certainly must be the constant shifting of beat subdivision required: the first beat is divided into seven parts, the second beat is divided into two parts, the first of which itself is divided into four and the second of which is divided into three. Below the example, I have placed the numbers 4, 6, 7, 8 and 10 based on the divisions of the quarter-note beat throughout the passage. Example 7.2 shows a rhythmic adaptation of the ICSG, which captures the succession of beat divisions in each of the three phrases of Example 7.1. In this case,

164 the beat-division succession graphs for the second and third phrases prove to be isomorphic to each other. Furthermore, these two graphs appear to be closely related to the graph of the first phrase. Specifically, the direction of the arrow connecting the first and last nodes has been reversed, as has the direction of the arrow connecting the two nodes at the extreme right of the graph. These three graphs all share the same underlying non-directed graph. It is not difficult to imagine using these graphs as a practice tool to be used by performers of this piece in order to practice the range of beat division successions required. EXAMPLE 7.2: OCTET FOR WINDS (MVT. 2) MM. 1-5, HORN 2, BEAT-DIVISION
SUCCESSION GRAPHS

PHRASE 1:
6 7 8 10

PHRASE 2:
6 4 8 10

PHRASE 3:
6 7 10 8

The concept of underlying non-directed graphs has come up twice in this study: once near the end of Chapter 6, and once here. We might consider exploring the relationships between succession graphs which, though not equivalent or isomorphic, share identical underlying non-directed graphs. It would be possible to define a variety of arrow operations on such graphs. We might say that, in Example 7.2, the beat-

165 division succession graphs of Phrase 1 can be transformed to the graph of Phrase 2 by two applications of an operation called arrow reversal: moving 10 8 to 8 10 and 6 10 to 10 6. Three such operations could be defined. Arrow reversal is the operation that converts the adjacency (x,y) to (y,x). Arrow addition adds an element to the set of adjacencies in the graph: if adjacency (x,y) exists, then arrow addition allows a new graph to maintain that adjacency but also to add the adjacency (y, x). Finally, arrow deletion takes a symmetric adjacency, where (x, y) and (y, x) are both members of the set of the graphs adjacencies, and eliminates one of these adjacencies. From adapting the concept of an ICSG to showing successions of different beat divisions, it is easy to imagine other generalizations of this basic idea. In fact, any series of musical events which can be described by saying this can be followed by that can be represented by a succession graph. One pedagogically useful example is presented as Example 7.3.

EXAMPLE 7.3: GRAPH OF FUNCTIONAL CHORD SUCCESSIONS IN MAJOR MODE PIECES

This familiar diagram from Stefan Kostka and Dorothy Paynes textbook Tonal Harmony represents the normative harmonic functions in major keys. 3 Though the visual presentation of this diagram is quite different than the graphs I have used throughout this paper, it should be clear that the principle is the same: arrows connect certain legal chord successions, and the lack of an arrow connecting two Roman numerals indicates
3

Reprinted from Kostka and Payne 2009, 113.

166 an illegal chord succession. As another example, we might even attempt to represent the harmonization of different bass scale degrees from Franois Campions 1716 Regle de loctave in a graph like the one shown in Example 7.4. 4

EXAMPLE 7.4: CAMPIONS REGLE DE LOCTAVE PRESENTED AS A SUCCESSION

GRAPH

5 3

6 4 3

6 3

6 5

6 4 2

5 3

6 3

6 4 3 6

6 5

6 3

^ 1

^ 2

^ 3

^ 4

^ 4

^ 5

^ 6

^ 7

^ 7

A second area of inquiry opened up by this dissertation lies in the area of the INT-C, as introduced and discussed in Chapter 3. In that chapter, it was noted that an INT-C does not necessarily define a single row; it is possible for two distinct rows to share an INT-C. This property defines a situation analogous to the Z-relation of Fortes pc-set theory. An abstract study of Z-related rows might prove to be a fruitful area of future exploration. In fact, Sebastiano Bisciglia and Jeremiah Goyette, fellow graduate students at the Eastman School of Music, are in the early stages of writing dissertations dealing with these and other related issues. 5 Third, I hope that this dissertation represents a significant step in Denisov scholarship. Even so, there is still important work to be done. Most obviously, I have only addressed one specific area of his musical style. Considered in its entirety, his

Campions Regle de loctave (Rule of the octave) was one of many ways of harmonizing ascending and descending bass scales (and scale fragments) during the 18th century. See Christensen 1992, 91. 5 Sebastiano Bisciglia, personal correspondence with the author (email, November 27, 2009).
4

167 compositional output involves a wide array of different approaches, many of which are themselves worthy of further exploration. In particular, the twelve-tone works of Denisovs so-called second conservatory demonstrate a far more traditional approach to dodecaphonic composition than the works examined in this dissertation. 6 An in-depth examination of these works might provide some insight into the composers increasing familiarity with and growing understanding of Western styles of music. It would also be worthwhile to investigate in more detail the taxonomy of different gestures defined by Kholopov and Tsenova. Specifically, the concept of smooth threads could be analyzed in detail and clearly codified, as this gesture is one found in virtually all of Denisovs music, regardless of its stylistic orientation. Finally, the relative inaccessibility of Russian-language writings on 20th-century music has severely limited the ability of English-speaking music theorists to deal with much of the music of this time period. Denisov himself was a prolific author, with essays about the music of Webern, Lutosawski, Debussy, Schoenberg, Nono, Dallapiccola, Bartks string quartets, Shostakovichs orchestration, Prokofievs use of sonata form and many other topics. Of these essays, only New Music and Jazz and The Compositional Process have been published in English, and the article On Some Melodic Types in Modern Music has been published in German. Two writings in particular, seem to demand English translation: his analysis of Weberns Piano Variations Op. 27 7 and his study On Dodecaphony and the Problems of Modern Compositional Techniques, which has many points in common with the Kurbatskaya chapter

These second conservatory works include not only the Music for Eleven Wind Instruments and Kettledrums (1961), the Piano Variations (1961), mentioned in Chapter 1, but also the Concerto for Flute, Oboe, Piano, and Percussion (1963), the Sonata for Violin (1963), and the Italian Songs (1964). 7 This article was published for the first time in 1970 in the Italian music journal Collage, in Russian and Italian.

168 translated in this study. Translations of many of these sources (or portions of those sources) would prove to be important contributions to the historiography of Soviet and Russian twelve-tone theory, and the history of twelve-tone theory and non-tonal music in general. I believe that these translations would represent an important step toward the goal of beginning a productive dialogue on the topic of Soviet serialism between English- and Russian-speaking music theorists.

169

BIBLIOGRAPHY
Babbitt, Milton. Twelve-Tone Invariants as Compositional Determinants. Musical Quarterly 46/2 (1960), 246-259. Bailey, Kathryn. The Twelve-Note Music of Anton Webern: Old Forms in a New Language. New York: Cambridge University Press, 1991. Barasch, Elena. Kontsert Edisona Denisova. In Svetdobrovechnost: Pamyati Edisona Denisova, stati, vospominaniya, material, ed. by Valeriya Tsenova, 343-368. Moscow: Moskovskaya gosudarstvennaya konservatoriya im. P. I. Chaikovskogo, 1999. Beebe, Jon P. Music for Unaccompanied Solo Bassoon: An Annotated Bibliography. Jefferson, North Carolina: McFarland and Company, Inc., 1990. Bradshaw, Susan. The Music of Edison Denisov. Tempo 151:2 (December 1984), 2-9. Buchler, Michael. Relative Saturation of Subsets and Interval Cycles as a Means for Determining Set-Class Similarity. Ph.D. dissertation, University of Rochester, 1997. Cain, Joren. The Saxophone Sonatas of Edison Denisov: A Study of Continuity. The Saxophone Symposium 24 (1999), 26-40. Chartrand, Gary, and Linda Lesniak. Graphs and Digraphs, 4th ed. Boca Raton, FL: Chapman & Hall, 2005. Child, Peter. Voice-Leading Patterns and Interval Collections in Late Shostakovich: Symphony No. 15. Music Analysis 12/1 (March 1993), 71-88. Christensen, Thomas. The Rgle de l'Octave in Thorough-Bass Theory and Practice. Acta Musicologica 64/2 (July-December 1992), 91-117. Dabney, Denise. A Multicultural Approach to Edison Denisovs Sonata for Alto Saxophone and Piano: Diverse Analytical Insights. The Saxophone Symposium, 20 (1995), 10-13. Denisov, Edison. New Music and Jazz. The World of Music, 10/3 (1968), 30-37. __________. Dodekafoniya i problem sovremennoy kompozitorskoy tekhniki. Muzka i sovremennost 6 (1969), 478-525. __________. Variatsii op. 27 dlya fortepiano A. Veberna. Collage 9 (1970), 46-71. __________. The Compositional Process. Tempo 105 (June 1973), 2-11.

170 __________. O kompozitsionnom protsesse. In Sovremennaya muzka i problem evolyutsii kompozitorskoy tekhniki, 10-18. Moscow: Sovetskiy Kompozitor, 1986. __________. Sovremennaya muzka i problem evolyutsii kompozitorskoy tekhniki. Moscow: Sovetskiy Kompozitor, 1986. __________. Novaya tekhnikaEto ne moda. In Svet-dobro-vechnost: pamyati Edisona Denisova, stati, vospominaniya, material, ed. by Valeriya Tsenova, 33-38. Moscow: Moskovskaya gosudarstvennaya konservatoriya im. P.I. Chikovskogo, 1999. Denisov, Edison, and Grigori Pantijelew. Edison Denissow im Gesprch mit Grigori Pantijelew. In Sowjetische Musik im Licht der Perestroika, ed. by Hermann Danuser, Hannelore Gerlach, and Jrgen Kchel, 242-247. Laaber: Laaber-Verlag, 1990. Denisov, Edison, and Jean-Pierre Armengaud. Entretiens avec Denisov: Un compositeur sous le regime sovietique. Paris: Edition Plume, 1993. Denisov, Edison, and Dmitriy Shulgin. Priznaniye Edisona Denisova: Po materialm besed. Moscow: Kompozitor, 1998. Fay, Laurel. Shostakovich: A Life. Oxford: Oxford University Press, 2000. Forte, Allen. The Structure of Atonal Music. New Haven: Yale University Press, 1973. Gulyanitskaya, Natalya. Vvedeniye v sovremennuyu garmoniyu. Moscow: Muzka, 1984. Haimo, Ethan. Schoenberg's Serial Odyssey: The Evolution of His Twelve-Tone Method, 19141928. New York: Oxford University Press, 1990. Hakobian, Levon. Music of the Soviet Age. Stockholm: Melos Music Literature, 1998. Hall, Patricia. The Progress of a Method: Bergs Tone Rows for Lulu. Musical Quarterly 71/4 (1985), 500-519. Haray, Frank. Graph Theory. Reading, MA: Addison-Wesley Publishing Company, 1969. (Third printing, 1972). Headlam, Dave. The Derivation of Rows in Lulu. Perspectives of New Music 24/1 (Autumn-Winter 1985), 198-233. __________. Row Derivation and Contour Association in Bergs Der Wein. Perspectives of New Music 28/1 (Winter 1990), 256-292. __________. The Music of Alban Berg. New Haven: Yale University Press, 1996. Helton, Jonathan. Edison Denisovs Sonata for Saxophone and Piano: An Analysis for the Performer. The Saxophone Symposium 25 (2000), 16-38.

171 Ilomki, Tuukka. Aspects of Pitch Organization in Schnbergs Variations for Orchestra, op. 31. Lietuvos Muzikologija 8 (2007), 30-43. __________. On the Similarity of Twelve-Tone Rows. Ph.D. dissertation, Sibelius Academy, 2008. Ho, Allan and Dmitry Feofanov. Shostakovich Reconsidered. London: Toccata Press, 1998. Jarman, Douglas. The Music of Alban Berg. Berkeley: University of California Press, 1979. Kholopov, Yuri. Andrei Volkonsky the Initiator: A Profile of His Life and Work. In Underground Music from the Former U.S.S.R., ed. by Valeriya Tsenova, 1-20. Amsterdam: Harwood Academic Publishers, 1997. Kholopov, Yuri, and Valeriya Tsenova. Edison Denisov. Moscow: Kompozitor, 1993. In English as Edison Denisov: The Russian Voice in European New Music. Translated by Romela Kohanovskaya. Studia Slavica Musicologica 28. Berlin: Ernst Kuhn, 2002. There is an earlier, less complete translation of this book: Edison Denisov. Translated by Romela Kohanovskaya. Chur, Switzerland: Harwood Academic, 1995. Kholopova, Valentina. Ob odnom printsipe khromatiki v muzke XX veka. Problem muzkalnoy nauki 2 (1973), 331-344. Kostka, Stefan and Dorothy Payne. Tonal Harmony, 6th ed. New York: McGraw Hill, 2009. Krenek, Ernst. Studies in Counterpoint Based on the Twelve-Tone Technique. New York: G. Schirmer, 1940. Kurbatskaya, Svetlana. Seriynaya muzka: vopros istorii, teorii, estetiki. Moscow: Sfera, 1996. Kurbatskaya, Svetlana, and Yuri Kholopov. Pyer Bulez-Edison Denisov: Analiticheskiye ocherki. Moscow: Sfera, 1998. Lewin, David. Generalized Musical Intervals and Transformations. New Haven: Yale University Press, 1987. __________. Klumpenhouwer Networks and Some Isographies that Involve Them. Music Theory Spectrum 12/1 (1990), 83120. Londeix, Jean-Marie. Sonate for Alto Saxophone and Piano by Edison Denisov. In Jean-Marie Londeix: Master of the Modern Saxophone, ed. by James C. Umble, 222225. New Jersey: Roncorp, 2000. Lyman, Jeffrey. After Shostakovich, What Next? New Russian Soviet Music for Bassoon. International Double Reed Society Journal 19/4 (1996), 53-67.

172 Morris, Robert. Composition with Pitch Classes: A Theory of Compositional Design. New Haven: Yale University Press, 1987. __________. Class Notes for Advanced Atonal Music Theory. Lebanon, NH: Frog Peak Music, 2001. __________. Some Musical Applications of Minimal Graph Cycles. Paper presented at the annual conference of the Society of Music Theory, Baltimore, 2007. Moseley, Brian. Getting RICH: Chains and Space in Weberns Op. 22, ii. Paper presented at the annual conference of the Society of Music Theory, Montral, 2009. Perle, George. The Music of Lulu: A New Analysis. Journal of the American Musicological Society 12/2-3 (1959), 182-200. Reich, Willi. Alban Bergs Lulu. Musical Quarterly 22 (1936), 383-401. Robinson, David and L. R. Foulds. Digraphs: Theory and Techniques. New York: Gordon and Breach, Science Publishers, Inc., 1980. Schmelz, Peter. Listening, Memory, and the Thaw: Unofficial Music and Society in the Soviet Union, 1956-1974. Ph.D. dissertation, University of California, Berkeley, 2002. __________. Shostakovich's Twelve-Tone Compositions and the Politics and Practice of Soviet Serialism. In Shostakovich and His World, ed. by Laurel Fay, 303-354. Princeton: Princeton University Press, 2004. __________. Andrey Volkonsky and the Beginnings of Unofficial Music in the Soviet Union. Journal of the American Musicological Society 58 (Spring 2005), 139-207. __________. What Was Shostakovich, and What Came Next? Journal of Musicology 24/3 (Summer 2007), 297338. __________. After Prokofiev. In Sergey Prokofiev and His World, ed. by Simon Morrison, 493-529. Princeton: Princeton University Press, 2008. __________. Such Freedom, If Only Musical. Oxford: Oxford University Press, 2009. Schoenberg, Arnold. Composition with Twelve Tones (1). In Style and Idea, ed. by Leonard Stein, 214-245. Berkeley: University of California Press, 1975. Schwarz, Boris. Arnold Schoenberg in Soviet Russia. Perspectives of New Music 4/1 (FallWinter 1965), 86-94. __________. Music and Musical Life in Soviet Russia, 1917-1970. London: Barrie and Jenkins, 1972.

173

Scott, Damon & Eric Isaacson. The Interval Angle: A Similarity Measure for PitchClass Sets. Perspectives of New Music 36/2 (1998), 107-142. Starr, Daniel Victor. Derivation and Polyphony in Twelve-Tone Music. Ph.D. dissertation, Princeton University, 1980. Taruskin, Richard. Defining Russia Musically: Historical and Hermeneutical Essays. Princeton: Princeton University Press, 1997. Tertz, Abram. On Socialist Realism. New York: Pantheon Books, 1960. Tsenova, Valeriya. Prostranstvo Edisona Denisova: K 70-Letiiu so Dnia Rozhdeniia: 1929-1996: Materialy Nauchnoi Konferentsii. Moskva: Moskovskaia gos. konservatoriia imeni P.I. Chaikovskog, 1999. Tsenova, Valeriya, ed. Svetdobrovechnost: Pamyati Edisona Denisova, stati, vospominaniya, material. Moscow: Moskovskaya gosudarstvennaya konservatoriya im. P. I. Chaikovskogo, 1999. Umble, James C. Jean-Marie Londeix: Master of the Modern Saxophone. New Jersey: Roncorp, 2000. Volkov, Solomon, ed. Testimony: The Memoirs of Dmitri Shostakovich. New York: Harper and Row, 1979. Whittall, Arnold. The Cambridge Introduction to Serialism. New York: Cambridge University Press, 2008. Wilson, Robin J. Introduction to Graph Theory. London: Longman Group, Ltd., 1975. Yakub, Roman. An Interview with Edisson Denisov. Ex Tempore 20/1 (SpringSummer 2002), 112-118. BIBLIOGRAPHY FOR APPENDIX 1 (KURBATSKAYA TRANSLATION) Babbitt, Milton. The Function of Set Structure in the Twelve-Tone System. Ph.D. dissertation, Princeton University, 1992. [Cited by Kurbatskaya as Babbitt 1946] Brindle, Reginald Smith. Serial Composition. New York: Oxford University Press, 1966. Denisov, Edison. Dodekafoniya i problem sovremennoy kompozitorskoy tekhniki. Muzka i sovremennost 6 (1969), 478-525. Forte, Allen. A Theory of Set-Complexes for Music, Journal of Music Theory 8/2 (Winter, 1964), 136-183. Gershkovich, Filipp. O Muzke: stati, zametki, pisma, vospominaniya. Sovetskiy

174 Kompozitor, 1991. Gulyanitskaya, Natalya. Vvedeniye v sovremennuyu garmoniyu. Moscow: Muzka, 1984. Kholopov, Yuri. Zadaniya po garmoniy. Moscow: Muzka, 1983. Kohoutek, Ctirad. Tekhnika kompozitsiy v muzke XX veka. Moscow: Muzka, 1976. Leibowitz, Ren. Schoenberg and His School: The Contemporary Stage of the Language of Music. New York: Da Capo Press, 1975. Perle, George. Serial Composition and Atonality; An Introduction to the Music of Schoenberg, Berg, and Webern, 2nd ed. Berkeley: University of California Press, 1968. Peyser, Joan. Twentieth-Century Music: The Sense behind the Sound. New York: Schirmer, 1980. Rognoni, Luigi. The Second Vienna School: Expressionism and Dodecaphony. London: Calder, 1977. Schffer, Boguslaw. Klasycy dodekafonii. Krakw: Polskie Wydawnictwo Muzyczne, 1964. Schoenberg, Arnold. Style and Idea, 1st ed. New York: Philosophical Library, 1950. Wellesz, Egon. The Origins of Schnbergs Twelve-Tone System. Washington DC: Library of Congress, 1958. Webern, Anton. The Path to the New Music, ed. by Willi Reich. Translated by Leo Black. Bryn Mawr, PA: Theodore Presser Co., in Association with Universal Edition. Reprinted London: Universal Edition, 1975.

175

APPENDIX 1: KURBATSKAYA TRANSLATION

Serial Music: Questions of History, Theory, and Aesthetics Svetlana Kurbatskaya [Notes on the translation: - I am deeply indebted to Dr. Peter Schmelz and Dr. Truman Bullard for their careful readings of this translation, the suggestions they offered, and the insight they provided into the mechanics of Russian translation. - The phrase vsotnaya organizatsiay translates literally as pitch organization, but I have modified this to pitch-class organization, as I believe this Western theoretical term more accurately conveys Kurbatsakayas intended meaning. Likewise, I have also translated the word vsota as pitch-class rather than the more literal pitch.] [pg. 32] Chapter 2: On the System of Terminology

The terminological system of serial and twelve-tone music evolved just as the methods of composing with the twelve-tone system did. Existing studies have developed the fundamental concepts of serialism [seriynost ] as an independent musical area and in the context of a more encompassing phenomenon (twelve-toneness) [dvenadtsatitonovost ]. When summarizing musical material from the end of the 20th century, it seems possible to encompass twelve-toneness, dodecaphony, and serialism as a unified whole, represented with an appropriate terminological system. Twelve-toneness is a property of the musical material according to which each of the twelve tones of the chromatic scale can be used as independent units of the musical fabric. Accordingly, twelve-toneness (dodecatonicism [dodekatonika], twelve-tone harmony) is a system of thinking based on the autonomy of each of the twelve pitch classes. It includes a wide spectrum of variously organized contemporary stylistic phenomena and techniques:

176 1) Free atonality a special type of pitch-class organization [vsotnaya organizatsiya] that does not reflect the standards of classical major and minor tonality, or, in general, to the centralizing action of a single foundation pitch [yedinoye ustoy i.e., tonic pitch]. New structural laws governing harmony are based on the equality of all pitch classes [vsota], and the emancipation of dissonance. 1 As examples, consider the Erwartung of A. Schoenberg, along with the Five Pieces on texts of Stephan George and the instrumental cycles Op. 6, Op. 9 and Op. 10 of A. Webern. The hemitonic (twelvesemitonal) method of symmetrical interval groups is the principle of atonal harmony, for example, in Weberns Op. 10. 2 Thus, the opening section of Piece No. 3 (Op. 10) rests on a thirteen-note complex (eleven different pitch classes), including homogeneous groups that are distributed between a melody and a chordal background:

1 For definitions of the concept of atonality and the analysis of some structural types of atonal harmony see Perle 1968 (Ch. 1); Kholopov 1983/1 (225-236), Gulyanitskaya 1984 (160-193), Filatova 1990. 2 This principle is specific to the twelve tone harmony of Webern.

177 [pg. 33] 2) Technique of tonal centers systems fashioned from the functional connections between dissonant sonic elements on the basis of an individually selected simultaneity [sozvuchiye]. One of the examples of such a pitch-class system (vsotnaya sistema) is the dissonant tonality of Skryabins late works (Op. 56) - the Skryabin scale, in which the Prometheus chord figures as the tonic, and the entire musical fabric is worked out by its regular repetitions in specific interval relationships (as in the Seventh Piano Sonata).

[pg. 34]

3) Technique of synthetic chords (N. Roslavets term) - a version of the technique of tonal centers (#2), where a chord [gruppa zvukov] functions as the source of vertical and horizontal material (in the work, or parts thereof). It is a predecessor of the dodecaphonic series, but is distinguished from it: it uses fewer than twelve notes, allows for octave doubling, and allows for changing synthetic chords during the course of the piece (N. Roslavets Three Compositions for piano, and vocal cycle Sad Landscapes).

178 4) Twelve-note chords independent chords (polychords) made up of all twelve pitch classes with or without repetitions (A. Berg, Five Orchestral Songs, Op. 4, no. 3; A. Webern, Orchestral Piece Op. 6, no. 5).

5) Technique of twelve-tone rows [12-tonovye ryad ] A row is the horizontal occurrence of non-repeating pitch classes (or, at least, with minimal repetitions), which is used as a melodic construction and is not the sole source of the musical fabric. 3 Consider the twelve-step rows [12-stupennye ryad] in the last of the Five Orchestral Songs (Op. 4) of A. Berg, No. 16 from The Polyphonic Notebook of R. Shchedrin, and the 14th Symphony of D. Shostakovich (respectively, Examples 4a, b, c.)

[pg. 35]

6) Technique of twelve-tone fields (Nikolai Obukhov, Liturgical Poems on texts by K. Balmont; A. Webern Six Bagatelles, Op. 9). A twelve-note semitonal field [12-zvukovoye
3

It is curious that melodic occurrences of twelve pitch-classes (with or without repetitions) appeared in music long before the development of the twelve-tone system. For example, the subject from the B minor fugue in Book 1 of J.S. Bachs Well-Tempered Clavier, and the introduction of F. Liszts Faust Symphony.

179 polutonovoye pole] is the set of twelve pitch classes (without or with repetitions), distributed vertically and horizontally (Kholopov 1983/1, p. 216). It concerns a special type of a twelve-tone harmonic system, where the sonic structures are formed in a mixed measurement (along a vertical line, horizontal, and also diagonal), systematically filling in the chromatic space. 4 For example, the beginning of Weberns Op. 9, No. 1 Bagatelle:

[pg. 36] 7) Technique of tropes (in the music theoretical system of J.M. Hauer) a form of twelve-tone composition based on dividing the twelve pitch-classes into two hexachords (all in all, there are 44 possible tropes). 5

8) Serial technique [seriynaya tekhnika] one of the forms of contemporary musical composition where the entire musical fabric is derived from an invariant series and its continuous recurrences (in basic and derived forms). A series is: 1) a specifically structured and constantly reproduced set of twelve (or fewer) pitch classes out of which all the sonic material is derived 6 ; 2) the ordered

The chromatic field can be incomplete that is, it can contain fewer than twelve pitch classes. For the basic conditions of the system, and examples from Hauers compositions, see Appendix 2. [Kurbatskayas Appendix 2 is not included in this translation.] 6 Let us give some existing definitions of series: 1) A series is the ordering of sonic material; the linear sequence of tones, which determines their mutual relations. (Schffer 1964, cz. 2, s. 6)
4 5

180 sequence of elements at various (including non-pitch) levels of a twelve-tone composition, fulfilling the role of the repeated invariant. This definition of series considers both a narrow and a wide-ranging concept which were uncovered in the course of the evolution of serial thinking (see below serialism, #12): narrow the series only as the basic pitch class structure; wide the series as the invariant model of sonic order at a different level (or on all levels) of a musical whole. When speaking about a pitch class series, it is necessary to note two crucial points. First, the basic value for the process (and result) of a composition does not have as much to do with the initial linear ordering of pitch classes as it does with the content of their intervallic relations. Specifically, they [the intervallic relations] realize themselves in the sonic fabric, creating the foundation for the logic governing the arrangement of pitch classes [vsotnaya logika] and, to a large extent, creating musical sense in serial works. Secondly, in the process of development of serial techniques, the regular repetition of a complete series has lost importance as an underlying principle (which is essential, for example, in the strict dodecaphony of the Schoenbergian/Webernian type). It became possible to derive the entire musical fabric from a single series by other methods in particular, by the repeating of small segments or other interval structures, which are genetically connected to the series but not taken directly from it.

2) 3) 4) 5)

A series is a group of sounds (row) from which, through its repetition, the entire fabric of the work is derived. The term series indicates the row itself (with its repetitions) and its function. (Kohoutek 1976, c. 107) A series (from Lat. series is a row, Fr. serie, German Reihe or Grundgestalt) is a row of twelve (sometimes fewer) different pitch classes, the repetition and transformations of which forms the entire musical fabric of a work. (Yuri Kholopov, article in MES 1990, s. 494) A series is a set of all twelve pitch classes from which, through its recurrence, the entire musical fabric of a piece is derived. (Kholopov 1983/1, s. 218) A series is a basic sound group which contains twelve non-repeating tones of equal importance in a strictly ordered sequence. (Gulyanitskaya, 1984, s. 196)

181 The horizontally presented series takes the form of a row (as in the theme of Edison Denisovs Variations for Piano):

[pg. 37]

Hence, some foreign theories treat the definitions of series and row (see Example 6) synonymously. 7 However, there is a fundamental difference between the two terms: a series does not appear as a surface element (like a row), but determines the appearance of relationships in the work.

9) Dodecaphony a type of serial technique in which the entire musical fabric is derived from a twelve-tone series. 8 Dodecaphony (from the Greek meaning twelve, meaning sound) literally indicates twelve sounds [dvenadtsatizvuchiye], and,
In German, the different phenomena (row and series) are designated by the word Reihe (row) and Reihentechnik, respectively. In English and American usage, before the term series appeared, beginning with M. Babbitt (1946), the term set was used (see Babbitt 1946, also Forte 1964). Being borrowed from mathematical set theory, this term has two meanings in music theory: row and also formalized collection [formalizovannaya obshchnost]. The creator of musical set theory (M. Babbitt) has combined in it the theory of sets and the morphology of the serial-dodecaphonic method (see Tsaregradskaya 1988, p. 14-18). In the works of R. Brindle and G. Perle the terms series and set are used interchangeably (see Brindle 1966, Perle 1968. 8 In our national [Russian] literature, one encounters instances of definitions of dodecaphonic serial technique. For example, in N. Gulyanitskayas Introduction to Modern Harmony, the use of dodecaphony and serial is thought of as synonymous (see Gulyanistkaya, 1984, p. 194). The discussion, however, deals with two concepts the specific and the general. This distinction is provided for in Russian and foreign terminology: Dodecaphony and serial composition; Zwlftontechnik and Reihentechnik makes sense here, as it places dodecaphony and serial composition in opposition, where Reiche [misprint: certainly must be Reihe] indicates not only twelve-tone row but also series.
7

182 at first glance, it coincides with twelve-toneness. 9 We confront the well known terminological problem of the divergence of the meanings of the two concepts, very real, not only for Russian-language studies. Thus, in foreign theory, there are pairs of corresponding terms: Zwlftonmusik and Zwlftontechnik (Ger. twelve-tone music and twelve-tone technique) 10 ; Twelve-tone composition and dodecaphony. It is logical to distinguish the two concepts in the Russian terminological system (instead of equating them through a literal translation): Twelvetone composition as a system of thought, dodecaphony as the method of composition based on the synthesis of twelve-toneness with serialism.11 As an example of deriving the musical material from an initial twelve-tone series (in the course of its transformations [vidoinzmeneniye]), consider the beginning of one of A. Weberns Drei Lieder (op. 23) on poems by Hildegard Jone. [pg. 38]

This coincidence is present, for example, in B. Schffers definition of dodecaphony: Dodecaphony is the 12-tone techniquewhich is the basis of the music, founded on the material of twelve (different!) sounds. (Schffer 1964, cz. 2, s. 5). 10 N.B.: In E. Kreneks work Studies in Twelve-Tone Counterpoint (Zwlfton-Kontrapunkt-Studien) and J. Rufers Composition with Twelve Tones (Die Komposition mit zwlf Tonen), the discussion does not deal with twelve-toneness in general, but with dodecaphony. 11 Concerning this distinction, we mean, for example, the definition of dodecaphony in the Russian translation of Ctirad Kohouteks book, which includes techniques of twelve-tone serialism, techniques of an incomplete series under the conditions of the twelve-tone system, and also the technique of twelve-tone rows in complete and incomplete chromatic fields. (see Kohoutek 1976, p. 108)
9

183 10) Microserialism a type of serial technique in which the source (invariant) of the musical fabric is a microseries, that is, a series of fewer than twelve pitch classes. For an example, consider the C#-C-E quartet of A. Webern, two Bransle dances from the ballet Agon (Example 7a) and In Memoriam Dylan Thomas by I. Stravinsky, and the Symphony-Concerto for violoncello and orchestra by B. Britten (Example 7b):

It is appropriate to compare the concept of microseries with two other concepts, also relating to the ordering of a small number of pitch classes, but displaying other shades of meaning. An incomplete series is part of a complete twelve-tone series that functions independently [of the complete series] in a specific section of the work. A subseries is a segment of a series that functions as a series on a micro-level.

[pg. 39] All methods of serial development can be applied to a sub-series: horizontal and vertical reflection [obrashcheniye horizontal reflection is retrograde, vertical reflection is inversion], transposition, permutation. The classic examples are found in the works of A. Webern (for example, the Concerto for Nine Instruments, op. 24) in which isomorphic interval groups/sub-series are derived from one intonational kernel [zerno] which

184 effectively fulfills the role as sole source of the pitch class fabric (more about this in Chapter 3). Here, consider also the series from W. Lutosawskis Funeral Music: 12

The distinction is that a microseries exists as an independent unit, but a subseries is a component of a more powerful structure (a full series).

11) Total Serialism (serial technique) [serializm] the method of creating music by means of a series of two or more parameters (see: Kogoutek 1976, p. 107, MES 1990, p. 493). For example, Messiaens le de feu uses not only a twelve-tone pitch class series, but also a series of rhythms (also twelve units), dynamics (five units) and articulations (five units):

12) Serialism [seriynost ]

12

In Example 8, small letters designate the interval forms of the subseries, similar to the forms of a series (p prime, ri retrograde inversion). As the initial subseries quite often acts in relation to other groups as the prime form of the series acts to derivative forms, we use the corresponding labels: p, i, r, ri.

185 1) In the twelve-tone system, the principle of pitch class organization in which all music, without exception, is derived from an initial sequence of non-repeating pitch classes.

[pg. 40] 2) The principle of musical thought by a series (or multiple series) [serii]. The wider meaning of serialism (and of series as a formalized order [formalizovanny poryadok]) is connected with the principle of [total] serial organization [serializatsiya] on non-pitch musical parameters. 13 Having been born during the crisis of major/minor harmonic systems, serialism was considered as an alternative to the tonal thinking of the 18th-19th centuries (hence, the comparison of a series with harmony, the theme; hence, the tendency to write with the aid of a series as with the aid of major/minor melodies and chords). 14 Examination beyond the narrow framework of pitch class organization makes it possible to reveal the influence of serialism as one of the universal categories of musical thought in the 20th century, applied to a wide range of genetically related phenomena. Depending on the sphere of application, the concepts of series and serialism display different levels in their meaning. So, in speaking about Schoenbergian serialism, we primarily mean the entire complex of serial methods then available to the composer (i.e., his serial techniques in the realm of pitch organization). Speaking about the post-classical serialism of the second half of the 20th century, we no longer mean so much concrete techniques of structural organization, as much as
Both aspects of the concept of serialism (broad and narrow) are contained in the definition of Kohoutek: Serial technique is 1) a method of music composition, based on the repetitions (and development) of a series; 2) a method of music composition which uses only a pitch-class series. (Kohoutek 1976, 107). Schffers definition is an example of a limited treatment: Serial music is music based on material of a series of twelveor fewer than twelve pitches. (see Schffer 1964, 2, 5). 14 See Schoenberg 1950, Wellesz 1958, Denisov 1969, Webern 1975, Leibowitz 1975, Rognoni 1977, Peyser 1980, Gulyanitskaya 1984, Gershkovich 1991.
13

186 different manifestations of serial thought in general. The following chapter is dedicated to the consideration of the broad sphere of questions connected with serialism.

187

APPENDIX 2: GRAPH THEORY TERMINOLOGY

This appendix serves to introduce the basic terminology used in graph theory. The vocabulary used in graph theory is large and, in many cases, varies idiosyncratically with individual authors. For example, the words node, vertex, point, element, and dot have all been used to describe the same concept. For the most part, I will use the terminology established in Harary (1972). However, I will depart from his terminology in the most basic of terms: where Haray uses the words point and arc, I will use the words node and arrow, respectively, owing to the familiarity of these terms to music theorists, due in no small part to the pioneering work of David Lewin. 1

DEFINITION A2.1: GRAPH A graph consists of a finite, non-empty set N of nodes and a collection A of ordered pairs of distinct elements of N. The elements of A are visually represented by arrows connecting the indicated nodes. 2

I am using the word graph here for what should more strictly be called a digraph. In terms used by graph theorists, a graph is a set of nodes connected by lines. A digraph is a directed graph: a set of nodes connected by arrows. Because my examples deal almost exclusively with directed graphs, I will adopt the convention established in Morris (2007) and use the word graph for purposes of simplicity. In the even that I need to refer to a non-directed graph, I will use the term ordinary graph.
1 2

See Lewin 1987, particularly chapter 9. Harary 1972, 10. This is also what Lewin calls a node/arrow system, though Lewin does not require that the ordered pairs of A be distinct elements. See Lewin 1987, 193.

188 DEFINITION A2.2: LABELED AND UNLABELED GRAPHS A labeled graph is a graph in which the nodes represent concrete musical objects (intervalclasses, in most cases, though this need not necessarily be the case), and the nodes are labeled with names describing the objects they represent. An unlabeled graph shows a system of relationships between abstract musical objects.

David Lewin has proposed a similar distinction between networks and graphs. 3 While it should be clear that I am invoking a similar distinction, I will not explicitly use Lewins terminology presently. In graph-theoretic terms, a network is a graph (with or without labeled nodes) that is associated with a function that assigns a numerical value to each of the arrows in the graph. 4 While this definition does apply to Lewins transformational networks, none of the graphs I am considering here involve such a function. In most cases, the term labeled graph will be implied when the word graph is used.

DEFINITION A2.3: ADJACENCY For all elements (n1,n2) of A, we say that node n1 is adjacent to node n2 and that node n2 is adjacent from node n1. 5 If A contains both (n1,n2) and (n2,n1), then both n1 and n2 are adjacent to each other, and we say that the adjacency is symmetric.

Lewin 1987, 195-196. Harary 1972, 50. 5 Harary 1972, 198. Other authors define adjacency in digraphs differently. In the present example, Chartrand/Lesniak 2005 would say that node n1 is adjacent to n2, but n2 is not adjacent to n1. Robinson/Foulds would say that nodes n1 and n2 are adjacent if either (n1,n2) or (n2,n1) is an element of A. See Chartrand/Lesniak 2005, 26 and Robinson/Foulds 1980, 19. When dealing with ordinary graphs, all authors are consistent in saying that if (n1,n2) is an element of A, then we say that n1 and n2 are adjacent to each other.
3 4

189 DEFINITION A2.4: INDEGREE AND OUTDEGREE The number of arrows beginning at (pointing from) a given node is that nodes outdegree and the number of arrows ending at (pointing to) a given node is that nodes indegree.

Example A2.1 shows a three-node graph, similar to the ICSGs introduced in chapter 3, which will serve to put these terms into context. EXAMPLE A2.1

ic1

ic2

ic5

In the labeled graph shown in Example A2.1, the non-empty set N of nodes are labeled interval class 1, interval class 2, and interval class 5 (ic1, ic2, ic5). The collection A consists of the following ordered pairs of nodes: (ic1,ic2), (ic2,ic1), and (ic1,ic5). The adjacency between ic1 and ic2 is symmetric, but, while ic1 is adjacent to ic5, ic5 is not adjacent to ic1. In Example A2.1, ic1 has an outdegree of 2 and an indegree of 1. Both ic2 and ic5 have an indegree of 1.

DEFINITION A2.5: SINK AND SOURCE A node with an outdegree of 0 is called a sink, and a node with an indegree of 0 is called a source.

Sources and sinks only exist in interval class succession graphs for twelve-tone rows which do not graph the rows cyclic interval, as described in Chapter 3 (page 79). A source is found when a row only has one occurrence of a particular interval class, and

190 that interval class is the first entry in that rows INT-C. When a unique interval class is the last entry in a rows INT-C, the rows ICSG will have a sink. These two special types of nodes only occur under these particular circumstances.

DEFINITION A2.6: WALK A walk is an alternating sequence of nodes and arrows, n1, (n1,n2), n2, , (nx,ny), ny. There are three types of walks: a closed walk is a walk in which the first and last nodes are the same, 6 a spanning walk is a walk which contains all the nodes, and a path is a walk in which all the nodes are distinct.

DEFINITION A2.7: REACHABLE, CONNECTED GRAPH If there is some path from n1 to n2, then we say that node n2 is reachable from node n1. Furthermore, a graph is connected if, for every pair of nodes, nx and ny, there is either a path from nx to ny, or from ny to nx, or both. 7

DEFINITION A2.8: SYMMETRIC GRAPH A graph is called symmetric if, for every (nx,ny) in A, (ny,nx) is also in A.

A few examples of some Second Viennese School rows will serve to acquaint the reader with the terms from Definitions A2.5 A2.8. Example A2.2A shows the ICSG

Robert Morris refers to this as a cycle, but a cycle is usually defined as a specific kind of closed walk within a digraph. Harary 1972 defines a cycle as a nontrivial closed walk with all points distinct (except the first and last) (emphasis mine). Morris 2007 does not observe this requirement. 7 This definition essentially considers the digraph as an ordinary graph, with lines connecting nodes instead of arrows. Harary actually defines three different types of digraph connectedness: strongly connected, unilaterally connected, and weakly connected. My definition corresponds to Hararys definition of weakly connected digraphs. See Harary 1972, 199.
6

191 for the primary row of Bergs Lyric Suite. This is an example of a symmetric graph: all the arrows are double-ended. As can be seen in Example A2.2A, the INT-C for the Lyric Suite row forms a palindrome. It may seem that this palindromic INT-C is the reason for the symmetric nature of the ICSG, but this is not precisely the case. Example A2.2B shows the ICSG for the row of Schoenbergs Violin Fantasy, Op. 47. Like the Lyric Suite ICSG, this graph is also symmetric, but the INT-C associated with this row is not a palindrome. While it is true that every row with a palindromic INT-C has a symmetric ICSG, Example A2.2B has shown that it is not true that every row with a symmetric ICSG has a palindromic INT-C. The primary row from Bergs Lulu is shown in Example A2.2C. The ic4 node in this ICSG is a source. Example A2.2D shows the row and ICSG for Schoenbergs Wind Quintet, Op. 26. In this graph, the ic3 node is a sink.

EXAMPLE A2.2: VARIOUS ICSGS A. BERG, LYRIC SUITE (PRIMARY ROW) Row: < 5 4 0 9 7 2 8 1 3 6 t e > INT-C: < 1 4 3 2 5 6 5 2 3 4 1 >

B. SCHOENBERG, VIOLIN FANTASY, OP. 47 Row: < t 9 1 e 5 7 3 4 0 2 8 6 > INT-C: < 1 4 2 6 3 4 1 4 2 6 2 > C. BERG, LULU (PRIMARY ROW) Row: < 0 4 5 2 7 9 6 8 e t 3 1 > INT-C: < 4 1 3 5 2 3 2 3 1 5 2 > D. SCHOENBERG, WIND QUINTET, OP. 26 Row: < 3 7 9 e 1 0 t 2 4 6 8 5 > INT-C: < 4 2 2 2 1 2 4 2 2 2 3 >